Rivi?res,fleuves

in VML) is created, and when the drawPoint function is called with a specific style for the point, it references the specified symbol with in SVG and in VML. http://dev.openlayers.org/sandbox/camptocamp/vectorSymbols/examples/vector-symbols.html We are thinking about creating a single library, in order to be generic and so that people can add their own symbols in one file and not both SVG and VML. Problems we meet are : - We would need to draw complex symbols. SVG element allows it since several or predefined can be included inside but in VML, the element isn't so permissive since we can't include several shapes or groups of elements inside. So only simple symbols (one shape, color, stroke-width, etc) would be allowed... - which language is better to write the single library : SVG since SVG symbols are quite spread on the web (with problem above in mind) or javascript to have a generic way to encode symbols, or another language? - A single library involves to develop a way to transform SVG into VML or a certain language into SVG and VML, and there are significant differences between the two (shapetype vs symbol, path element accepting decimal numbers in SVG and not in VML, etc) Any ideas about these problems? So do you think it's worth it or offering predefined symbols and letting people adding their code into SVG library and VML library would be satisfactory? Regards, Barbara From ian at planetmayo.com Thu Oct 18 13:14:31 2007 From: ian at planetmayo.com (Ian Mayo) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Developement of an external vector symbol Message-ID: Barbara, I'm afraid I can't offer an opinion on your languages challenge. It would be great if there was a pre-existing file-format for vector symbols, then we just specify filenames rather than hard-code the vectored shapes. I'm currenlty working on provision of per-entry icons (images) for georss feeds. [1]. I've made good progress, but may have to pause my efforts for the next 7 days, see my demo at: [2]. Had you considered using the symbols on a per-entry fashion in a georss feed? I suspect it's fortunate I've got a self-induced development pause: it would seem to be worth offering per-entry symbols as well as icons. Appreciate your thoughts/opinions. Ian. 1. http://trac.openlayers.org/ticket/1073 2. http://dev.openlayers.org/sandbox/ianmayo/branches/rss/openlayers/examples/georss.html -- Ian Mayo PlanetMayo Ltd From erik.uzureau at metacarta.com Thu Oct 18 13:22:03 2007 From: erik.uzureau at metacarta.com (Erik Uzureau) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <926726.10161.qm@web51402.mail.re2.yahoo.com> References: <926726.10161.qm@web51402.mail.re2.yahoo.com> Message-ID: <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> I like this idea a lot. we make a /lang directory and a separate file for each language. also the use of JSON I think is crafty. the file could look something like: OpenLayers.Strings['english'] = { 'foo': "Foo for president", 'bar': "Bar for gardner" }; OpenLayers.String.setLanguage = function(lang) { OpenLayers.String._language = lang; } OpenLayers.String.get = function(strKey) { return OpenLayers.String[OpenLayers.String._language][strKey]; } and so: OpenLayers.getString('foo'); or OpenLayers.String.get('foo'); this is off the top of my head. On 10/18/07, Dan Little wrote: > > I wouldn't think this to be too hard. > > In most internationalized applications I've seen, there is typically a > "lang" or "strings" file that contains all of the strings for a given > application. Then when the code needs a string it will call a global > variable containing that error/warning/info message. This way all of the > needed translation is isolated to a single file. Changing languages becomes > a matter of loading an alternate language file. > > So there could be a collection of language files: > strings.en.js > strings.es.js > strings.fr.js > strnigs.de.js > > And then provide some basic facilities for switching between them. In fact, > each of those files could be a JSON object that gets loaded to a global > object/variable. > > I may be just spewing bad ideas... but so long as the creative juices were > flowing I'd thought I'd share. > > -Duck > > > ----- Original Message ---- > From: Jachym Cepicky > To: Bart van den Eijnden (OSGIS) > Cc: "dev@openlayers" > Sent: Thursday, October 18, 2007 8:58:42 AM > Subject: Re: [OpenLayers-Dev] status of translation possibilities > > Bart van den Eijnden (OSGIS) p??e v ?t 18. 10. 2007 v 15:44 +0200: > > Hi Chris, > > > > I think it is acceptable to have a global setting. I can't imagine a > > common use case (for myself) where one map would be in English and > > another one in Dutch on the same page. It is important though that it is > > possible to switch languages. > > > +1 > > jachym > > > Best regards, > > Bart > > > > Christopher Schmidt schreef: > > > On Thu, Oct 18, 2007 at 03:23:50PM +0200, Jachym Cepicky wrote: > > > > > >> Hi, > > >> > > >> I would like to ask about the status of OpenLayers language support. > > >> There are not many strings, which could be translated, however, some > are > > >> there and IMHO we should (de)find a way, how to make this possible. Was > > >> there some discussion to this topic? > > >> > > >> IMHO, new map attribute "lang" should be defined, as well as new method > > >> OpenLayer.Util.getOLString("English text") (or similar). > > >> > > >> Any suggestions? Should I write something like this? Is there already > > >> some code out there? > > >> > > > > > > No code yet, but I agree that now is the time. > > > > > > > http://trac.openlayers.org/wiki/RFC/Internationalization > > > > > > Is my start at an RFC, which never went far enough. > > > > > > My belief is that we should make this a global setting, rather than a > > > per-map configuration. This would mean you could not have multiple > > > languages on the same page at the same time, but would also mean you > > > don't need a reference to the map object in order to get a correct > > > translation string. I don't know if the trade off is worth it. What do > > > other people think? > > > > > > Regards, > > > > > > > > -- > Jachym Cepicky > e-mail: jachym.cepicky@gmail.com > URL: http://les-ejk.cz > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > > > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around > http://mail.yahoo.com > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > From erik.uzureau at metacarta.com Thu Oct 18 13:23:27 2007 From: erik.uzureau at metacarta.com (Erik Uzureau) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> Message-ID: <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> or maybe to keep the namespaces straight and clean, more like this: OpenLayers.Strings.languages['english'] = { 'foo': "Foo for president", 'bar': "Bar for gardner" }; OpenLayers.String.setLanguage = function(lang) { OpenLayers.Strings.language = lang; } OpenLayers.String.get = function(strKey) { return OpenLayers.Strings.languages[OpenLayers.String._language][strKey]; } and so: OpenLayers.getString('foo'); or OpenLayers.String.get('foo'); On 10/18/07, Erik Uzureau wrote: > I like this idea a lot. we make a /lang directory and a separate file for > each language. also the use of JSON I think is crafty. > > the file could look something like: > > OpenLayers.Strings['english'] = { > 'foo': "Foo for president", > 'bar': "Bar for gardner" > }; > > OpenLayers.String.setLanguage = function(lang) { > OpenLayers.String._language = lang; > } > > > OpenLayers.String.get = function(strKey) { > return OpenLayers.String[OpenLayers.String._language][strKey]; > } > > > and so: > > > OpenLayers.getString('foo'); > > or > > OpenLayers.String.get('foo'); > > > > this is off the top of my head. > > > On 10/18/07, Dan Little wrote: > > > > I wouldn't think this to be too hard. > > > > In most internationalized applications I've seen, there is typically a > > "lang" or "strings" file that contains all of the strings for a given > > application. Then when the code needs a string it will call a global > > variable containing that error/warning/info message. This way all of the > > needed translation is isolated to a single file. Changing languages becomes > > a matter of loading an alternate language file. > > > > So there could be a collection of language files: > > strings.en.js > > strings.es.js > > strings.fr.js > > strnigs.de.js > > > > And then provide some basic facilities for switching between them. In fact, > > each of those files could be a JSON object that gets loaded to a global > > object/variable. > > > > I may be just spewing bad ideas... but so long as the creative juices were > > flowing I'd thought I'd share. > > > > -Duck > > > > > > ----- Original Message ---- > > From: Jachym Cepicky > > To: Bart van den Eijnden (OSGIS) > > Cc: "dev@openlayers" > > Sent: Thursday, October 18, 2007 8:58:42 AM > > Subject: Re: [OpenLayers-Dev] status of translation possibilities > > > > Bart van den Eijnden (OSGIS) p??e v ?t 18. 10. 2007 v 15:44 +0200: > > > Hi Chris, > > > > > > I think it is acceptable to have a global setting. I can't imagine a > > > common use case (for myself) where one map would be in English and > > > another one in Dutch on the same page. It is important though that it is > > > possible to switch languages. > > > > > +1 > > > > jachym > > > > > Best regards, > > > Bart > > > > > > Christopher Schmidt schreef: > > > > On Thu, Oct 18, 2007 at 03:23:50PM +0200, Jachym Cepicky wrote: > > > > > > > >> Hi, > > > >> > > > >> I would like to ask about the status of OpenLayers language support. > > > >> There are not many strings, which could be translated, however, some > > are > > > >> there and IMHO we should (de)find a way, how to make this possible. Was > > > >> there some discussion to this topic? > > > >> > > > >> IMHO, new map attribute "lang" should be defined, as well as new method > > > >> OpenLayer.Util.getOLString("English text") (or similar). > > > >> > > > >> Any suggestions? Should I write something like this? Is there already > > > >> some code out there? > > > >> > > > > > > > > No code yet, but I agree that now is the time. > > > > > > > > > > http://trac.openlayers.org/wiki/RFC/Internationalization > > > > > > > > Is my start at an RFC, which never went far enough. > > > > > > > > My belief is that we should make this a global setting, rather than a > > > > per-map configuration. This would mean you could not have multiple > > > > languages on the same page at the same time, but would also mean you > > > > don't need a reference to the map object in order to get a correct > > > > translation string. I don't know if the trade off is worth it. What do > > > > other people think? > > > > > > > > Regards, > > > > > > > > > > > > -- > > Jachym Cepicky > > e-mail: jachym.cepicky@gmail.com > > URL: http://les-ejk.cz > > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > > > > > > > > __________________________________________________ > > Do You Yahoo!? > > Tired of spam? Yahoo! Mail has the best spam protection around > > http://mail.yahoo.com > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev > > > > > From erik.uzureau at metacarta.com Thu Oct 18 13:28:54 2007 From: erik.uzureau at metacarta.com (Erik Uzureau) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jslint cleanup committed In-Reply-To: <20071018130448.GB14361@metacarta.com> References: <20071018130448.GB14361@metacarta.com> Message-ID: <6ae3fb590710181028q6c4c1fc9g6b2f7a2d750c9761@mail.gmail.com> I agree that the firebug files we should leave as-is so we can make upgrading easier (though it would certainly be worth posting a cleanup patch to their project. do onto others...) regarding the rico bits, I say "yes". We've modified them pretty significantly anyways, and I think we're now using them for everything we'll ever need from them. Which is to say, the only reason they're in there is for rounded corners, and I don't anticipate great leaps of innovation in the world of corner rounding. So yeah, let's go ahead and clean that up. Thanks so much for your work on this, Paul. r5002 is a great patch, a non-functional but professionalizing step forward. Muito muito obrigado!!! Erik On 10/18/07, Christopher Schmidt wrote: > On Thu, Oct 18, 2007 at 09:01:57AM -0400, Paul Spencer wrote: > > I have added missing semi-colons and curly braces around all block > > scopes in: > > > > http://trac.openlayers.org/changeset/5002/ > > > > jslint reports the remaining warnings: > > > > /lib/Firebug/firebug.js: 0 error(s), 45 warning(s) > > /lib/Firebug/firebugx.js: 0 error(s), 2 warning(s) > > /lib/Rico/Color.js: 0 error(s), 18 warning(s) > > /lib/Rico/Corner.js: 0 error(s), 36 warning(s) > > > > Do you want me to do anything about these? > > I would prefer we leave the firebug files alone -- I don't think we've > touched those at all. > > I have no strong feelings on Rico: we've already changed those files to > put them in the OpenLayers namespace, further changes don't bother me > much, but I'm also happy to just leave them. > > Regards, > -- > Christopher Schmidt > MetaCarta > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > From erik.uzureau at metacarta.com Thu Oct 18 13:38:23 2007 From: erik.uzureau at metacarta.com (Erik Uzureau) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Developement of an external vector symbol library In-Reply-To: <47177BD1.2010508@camptocamp.com> References: <47177BD1.2010508@camptocamp.com> Message-ID: <6ae3fb590710181038u1b66ca7eld29382125278e9d7@mail.gmail.com> Dear Barbara, This project sounds *excellent*. This is just the sort of thing that OpenLayers needs in order to step it up to the next level. In a previous thread, I have proposed the creation of an Addins directory, which all things going right, should probably be together within the next week or so. I think your proposal for an OpenLayers Icons Repository would be a *great* candidate for the Addins directory. I'm sure many people developing OL apps would love to have access to these symbols. Erik On 10/18/07, Barbara Philippot wrote: > Hi, > > > We are trying to improve vector rendering for points geometry, allowing > users to display symbols such as a square, a triangle, a plane, a church > and so on instead of just a circle. > > At present, we have a library in SVG (using the element) and a > library in VML (using the element). Symbols are loaded when > the root element

in VML) is created, and when the > drawPoint function is called with a specific style for the point, it > references the specified symbol with in SVG and > in VML. > > http://dev.openlayers.org/sandbox/camptocamp/vectorSymbols/examples/vector-symbols.html > > We are thinking about creating a single library, in order to be generic > and so that people can add their own symbols in one file and not both > SVG and VML. > > Problems we meet are : > > - We would need to draw complex symbols. SVG element allows it > since several or predefined can be included inside but in > VML, the element isn't so permissive since we can't include > several shapes or groups of elements inside. So only simple symbols (one > shape, color, stroke-width, etc) would be allowed... > > - which language is better to write the single library : SVG since SVG > symbols are quite spread on the web (with problem above in mind) or > javascript to have a generic way to encode symbols, or another language? > > - A single library involves to develop a way to transform SVG into VML > or a certain language into SVG and VML, and there are significant > differences between the two (shapetype vs symbol, path element accepting > decimal numbers in SVG and not in VML, etc) > > Any ideas about these problems? > > So do you think it's worth it or offering predefined symbols and letting > people adding their code into SVG library and VML library would be > satisfactory? > > > Regards, > > Barbara > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > From crschmidt at metacarta.com Thu Oct 18 14:31:04 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> Message-ID: <20071018183104.GB31939@metacarta.com> On Thu, Oct 18, 2007 at 12:23:27PM -0500, Erik Uzureau wrote: > or maybe to keep the namespaces straight and clean, more like this: > > OpenLayers.Strings.languages['english'] = { > 'foo': "Foo for president", > 'bar': "Bar for gardner" > }; This seems fine to me. > OpenLayers.String.setLanguage = function(lang) { > OpenLayers.Strings.language = lang; > } setLanguage, if we're going to provide a method for it, is going to need to be more complex than that. We'd have to loop through the controls, and call redraw() on them, to update their text, and I'm not sure that all controls would support a redraw correctly. > OpenLayers.String.get = function(strKey) { > return OpenLayers.Strings.languages[OpenLayers.String._language][strKey]; > } It needs to be more complex than this to support interpolation, and I think it will need to: as the RFC says: "The OpenLayers.String object will have an additional function, translate. Translate will accept a translation key -- a single string which can be used to look up the string -- and any number of additional arguments. If additional arguments are passed, they will be interpolated in the string in order." MapBuilder already has some code that does this, so we can work with them to integrate those bits of it into OpenLayers. It seems there is a general agreement that we need to do this. There are more technical details to work out: * What is the expectation we provide for updating existing langauge strings in the code? For example, controls will probably soon have the ability to display titles in a control.panel object. However, the title property is set on the div when it's created. Do we need to make a redraw() that works for all controls that would use text? Does this block integration of i18n support into core? (See also: CSS support for display, which we didn't block on.) * Do we block putting the library into core until we have all the strings pulled out? (I think that given how little text we have, I think this is 'yes'.) * Do we block all future commits that include text on getting it using the translation system? I think the answer to this is also yes. How far do we go before we put code in trunk? I think that some people who know more about i18n than me should start pondering all the things that an english speaker like me is going to screw up :) MapBuilder's experience here is, again, instructive. Regards, -- Christopher Schmidt MetaCarta From jachym.cepicky at gmail.com Thu Oct 18 15:08:25 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <20071018183104.GB31939@metacarta.com> References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> Message-ID: <1192734505.8316.30.camel@mellon> Hi, Christopher Schmidt p??e v ?t 18. 10. 2007 v 14:31 -0400: > On Thu, Oct 18, 2007 at 12:23:27PM -0500, Erik Uzureau wrote: > > or maybe to keep the namespaces straight and clean, more like this: > > > > > > OpenLayers.Strings.languages['english'] = { > > 'foo': "Foo for president", > > 'bar': "Bar for gardner" > > }; This is IMHO bad, because it makes things untransparent. I would suggest using whole english strings as key: OpenLayers.Strings.languages['czech'] = { 'Foo for president': "Foo prezidentem!", 'Bar for gardner': "Bar zahradn?kem!" }; more comments below > This seems fine to me. > > > OpenLayers.String.setLanguage = function(lang) { > > OpenLayers.Strings.language = lang; > > } > > setLanguage, if we're going to provide a method for it, is going to need > to be more complex than that. We'd have to loop through the controls, > and call redraw() on them, to update their text, and I'm not sure that > all controls would support a redraw correctly. > > > OpenLayers.String.get = function(strKey) { > > return OpenLayers.Strings.languages[OpenLayers.String._language][strKey]; > > } > > It needs to be more complex than this to support interpolation, and I > think it will need to: as the RFC says: > > "The OpenLayers.String object will have an additional function, > translate. Translate will accept a translation key -- a single string > which can be used to look up the string -- and any number of additional > arguments. If additional arguments are passed, they will be interpolated > in the string in order." > > MapBuilder already has some code that does this, so we can work with > them to integrate those bits of it into OpenLayers. > > It seems there is a general agreement that we need to do this. There are > more technical details to work out: > > * What is the expectation we provide for updating existing langauge > strings in the code? For example, controls will probably soon have > the ability to display titles in a control.panel object. However, the > title property is set on the div when it's created. Do we need to > make a redraw() that works for all controls that would use text? Does > this block integration of i18n support into core? (See also: CSS > support for display, which we didn't block on.) +0 > * Do we block putting the library into core until we have all the > strings pulled out? (I think that given how little text we have, I > think this is 'yes'.) -1 > * Do we block all future commits that include text on getting it using > the translation system? I think the answer to this is also yes. +1 > > How far do we go before we put code in trunk? There should be just general support for this. Methods for setting the language of the map/OpenLayers (Something like ProxyHost) and support for translations files. Translation is a job, which will never be marked as "done". It will always change, the translators never translate everything, and so on. > > I think that some people who know more about i18n than me should start > pondering all the things that an english speaker like me is going to > screw up :) MapBuilder's experience here is, again, instructive. Few more comments: I agree, that there are only few strings in current version. However, this is a job, which will never be marked as done :) Also, one has to distinguish between debugging messages nad standard user messages/strings. Also translation will never be completed just my 2 cents jachym -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071018/bc3ebe47/attachment.bin From pspencer at dmsolutions.ca Thu Oct 18 15:46:21 2007 From: pspencer at dmsolutions.ca (Paul Spencer) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <20071018183104.GB31939@metacarta.com> References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> Message-ID: On 18-Oct-07, at 2:31 PM, Christopher Schmidt wrote: > setLanguage, if we're going to provide a method for it, is going to > need > to be more complex than that. We'd have to loop through the controls, > and call redraw() on them, to update their text, and I'm not sure that > all controls would support a redraw correctly. > Could we use an event that is triggered and then controls that have strings can register (or be registered) for the language changing? OpenLayers.String.events.register('languagechange', OpenLayers.Function.bind(this, this.onLanguageChange)); Then it is a matter of implementing support in specific controls that need it or can use it. And if there are things other than controls that can use the current language, they can take advantage easily. For instance, some layer types might be able to use the current language for labeling. And applications (or application frameworks) around OpenLayers could also take advantage of it. We will probably want a LanguageSwitcher control that sets the language and tracks the current language too, mostly for testing and development. We need a way of indicating what languages to support in any given application, and of enumerating the supported languages through the OpenLayers API. We need a way to extend the base set of translations with other, application-specific translations. Cheers Paul +-----------------------------------------------------------------+ |Paul Spencer pspencer@dmsolutions.ca | +-----------------------------------------------------------------+ |Chief Technology Officer | |DM Solutions Group Inc http://www.dmsolutions.ca/ | +-----------------------------------------------------------------+ From crschmidt at metacarta.com Thu Oct 18 16:01:13 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> Message-ID: <20071018200113.GA7249@metacarta.com> On Thu, Oct 18, 2007 at 03:46:21PM -0400, Paul Spencer wrote: > > On 18-Oct-07, at 2:31 PM, Christopher Schmidt wrote: > > >setLanguage, if we're going to provide a method for it, is going to > >need > >to be more complex than that. We'd have to loop through the controls, > >and call redraw() on them, to update their text, and I'm not sure that > >all controls would support a redraw correctly. > > > > Could we use an event that is triggered and then controls that have > strings can register (or be registered) for the language changing? > > OpenLayers.String.events.register('languagechange', > OpenLayers.Function.bind(this, this.onLanguageChange)); I like this idea. I hadn't thought of it, and it makes sense. > We will probably want a LanguageSwitcher control that sets the > language and tracks the current language too, mostly for testing and > development. I'm not convinced that we need that, but if it gets built as part of development, it would probably be useful. > We need a way of indicating what languages to support in any given > application, and of enumerating the supported languages through the > OpenLayers API. for (var key in OpenLayers.Strings.languages) { alert(key); } > We need a way to extend the base set of translations with other, > application-specific translations. OpenLayers.Strings.languages['eng']['My Application String!'] = 'My application String!'; OpenLayers.Strings.languages['spa']['My Application String!'] = '¡String Applicacion de mio!'; Regards, -- Christopher Schmidt MetaCarta From arnd.wippermann at web.de Thu Oct 18 16:09:19 2007 From: arnd.wippermann at web.de (Arnd Wippermann) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] vector layer does not draw at lower resolutions. In-Reply-To: <13219517.post@talk.nabble.com> References: <13219517.post@talk.nabble.com> Message-ID: <13282676.post@talk.nabble.com> Use this map.addLayers([layer,vlayer]); instead of map.addLayer(layer,vlayer); with your code only one layer will added to the map. Arnd Wippermann http://gis.ibbeck.de/ginfo -- View this message in context: http://www.nabble.com/vector-layer-does-not-draw-at-lower-resolutions.-tf4629642.html#a13282676 Sent from the OpenLayers Dev mailing list archive at Nabble.com. From erik.uzureau at metacarta.com Thu Oct 18 17:24:21 2007 From: erik.uzureau at metacarta.com (Erik Uzureau) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <20071018183104.GB31939@metacarta.com> References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> Message-ID: <6ae3fb590710181424n51ff8a76y6b05248644cfbd6f@mail.gmail.com> On 10/18/07, Christopher Schmidt wrote: > On Thu, Oct 18, 2007 at 12:23:27PM -0500, Erik Uzureau wrote: > > or maybe to keep the namespaces straight and clean, more like this: > > > > > > OpenLayers.Strings.languages['english'] = { > > 'foo': "Foo for president", > > 'bar': "Bar for gardner" > > }; > > This seems fine to me. > > > OpenLayers.String.setLanguage = function(lang) { > > OpenLayers.Strings.language = lang; > > } > > setLanguage, if we're going to provide a method for it, is going to need > to be more complex than that. We'd have to loop through the controls, > and call redraw() on them, to update their text, and I'm not sure that > all controls would support a redraw correctly. > > > OpenLayers.String.get = function(strKey) { > > return OpenLayers.Strings.languages[OpenLayers.String._language][strKey]; > > } > > It needs to be more complex than this to support interpolation, and I > think it will need to: as the RFC says: > > "The OpenLayers.String object will have an additional function, > translate. Translate will accept a translation key -- a single string > which can be used to look up the string -- and any number of additional > arguments. If additional arguments are passed, they will be interpolated > in the string in order." > > MapBuilder already has some code that does this, so we can work with > them to integrate those bits of it into OpenLayers. ah... you mean, like printf stylie? > It seems there is a general agreement that we need to do this. There are > more technical details to work out: > > * What is the expectation we provide for updating existing langauge > strings in the code? For example, controls will probably soon have > the ability to display titles in a control.panel object. However, the > title property is set on the div when it's created. Do we need to > make a redraw() that works for all controls that would use text? Does > this block integration of i18n support into core? (See also: CSS > support for display, which we didn't block on.) > * Do we block putting the library into core until we have all the > strings pulled out? (I think that given how little text we have, I > think this is 'yes'.) > * Do we block all future commits that include text on getting it using > the translation system? I think the answer to this is also yes. > > How far do we go before we put code in trunk? > > I think that some people who know more about i18n than me should start > pondering all the things that an english speaker like me is going to > screw up :) MapBuilder's experience here is, again, instructive. > > Regards, > -- > Christopher Schmidt > MetaCarta > From erik.uzureau at metacarta.com Thu Oct 18 17:29:41 2007 From: erik.uzureau at metacarta.com (Erik Uzureau) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <20071018200113.GA7249@metacarta.com> References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> <20071018200113.GA7249@metacarta.com> Message-ID: <6ae3fb590710181429n182c547bld43f7be10bbcaedc@mail.gmail.com> On 10/18/07, Christopher Schmidt wrote: > On Thu, Oct 18, 2007 at 03:46:21PM -0400, Paul Spencer wrote: > > > > On 18-Oct-07, at 2:31 PM, Christopher Schmidt wrote: > > > > >setLanguage, if we're going to provide a method for it, is going to > > >need > > >to be more complex than that. We'd have to loop through the controls, > > >and call redraw() on them, to update their text, and I'm not sure that > > >all controls would support a redraw correctly. > > > > > > > Could we use an event that is triggered and then controls that have > > strings can register (or be registered) for the language changing? > > > > OpenLayers.String.events.register('languagechange', > > OpenLayers.Function.bind(this, this.onLanguageChange)); > > I like this idea. I hadn't thought of it, and it makes sense. yes, i like it to. the only thing i dont like about it is that it seems like it might sort of be a burden to control developers to have to deal with registering events. to that end, we might consider such a thing as making the base OpenLayers.Control's constructor register the listener and call a stub function udpateLanguage() or something like that. In OpenLayers.Control, the stub function would do nothing. Then if controls _do_ wish to update on language changes, they need only implement the updateLanguage(). This might help in the case of a control that is subclassed from another control -- the updateLanguage() calls would be chained up, instead of having to register a separate listener for each class. I dunno. This is just the first thing that comes to mind... that as a control developer it seems like kind of a PITA and or unnecessary overhead to be registering more listeners. --E > > > We will probably want a LanguageSwitcher control that sets the > > language and tracks the current language too, mostly for testing and > > development. > > I'm not convinced that we need that, but if it gets built as part of > development, it would probably be useful. > > > We need a way of indicating what languages to support in any given > > application, and of enumerating the supported languages through the > > OpenLayers API. > > for (var key in OpenLayers.Strings.languages) { > alert(key); > } > > > We need a way to extend the base set of translations with other, > > application-specific translations. > > OpenLayers.Strings.languages['eng']['My Application String!'] = 'My > application String!'; > > OpenLayers.Strings.languages['spa']['My Application String!'] = > '?String Applicacion de mio!'; > > Regards, > -- > Christopher Schmidt > MetaCarta > From gertjan at West.NL Fri Oct 19 07:05:18 2007 From: gertjan at West.NL (Gertjan van Oosten) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <6ae3fb590710181424n51ff8a76y6b05248644cfbd6f@mail.gmail.com> <20071018183104.GB31939@metacarta.com> <17AD4DD1-AF6F-44AE-9D72-3D6CFB1EB2F0@dmsolutions.ca> <20071018133356.GA19498@metacarta.com> References: <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> <926726.10161.qm@web51402.mail.re2.yahoo.com> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <17AD4DD1-AF6F-44AE-9D72-3D6CFB1EB2F0@dmsolutions.ca> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> Message-ID: <20071019110518.GA5597@West.NL> As quoted from Christopher Schmidt : > It needs to be more complex than this to support interpolation, and I > think it will need to: as the RFC says: [ http://trac.openlayers.org/wiki/RFC/Internationalization ] > "The OpenLayers.String object will have an additional function, > translate. Translate will accept a translation key -- a single string > which can be used to look up the string -- and any number of additional > arguments. If additional arguments are passed, they will be interpolated > in the string in order." > > MapBuilder already has some code that does this, so we can work with > them to integrate those bits of it into OpenLayers. As I wrote on the I18N thread mid-July (see below), OpenLayers can just copy the interpolation function I wrote for MapBuilder. So interpolation capabilities are more or less a given. The only question remaining is how to get the i18n'ed strings into (and out of) the Openlayers code. I have never come across a real-life use case where you would want i18n data to be displayed in mixed-language in one application, so a single, global language setting would suffice for the majority of cases. Note that even when you have i18n texts, you still need to take into account that the developer adding new strings to the code and the people doing the translation thereof into the other supported languages are generally not the same. Which means you need some sort of fall-back mechanism for message keys that remain as yet untranslated. In MapBuilder the fall back is on the default language "en". To start off, I guess simple language tags ("en", "nl", "fr", ...) will suffice. If and when people want to support language subtags ("en-GB", "en-US", "nl-NL", "nl-BE", "fr-FR", "fr-CA", ...) you should also incorporate a fall back from the subtagged language on the more general language tag. For MapBuilder this is still on my TODO list; there has been no demand for it yet. [See http://communitymapbuilder.org/display/MAP/Add+Translations for complete information on how MapBuilder does things.] Anyway, to quote myself on interpolation ;-) : [...] the function mbFormatMessage() I wrote, copied below from trunk/lib/util/Util.js for your convenience, can be used straight out-of-the-box to get simple interpolation. Since at the top of the file it clearly states that it is LGPL, nothing is stopping you. 8-) /** * Format a message with the extra arguments.
* E.g. if called as: mbFormatMessage("{1} is {0} {2}, {1}", "a good", "this", "test")
* the formatted message returned is: "this is a good test, this" * @param messageFormat the message format string * @param varArgs optional extra parameters for formatting the message * @return the formatted message */ function mbFormatMessage(messageFormat) { var message = messageFormat; var varArgs = [].slice.call(arguments, mbFormatMessage.length); for (var i in varArgs) { var parm = new RegExp("\\{" + i + "\\}", "g"); message = message.replace(parm, varArgs[i]); } return message; } Regards, -- -- Gertjan van Oosten, gertjan@West.NL, West Consulting B.V., +31 15 2191 600 From gertjan at West.NL Fri Oct 19 07:17:04 2007 From: gertjan at West.NL (Gertjan van Oosten) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Layers into Groups In-Reply-To: <4715D475.4080409@minst.net> References: <1192447207.7829.52.camel@mellon> <471355A2.9040803@ominiverdi.com> <4713F885.9060004@gmail.com> <4715D475.4080409@minst.net> Message-ID: <20071019111704.GB5597@West.NL> As quoted from "Steven M. Ottens" : > The mapbuilder code uses > //wmc:ResourceList/wmc:Layer[wmc:Extension/wmc:Group/@name=$GroupName] > So this means that it selects the layers which have a group-extension. The grouping thing is only supported in the OWSContext spec I believe (not in the WMC). The future of the OWSContext spec is unclear to me, BTW. > I believe the structure is hierarchical, so using group/layer approach > should be consistent with the OWS spec. > But I'm not the real expert, I though gertjan did the grouping code, but > could be wrong. I don't think I did that, I seem to recall Olivier (Cc'ed) did that? Regards, -- -- Gertjan van Oosten, gertjan@West.NL, West Consulting B.V., +31 15 2191 600 From crschmidt at metacarta.com Fri Oct 19 07:47:21 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <20071019110518.GA5597@West.NL> References: <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> <926726.10161.qm@web51402.mail.re2.yahoo.com> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <17AD4DD1-AF6F-44AE-9D72-3D6CFB1EB2F0@dmsolutions.ca> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <20071019110518.GA5597@West.NL> Message-ID: <20071019114721.GE1322@metacarta.com> On Fri, Oct 19, 2007 at 01:05:18PM +0200, Gertjan van Oosten wrote: > Note that even when you have i18n texts, you still need to take into > account that the developer adding new strings to the code and the people > doing the translation thereof into the other supported languages are > generally not the same. Which means you need some sort of fall-back > mechanism for message keys that remain as yet untranslated. In > MapBuilder the fall back is on the default language "en". Good point. I had forgotten about that. We should update teh RFC to take that (and other things in this thread) into account. > Anyway, to quote myself on interpolation ;-) : Right, I remembered this message. Per the OpenLayers commit policy, we prefer (unless absolutely neccesary) to get a CLA for submissions from their author, so they can be released under ClearBSD. (Understaning the legal implications of LGPL is way more than I have the legal brain cells for.) I assume that's not really a problem, but I'm mentioniong it anyway :) I'll come back to it when we get that far in the code. Regards, -- Christopher Schmidt MetaCarta From jachym.cepicky at gmail.com Fri Oct 19 09:08:35 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <20071019114721.GE1322@metacarta.com> References: <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> <926726.10161.qm@web51402.mail.re2.yahoo.com> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <17AD4DD1-AF6F-44AE-9D72-3D6CFB1EB2F0@dmsolutions.ca> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <20071019110518.GA5597@West.NL> <20071019114721.GE1322@metacarta.com> Message-ID: <1192799315.7995.19.camel@mellon> Hi Christopher Schmidt p??e v P? 19. 10. 2007 v 07:47 -0400: > On Fri, Oct 19, 2007 at 01:05:18PM +0200, Gertjan van Oosten wrote: > > Note that even when you have i18n texts, you still need to take into > > account that the developer adding new strings to the code and the people > > doing the translation thereof into the other supported languages are > > generally not the same. Which means you need some sort of fall-back > > mechanism for message keys that remain as yet untranslated. In > > MapBuilder the fall back is on the default language "en". > > Good point. I had forgotten about that. We should update teh RFC to take > that (and other things in this thread) into account. > I was thinking, that rather then return default "en" translation, it could return the untranslated message back. Since I still suppose, translation keys will be strings in english, result will be the same. Jachym > > Anyway, to quote myself on interpolation ;-) : > > Right, I remembered this message. Per the OpenLayers commit policy, we > prefer (unless absolutely neccesary) to get a CLA for submissions from > their author, so they can be released under ClearBSD. (Understaning the > legal implications of LGPL is way more than I have the legal brain cells > for.) I assume that's not really a problem, but I'm mentioniong it > anyway :) I'll come back to it when we get that far in the code. > > Regards, -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071019/dd423f2a/attachment.bin From rohith.kota at uai.com Fri Oct 19 09:56:04 2007 From: rohith.kota at uai.com (Rohith Kota) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] vector layer does not draw at lower resoluti ons. Message-ID: <76758090F8686C47A44B6FF52514A1D308681C7B@hermes.uai.int> I did try as you have suggested, but without any luck. If you move the zoom bar to the lowest level and try to draw vector it does not draw rather it would be on zoom level 2 bar higher. thanks rohith kota 327-3925 -----Original Message----- From: dev-bounces@openlayers.org [mailto:dev-bounces@openlayers.org]On Behalf Of Arnd Wippermann Sent: Thursday, October 18, 2007 3:09 PM To: dev@openlayers.org Subject: Re: [OpenLayers-Dev] vector layer does not draw at lower resolutions. Use this map.addLayers([layer,vlayer]); instead of map.addLayer(layer,vlayer); with your code only one layer will added to the map. Arnd Wippermann http://gis.ibbeck.de/ginfo -- View this message in context: http://www.nabble.com/vector-layer-does-not-draw-at-lower-resolutions.-tf462 9642.html#a13282676 Sent from the OpenLayers Dev mailing list archive at Nabble.com. _______________________________________________ Dev mailing list Dev@openlayers.org http://openlayers.org/mailman/listinfo/dev This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the sender. This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail. From crschmidt at metacarta.com Fri Oct 19 10:18:27 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <1192799315.7995.19.camel@mellon> References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <17AD4DD1-AF6F-44AE-9D72-3D6CFB1EB2F0@dmsolutions.ca> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <20071019110518.GA5597@West.NL> <20071019114721.GE1322@metacarta.com> <1192799315.7995.19.camel@mellon> Message-ID: <20071019141827.GA11728@metacarta.com> On Fri, Oct 19, 2007 at 03:08:35PM +0200, Jachym Cepicky wrote: > Hi > Christopher Schmidt p????e v P?? 19. 10. 2007 v 07:47 -0400: > > On Fri, Oct 19, 2007 at 01:05:18PM +0200, Gertjan van Oosten wrote: > > > Note that even when you have i18n texts, you still need to take into > > > account that the developer adding new strings to the code and the people > > > doing the translation thereof into the other supported languages are > > > generally not the same. Which means you need some sort of fall-back > > > mechanism for message keys that remain as yet untranslated. In > > > MapBuilder the fall back is on the default language "en". > > > > Good point. I had forgotten about that. We should update teh RFC to take > > that (and other things in this thread) into account. > > > > I was thinking, that rather then return default "en" translation, it > could return the untranslated message back. Since I still suppose, > translation keys will be strings in english, result will be the same. Right. Though perhaps support for a 'fallback' language in addition to just returning the translation string makes sense? It wouldn't be that hard, in any case -- a couple extra lines of code in String.translate() Regards, -- Christopher Schmidt MetaCarta From jachym.cepicky at gmail.com Fri Oct 19 10:48:31 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <20071019141827.GA11728@metacarta.com> References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <17AD4DD1-AF6F-44AE-9D72-3D6CFB1EB2F0@dmsolutions.ca> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <20071019110518.GA5597@West.NL> <20071019114721.GE1322@metacarta.com> <1192799315.7995.19.camel@mellon> <20071019141827.GA11728@metacarta.com> Message-ID: <1192805311.7995.23.camel@mellon> Christopher Schmidt p??e v P? 19. 10. 2007 v 10:18 -0400: > On Fri, Oct 19, 2007 at 03:08:35PM +0200, Jachym Cepicky wrote: > > > Good point. I had forgotten about that. We should update teh RFC to take > > > that (and other things in this thread) into account. > > > > > > > I was thinking, that rather then return default "en" translation, it > > could return the untranslated message back. Since I still suppose, > > translation keys will be strings in english, result will be the same. > > Right. Though perhaps support for a 'fallback' language in addition to > just returning the translation string makes sense? It wouldn't be that > hard, in any case -- a couple extra lines of code in String.translate() > > Regards, just to be sure: "support" yes, but not "assume", everything will always be translated even in "fallback" language jachym -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071019/a696021f/attachment.bin From barbara.philippot at camptocamp.com Fri Oct 19 11:18:32 2007 From: barbara.philippot at camptocamp.com (Barbara Philippot) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Developement of an external vector symbol In-Reply-To: References: Message-ID: <4718CAC8.7020705@camptocamp.com> Ian Mayo a ?crit : > Barbara, > I'm afraid I can't offer an opinion on your languages challenge. It > would be great if there was a pre-existing file-format for vector > symbols, then we just specify filenames rather than hard-code the > vectored shapes. > Yes, this is why we are trying to do. But people would have to specify two filenames (one in SVG, the other in VML) to replace the pre-existing symbol catalogue files and it's not still the ideal solution but perhaps it's the only one. > I'm currenlty working on provision of per-entry icons (images) for > georss feeds. [1]. I've made good progress, but may have to pause my > efforts for the next 7 days, see my demo at: [2]. Had you considered > using the symbols on a per-entry fashion in a georss feed? I suspect > it's fortunate I've got a self-induced development pause: it would > seem to be worth offering per-entry symbols as well as icons. > > Appreciate your thoughts/opinions. > Well, I hadn't considered using symbols in a georss feed and I don't know well Georss format. For the moment, symbols are associated with points geometry with a style.use="symbol" in the same way as images were associated with style.externalGraphic. So, I haven't looked in detail to the way you attach icons to GeoRss nodes but it seems that you would be able to use SVG and VML symbols as well as images. I hope to make progress before your development pause ends :-) > Ian. > > 1. http://trac.openlayers.org/ticket/1073 > 2. http://dev.openlayers.org/sandbox/ianmayo/branches/rss/openlayers/examples/georss.html > > From gertjan at West.NL Fri Oct 19 11:37:46 2007 From: gertjan at West.NL (Gertjan van Oosten) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <1192805311.7995.23.camel@mellon> References: <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <17AD4DD1-AF6F-44AE-9D72-3D6CFB1EB2F0@dmsolutions.ca> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <20071019110518.GA5597@West.NL> <20071019114721.GE1322@metacarta.com> <1192799315.7995.19.camel@mellon> <20071019141827.GA11728@metacarta.com> <1192805311.7995.23.camel@mellon> Message-ID: <20071019153746.GB6468@West.NL> As quoted from Jachym Cepicky : > Christopher Schmidt p???e v P? 19. 10. 2007 v 10:18 -0400: > > On Fri, Oct 19, 2007 at 03:08:35PM +0200, Jachym Cepicky wrote: > > > > Good point. I had forgotten about that. We should update teh RFC to take > > > > that (and other things in this thread) into account. > > > > > > I was thinking, that rather then return default "en" translation, it > > > could return the untranslated message back. Since I still suppose, > > > translation keys will be strings in english, result will be the same. > > > > Right. Though perhaps support for a 'fallback' language in addition to > > just returning the translation string makes sense? It wouldn't be that > > hard, in any case -- a couple extra lines of code in String.translate() > > just to be sure: "support" yes, but not "assume", everything will always > be translated even in "fallback" language In MapBuilder we use message keys to look up the message; think of mnemonical names like featureNotFound or pointProjectsToInfinity. If a messageKey is not found for the selected language, and the messageKey is not found for the default language, then the messageKey itself is returned as the message. That way you a. get at least some message for every key and b. can easily inspect visually that a messageKey has not yet been translated into even the default language. Regards, -- -- Gertjan van Oosten, gertjan@West.NL, West Consulting B.V., +31 15 2191 600 From gertjan at West.NL Fri Oct 19 11:41:17 2007 From: gertjan at West.NL (Gertjan van Oosten) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <20071019153746.GB6468@West.NL> References: <20071018133356.GA19498@metacarta.com> <17AD4DD1-AF6F-44AE-9D72-3D6CFB1EB2F0@dmsolutions.ca> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <20071019110518.GA5597@West.NL> <20071019114721.GE1322@metacarta.com> <1192799315.7995.19.camel@mellon> <20071019141827.GA11728@metacarta.com> <1192805311.7995.23.camel@mellon> <20071019153746.GB6468@West.NL> Message-ID: <20071019154117.GA8385@West.NL> > If a messageKey is not found for the selected language, and the > messageKey is not found for the default language, then the messageKey > itself is returned as the message. To be complete: if no messages at all are found (yes, that can happen with MapBuilder if e.g. a wrong widgetText directory is specified in the config) it returns "NoMsgsFound". Not sure if that could happen with OpenLayers though. Regards, -- -- Gertjan van Oosten, gertjan@West.NL, West Consulting B.V., +31 15 2191 600 From jachym.cepicky at gmail.com Fri Oct 19 11:44:35 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <20071019153746.GB6468@West.NL> References: <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <17AD4DD1-AF6F-44AE-9D72-3D6CFB1EB2F0@dmsolutions.ca> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <20071019110518.GA5597@West.NL> <20071019114721.GE1322@metacarta.com> <1192799315.7995.19.camel@mellon> <20071019141827.GA11728@metacarta.com> <1192805311.7995.23.camel@mellon> <20071019153746.GB6468@West.NL> Message-ID: <1192808675.7995.35.camel@mellon> Gertjan van Oosten p??e v P? 19. 10. 2007 v 17:37 +0200: > In MapBuilder we use message keys to look up the message; think of > mnemonical names like featureNotFound or pointProjectsToInfinity. > If a messageKey is not found for the selected language, and the > messageKey is not found for the default language, then the messageKey > itself is returned as the message. That way you a. get at least some > message for every key and b. can easily inspect visually that a > messageKey has not yet been translated into even the default language. > yes. only difference is, I'm suggesting not to use specified messageKeys, but strings itself, because it can be IMHO fast mess. Otherwise, the concept is exactly that, what I would expect. jachym -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071019/f30d7bc2/attachment.bin From tschaub at openplans.org Fri Oct 19 12:38:46 2007 From: tschaub at openplans.org (Tim Schaub) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jslint cleanup committed In-Reply-To: <20071018130448.GB14361@metacarta.com> References: <20071018130448.GB14361@metacarta.com> Message-ID: <4718DD96.8000102@openplans.org> Hey- Christopher Schmidt wrote: > On Thu, Oct 18, 2007 at 09:01:57AM -0400, Paul Spencer wrote: >> I have added missing semi-colons and curly braces around all block >> scopes in: >> >> http://trac.openlayers.org/changeset/5002/ >> >> jslint reports the remaining warnings: >> >> /lib/Firebug/firebug.js: 0 error(s), 45 warning(s) >> /lib/Firebug/firebugx.js: 0 error(s), 2 warning(s) >> /lib/Rico/Color.js: 0 error(s), 18 warning(s) >> /lib/Rico/Corner.js: 0 error(s), 36 warning(s) >> >> Do you want me to do anything about these? > > I would prefer we leave the firebug files alone -- I don't think we've > touched those at all. > We ship a patched version of firebug lite. See http://code.google.com/p/fbug/issues/detail?id=85 I'd rather not patch it any further, as it looks like even that patch might not get any attention. More importantly, we don't concatenate/compress firebug in the single file builds. If anybody does this, they should be scolded. So, no need to modify firebug (any more than I have) as far as I'm concerned. Thanks for fixing the other issues Paul. Tim > I have no strong feelings on Rico: we've already changed those files to > put them in the OpenLayers namespace, further changes don't bother me > much, but I'm also happy to just leave them. > > Regards, From val.cartei at gmail.com Fri Oct 19 12:38:50 2007 From: val.cartei at gmail.com (Val Cartei) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] use openlayers for images Message-ID: Instead of using maps, I'd like to extend the OpenLayers functions to be able to manipulate other images too. This is what I have got so far: http://dev.getconveyor.org/panorama_experiments/15 The main issue I am having at the moment is that I can't make the library not to use resolution info. The image should be displayed to its original resolution (as if displayed in a browser) and the overviewmap built accordingly. Has anyone suggestions? -- Best Regards, Valentina Cartei UK Tel. +44 (0)796 6882820 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071019/ab56fc07/attachment.html From pspencer at dmsolutions.ca Fri Oct 19 12:42:38 2007 From: pspencer at dmsolutions.ca (Paul Spencer) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jslint cleanup committed In-Reply-To: <4718DD96.8000102@openplans.org> References: <20071018130448.GB14361@metacarta.com> <4718DD96.8000102@openplans.org> Message-ID: On 19-Oct-07, at 12:38 PM, Tim Schaub wrote: >>> /lib/Rico/Color.js: 0 error(s), 18 warning(s) >>> /lib/Rico/Corner.js: 0 error(s), 36 warning(s) >>> >>> Do you want me to do anything about these? >> >> I would prefer we leave the firebug files alone -- I don't think >> we've >> touched those at all. >> > > We ship a patched version of firebug lite. See > http://code.google.com/p/fbug/issues/detail?id=85 > > I'd rather not patch it any further, as it looks like even that patch > might not get any attention. > > More importantly, we don't concatenate/compress firebug in the single > file builds. If anybody does this, they should be scolded. > for the record, I did fix the above warnings in Rico/* and did NOT touch firebug. As Tim indicates, it doesn't affect the final build. As of now, the final full build is 'lint free' with my current settings, which makes me very happy :) I may look at some of the other settings to see if there is anything else we could potentially benefit from in the newer version of jslint which I discovered last night ... Cheers Paul +-----------------------------------------------------------------+ |Paul Spencer pspencer@dmsolutions.ca | +-----------------------------------------------------------------+ |Chief Technology Officer | |DM Solutions Group Inc http://www.dmsolutions.ca/ | +-----------------------------------------------------------------+ From pspencer at dmsolutions.ca Fri Oct 19 12:46:35 2007 From: pspencer at dmsolutions.ca (Paul Spencer) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] use openlayers for images In-Reply-To: References: Message-ID: <118744D3-6118-4539-8D67-2299EDEBC7A3@dmsolutions.ca> I did something similar to show pre-built tiles in pixel space, how about something like: olMap = new OpenLayers.Map( 'map', {maxExtent: new OpenLayers.Bounds(0, 0, 4000, 4000), minResolution: 1.0, maxResolution: 1.0, controls: []}); Paul On 19-Oct-07, at 12:38 PM, Val Cartei wrote: > Instead of using maps, I'd like to extend the OpenLayers functions > to be able to manipulate other images too. > This is what I have got so far: > http://dev.getconveyor.org/panorama_experiments/15 > The main issue I am having at the moment is that I can't make the > library not to use resolution info. The image should be displayed > to its original resolution (as if displayed in a browser) and the > overviewmap built accordingly. > Has anyone suggestions? > > -- > Best Regards, > > Valentina Cartei > UK Tel. +44 (0)796 6882820 > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev +-----------------------------------------------------------------+ |Paul Spencer pspencer@dmsolutions.ca | +-----------------------------------------------------------------+ |Chief Technology Officer | |DM Solutions Group Inc http://www.dmsolutions.ca/ | +-----------------------------------------------------------------+ From olivier.terral at geomatys.fr Fri Oct 19 12:47:27 2007 From: olivier.terral at geomatys.fr (olivier.terral) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Layers into Groups In-Reply-To: <20071019111704.GB5597@West.NL> References: <1192447207.7829.52.camel@mellon> <471355A2.9040803@ominiverdi.com> <4713F885.9060004@gmail.com> <4715D475.4080409@minst.net> <20071019111704.GB5597@West.NL> Message-ID: <4718DF9F.1090002@geomatys.fr> Ola, Sorry , I think it's Vincent Schut who did that but I 've often manipulated Group layers . I don't know what is it specified in OWS spec but I know what is done in Mapbuilder. We have choosen to add in an Extension tag because the spec was not clear with groups and the Extension tag permits to add everything we want (that's I've understood). All groups are specified inside the General/Extension tags : and if a Layer is a member of a Group, it's specified in its extension tag like said Steven But actually, this way of define groups is very limited because there is only one level possible. May be we should improve this solution to permit a tree view as I've seen in the demo of Arm?n http://rips-uis.lubw.baden-wuerttemberg.de/rips/wms_viewer/test/myOpenLayer.php Gertjan van Oosten a ?crit : > As quoted from "Steven M. Ottens" : > >> The mapbuilder code uses >> //wmc:ResourceList/wmc:Layer[wmc:Extension/wmc:Group/@name=$GroupName] >> So this means that it selects the layers which have a group-extension. >> > > The grouping thing is only supported in the OWSContext spec I believe > (not in the WMC). The future of the OWSContext spec is unclear to me, BTW. > > >> I believe the structure is hierarchical, so using group/layer approach >> should be consistent with the OWS spec. >> But I'm not the real expert, I though gertjan did the grouping code, but >> could be wrong. >> > > I don't think I did that, I seem to recall Olivier (Cc'ed) did that? > > Regards, > From tschaub at openplans.org Fri Oct 19 12:53:52 2007 From: tschaub at openplans.org (Tim Schaub) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> Message-ID: <4718E120.6000401@openplans.org> Or we could do like others and give elements that contain strings that need translating a unique attribute. Simplified, this could look something like this: foo in my lang Then, switching languages would comb through all elements on the page, replacing the content of those with i18n attributes - determining the appropriate content based on the global lang setting and the value of the key in the appropriate dictionary. Not efficient, but simple. Tim PS - I don't think we should go too heavy here because I think OL as a mapping library should contain very few strings if any. How can we get strings out of the library you ask? Use templates to make things like the layer switcher - instead of cramming a bunch of text node creation code in the controls. Paul Spencer wrote: > On 18-Oct-07, at 2:31 PM, Christopher Schmidt wrote: > >> setLanguage, if we're going to provide a method for it, is going to >> need >> to be more complex than that. We'd have to loop through the controls, >> and call redraw() on them, to update their text, and I'm not sure that >> all controls would support a redraw correctly. >> > > Could we use an event that is triggered and then controls that have > strings can register (or be registered) for the language changing? > > OpenLayers.String.events.register('languagechange', > OpenLayers.Function.bind(this, this.onLanguageChange)); > > Then it is a matter of implementing support in specific controls that > need it or can use it. And if there are things other than controls > that can use the current language, they can take advantage easily. > For instance, some layer types might be able to use the current > language for labeling. And applications (or application frameworks) > around OpenLayers could also take advantage of it. > > We will probably want a LanguageSwitcher control that sets the > language and tracks the current language too, mostly for testing and > development. > > We need a way of indicating what languages to support in any given > application, and of enumerating the supported languages through the > OpenLayers API. > > We need a way to extend the base set of translations with other, > application-specific translations. > > Cheers > > Paul > > +-----------------------------------------------------------------+ > |Paul Spencer pspencer@dmsolutions.ca | > +-----------------------------------------------------------------+ > |Chief Technology Officer | > |DM Solutions Group Inc http://www.dmsolutions.ca/ | > +-----------------------------------------------------------------+ > > > > > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > !DSPAM:4033,4717b81f246325210051143! > From tschaub at openplans.org Fri Oct 19 13:18:44 2007 From: tschaub at openplans.org (Tim Schaub) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <6ae3fb590710181429n182c547bld43f7be10bbcaedc@mail.gmail.com> References: <926726.10161.qm@web51402.mail.re2.yahoo.com> <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> <20071018200113.GA7249@metacarta.com> <6ae3fb590710181429n182c547bld43f7be10bbcaedc@mail.gmail.com> Message-ID: <4718E6F4.6010103@openplans.org> I think we should organize thoughts with these use cases in mind. Please correct me as you see fit - but this is how I see things breaking down (descending order based on number of people in each group). 1) people who want a single language (theirs) 2) people who want dynamic translation I almost put 0) people who don't know about language since that is the case that OpenLayers currently considers - but I didn't want to make anybody angry. Anyway, I'd like to see things done so we don't sacrifice efficiency for people who want a single file build that just includes their single language. Having every element that contains a string registered as an event listener starts sounding a bit heavy for this case. And, in terms of parsing strings for formatting, isn't it enough to break elements up based on the portions that can be translated:

OpenLayers rocks

? This doesn't work for alert(str), but I happen to think we should purge alerts from the library. Being contrary, Tim PS- Also, I know the i18n attribute is too simple to deal with element attribute values that need translating themselves ( translate me

). In that case, we need multiple attributes in the i18n namespace (something akin to http://wiki.zope.org/zope3/ZPTInternationalizationSupport). Erik Uzureau wrote: > On 10/18/07, Christopher Schmidt wrote: >> On Thu, Oct 18, 2007 at 03:46:21PM -0400, Paul Spencer wrote: >>> On 18-Oct-07, at 2:31 PM, Christopher Schmidt wrote: >>> >>>> setLanguage, if we're going to provide a method for it, is going to >>>> need >>>> to be more complex than that. We'd have to loop through the controls, >>>> and call redraw() on them, to update their text, and I'm not sure that >>>> all controls would support a redraw correctly. >>>> >>> Could we use an event that is triggered and then controls that have >>> strings can register (or be registered) for the language changing? >>> >>> OpenLayers.String.events.register('languagechange', >>> OpenLayers.Function.bind(this, this.onLanguageChange)); >> I like this idea. I hadn't thought of it, and it makes sense. > > yes, i like it to. the only thing i dont like about it is that it > seems like it might sort of be a burden to control developers to have > to deal with registering events. > > to that end, we might consider such a thing as making the base > OpenLayers.Control's constructor register the listener and call a stub > function udpateLanguage() or something like that. In > OpenLayers.Control, the stub function would do nothing. Then if > controls _do_ wish to update on language changes, they need only > implement the updateLanguage(). > > This might help in the case of a control that is subclassed from > another control -- the updateLanguage() calls would be chained up, > instead of having to register a separate listener for each class. > > I dunno. This is just the first thing that comes to mind... that as a > control developer it seems like kind of a PITA and or unnecessary > overhead to be registering more listeners. > > --E > > >>> We will probably want a LanguageSwitcher control that sets the >>> language and tracks the current language too, mostly for testing and >>> development. >> I'm not convinced that we need that, but if it gets built as part of >> development, it would probably be useful. >> >>> We need a way of indicating what languages to support in any given >>> application, and of enumerating the supported languages through the >>> OpenLayers API. >> for (var key in OpenLayers.Strings.languages) { >> alert(key); >> } >> >>> We need a way to extend the base set of translations with other, >>> application-specific translations. >> OpenLayers.Strings.languages['eng']['My Application String!'] = 'My >> application String!'; >> >> OpenLayers.Strings.languages['spa']['My Application String!'] = >> '?String Applicacion de mio!'; >> >> Regards, >> -- >> Christopher Schmidt >> MetaCarta >> > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > !DSPAM:4033,4717d058293873327367457! > From tschaub at openplans.org Fri Oct 19 13:30:26 2007 From: tschaub at openplans.org (Tim Schaub) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jslint cleanup committed In-Reply-To: References: <20071018130448.GB14361@metacarta.com> <4718DD96.8000102@openplans.org> Message-ID: <4718E9B2.9010308@openplans.org> Hey- Paul Spencer wrote: > > for the record, I did fix the above warnings in Rico/* and did NOT touch > firebug. As Tim indicates, it doesn't affect the final build. As of > now, the final full build is 'lint free' with my current settings, which > makes me very happy :) I may look at some of the other settings to see > if there is anything else we could potentially benefit from in the newer > version of jslint which I discovered last night ... > Paul, I'd also be interested in the result of disallowing use of undefined variables. Perhaps you already talked about this and I missed it. What would be really cool is if jslint could be tweaked to look for variable assignment outside of the OpenLayers namespace - something we occasionally stumble across and have no way of testing for. (I know a case where window.i == 2 and I cannot figure out why.) This might even be caught by the "no undefined variables" filter if it is initialized without the var keyword (which is likely). Tim > Cheers > > Paul > > +-----------------------------------------------------------------+ > |Paul Spencer pspencer@dmsolutions.ca | > +-----------------------------------------------------------------+ > |Chief Technology Officer | > |DM Solutions Group Inc http://www.dmsolutions.ca/ | > +-----------------------------------------------------------------+ > > > > > > > !DSPAM:4033,4718de8133771961014482! > From pspencer at dmsolutions.ca Fri Oct 19 15:33:40 2007 From: pspencer at dmsolutions.ca (Paul Spencer) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jslint cleanup committed In-Reply-To: <4718E9B2.9010308@openplans.org> References: <20071018130448.GB14361@metacarta.com> <4718DD96.8000102@openplans.org> <4718E9B2.9010308@openplans.org> Message-ID: <3F98421E-0C63-4763-B2C2-3F727AE79D11@dmsolutions.ca> For those of you that are not on IRC, Tim and I worked this out this afternoon and I committed: http://trac.openlayers.org/changeset/5028 which cleans up some 29 instances of local scope variables used without 'var', making them global in scope. This is now part of my regularly scheduled routine :) Cheers Paul On 19-Oct-07, at 1:30 PM, Tim Schaub wrote: > Hey- > > Paul Spencer wrote: >> >> for the record, I did fix the above warnings in Rico/* and did NOT >> touch >> firebug. As Tim indicates, it doesn't affect the final build. As of >> now, the final full build is 'lint free' with my current settings, >> which >> makes me very happy :) I may look at some of the other settings >> to see >> if there is anything else we could potentially benefit from in the >> newer >> version of jslint which I discovered last night ... >> > > Paul, I'd also be interested in the result of disallowing use of > undefined variables. Perhaps you already talked about this and I > missed it. > > What would be really cool is if jslint could be tweaked to look for > variable assignment outside of the OpenLayers namespace - something we > occasionally stumble across and have no way of testing for. (I know a > case where window.i == 2 and I cannot figure out why.) This might > even > be caught by the "no undefined variables" filter if it is initialized > without the var keyword (which is likely). > > Tim > >> Cheers >> >> Paul >> >> +-----------------------------------------------------------------+ >> |Paul Spencer pspencer@dmsolutions.ca | >> +-----------------------------------------------------------------+ >> |Chief Technology Officer | >> |DM Solutions Group Inc http://www.dmsolutions.ca/ | >> +-----------------------------------------------------------------+ >> >> >> >> >> >> >> !DSPAM:4033,4718de8133771961014482! >> > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev +-----------------------------------------------------------------+ |Paul Spencer pspencer@dmsolutions.ca | +-----------------------------------------------------------------+ |Chief Technology Officer | |DM Solutions Group Inc http://www.dmsolutions.ca/ | +-----------------------------------------------------------------+ From pspencer at dmsolutions.ca Fri Oct 19 15:35:15 2007 From: pspencer at dmsolutions.ca (Paul Spencer) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jslint cleanup committed In-Reply-To: <3F98421E-0C63-4763-B2C2-3F727AE79D11@dmsolutions.ca> References: <20071018130448.GB14361@metacarta.com> <4718DD96.8000102@openplans.org> <4718E9B2.9010308@openplans.org> <3F98421E-0C63-4763-B2C2-3F727AE79D11@dmsolutions.ca> Message-ID: <62EC14F9-9730-4628-A62A-9AF684F8E6F7@dmsolutions.ca> 19 instances, not 29. Next, I will get a calculator and learn to add ... On 19-Oct-07, at 3:33 PM, Paul Spencer wrote: > For those of you that are not on IRC, Tim and I worked this out > this afternoon and I committed: > > http://trac.openlayers.org/changeset/5028 > > which cleans up some 29 instances of local scope variables used > without 'var', making them global in scope. This is now part of my > regularly scheduled routine :) > > Cheers > > Paul > > On 19-Oct-07, at 1:30 PM, Tim Schaub wrote: > >> Hey- >> >> Paul Spencer wrote: >>> >>> for the record, I did fix the above warnings in Rico/* and did >>> NOT touch >>> firebug. As Tim indicates, it doesn't affect the final build. >>> As of >>> now, the final full build is 'lint free' with my current >>> settings, which >>> makes me very happy :) I may look at some of the other settings >>> to see >>> if there is anything else we could potentially benefit from in >>> the newer >>> version of jslint which I discovered last night ... >>> >> >> Paul, I'd also be interested in the result of disallowing use of >> undefined variables. Perhaps you already talked about this and I >> missed it. >> >> What would be really cool is if jslint could be tweaked to look for >> variable assignment outside of the OpenLayers namespace - >> something we >> occasionally stumble across and have no way of testing for. (I >> know a >> case where window.i == 2 and I cannot figure out why.) This might >> even >> be caught by the "no undefined variables" filter if it is initialized >> without the var keyword (which is likely). >> >> Tim >> >>> Cheers >>> >>> Paul >>> >>> +-----------------------------------------------------------------+ >>> |Paul Spencer pspencer@dmsolutions.ca | >>> +-----------------------------------------------------------------+ >>> |Chief Technology Officer | >>> |DM Solutions Group Inc http://www.dmsolutions.ca/ | >>> +-----------------------------------------------------------------+ >>> >>> >>> >>> >>> >>> >>> !DSPAM:4033,4718de8133771961014482! >>> >> >> _______________________________________________ >> Dev mailing list >> Dev@openlayers.org >> http://openlayers.org/mailman/listinfo/dev > > +-----------------------------------------------------------------+ > |Paul Spencer pspencer@dmsolutions.ca | > +-----------------------------------------------------------------+ > |Chief Technology Officer | > |DM Solutions Group Inc http://www.dmsolutions.ca/ | > +-----------------------------------------------------------------+ > > > > > +-----------------------------------------------------------------+ |Paul Spencer pspencer@dmsolutions.ca | +-----------------------------------------------------------------+ |Chief Technology Officer | |DM Solutions Group Inc http://www.dmsolutions.ca/ | +-----------------------------------------------------------------+ From lorenzo at ominiverdi.com Fri Oct 19 16:09:05 2007 From: lorenzo at ominiverdi.com (Lorenzo Becchi) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Layers into Groups In-Reply-To: <4718DF9F.1090002@geomatys.fr> References: <1192447207.7829.52.camel@mellon> <471355A2.9040803@ominiverdi.com> <4713F885.9060004@gmail.com> <4715D475.4080409@minst.net> <20071019111704.GB5597@West.NL> <4718DF9F.1090002@geomatys.fr> Message-ID: <47190EE1.3030206@ominiverdi.com> Hi Olivier, > May be we should improve this solution to permit a tree view as I've > seen in the demo of Arm?n > I think too we shoud choose something all together and try to use osgeo-standards mailing list to make lobby and help OGC to go in a good direction. As we are dealing with OGC standards, why don't we use the solution already used by WMS? well, it's a bit I'm not revisioning the specs but should be something like this: my group of level 1 my group of layer 2 layer_1 .... layer_2 .... layer3 ciao Lorenzo olivier.terral wrote: > Ola, > > Sorry , I think it's Vincent Schut who did that but I 've often > manipulated Group layers . > > I don't know what is it specified in OWS spec but I know what is done > in Mapbuilder. > We have choosen to add in an Extension tag because the spec was > not clear with groups and the Extension tag permits to add everything we > want (that's I've understood). > > All groups are specified inside the General/Extension tags : > > > > > > > > > > > > and if a Layer is a member of a Group, it's specified in its extension > tag like said Steven > > > > > > But actually, this way of define groups is very limited because there is > only one level possible. > > May be we should improve this solution to permit a tree view as I've > seen in the demo of Arm?n > > http://rips-uis.lubw.baden-wuerttemberg.de/rips/wms_viewer/test/myOpenLayer.php > > > Gertjan van Oosten a ?crit : > >> As quoted from "Steven M. Ottens" : >> >> >>> The mapbuilder code uses >>> //wmc:ResourceList/wmc:Layer[wmc:Extension/wmc:Group/@name=$GroupName] >>> So this means that it selects the layers which have a group-extension. >>> >>> >> The grouping thing is only supported in the OWSContext spec I believe >> (not in the WMC). The future of the OWSContext spec is unclear to me, BTW. >> >> >> >>> I believe the structure is hierarchical, so using group/layer approach >>> should be consistent with the OWS spec. >>> But I'm not the real expert, I though gertjan did the grouping code, but >>> could be wrong. >>> >>> >> I don't think I did that, I seem to recall Olivier (Cc'ed) did that? >> >> Regards, >> >> > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > From cameron.shorter at gmail.com Fri Oct 19 17:34:41 2007 From: cameron.shorter at gmail.com (Cameron Shorter) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <20071019110518.GA5597@West.NL> References: <6ae3fb590710181022n5151ef06wadc5d87b5f2b32d@mail.gmail.com> <6ae3fb590710181023j1bf6cb0cmcfe58b201207bdc2@mail.gmail.com> <20071018183104.GB31939@metacarta.com> <926726.10161.qm@web51402.mail.re2.yahoo.com> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <17AD4DD1-AF6F-44AE-9D72-3D6CFB1EB2F0@dmsolutions.ca> <1192713830.8316.9.camel@mellon> <20071018133356.GA19498@metacarta.com> <20071019110518.GA5597@West.NL> Message-ID: <471922F1.1050509@gmail.com> I think this is another candidate for functionality from Mapbuilder to be migrated to Openlayers or maybe the new Openlayers Addins directory. Mapbuilder can then access functionality by calling Openlayers. It would probably involve a little reformatting of code to fit the Openlayers style guide. Gertjan would be able to talk to this further. Gertjan van Oosten wrote: > As quoted from Christopher Schmidt : > >> It needs to be more complex than this to support interpolation, and I >> think it will need to: as the RFC says: >> > [ http://trac.openlayers.org/wiki/RFC/Internationalization ] > >> "The OpenLayers.String object will have an additional function, >> translate. Translate will accept a translation key -- a single string >> which can be used to look up the string -- and any number of additional >> arguments. If additional arguments are passed, they will be interpolated >> in the string in order." >> >> MapBuilder already has some code that does this, so we can work with >> them to integrate those bits of it into OpenLayers. >> > > As I wrote on the I18N thread mid-July (see below), OpenLayers can just > copy the interpolation function I wrote for MapBuilder. So > interpolation capabilities are more or less a given. The only question > remaining is how to get the i18n'ed strings into (and out of) the > Openlayers code. I have never come across a real-life use case where > you would want i18n data to be displayed in mixed-language in one > application, so a single, global language setting would suffice for the > majority of cases. > > Note that even when you have i18n texts, you still need to take into > account that the developer adding new strings to the code and the people > doing the translation thereof into the other supported languages are > generally not the same. Which means you need some sort of fall-back > mechanism for message keys that remain as yet untranslated. In > MapBuilder the fall back is on the default language "en". > > To start off, I guess simple language tags ("en", "nl", "fr", ...) will > suffice. If and when people want to support language subtags ("en-GB", > "en-US", "nl-NL", "nl-BE", "fr-FR", "fr-CA", ...) you should also > incorporate a fall back from the subtagged language on the more general > language tag. For MapBuilder this is still on my TODO list; there has > been no demand for it yet. > > [See http://communitymapbuilder.org/display/MAP/Add+Translations > for complete information on how MapBuilder does things.] > > Anyway, to quote myself on interpolation ;-) : > > > [...] the function mbFormatMessage() I wrote, copied below from > trunk/lib/util/Util.js for your convenience, can be used straight > out-of-the-box to get simple interpolation. Since at the top of the > file it clearly states that it is LGPL, nothing is stopping you. 8-) > > /** > * Format a message with the extra arguments.
> * E.g. if called as: mbFormatMessage("{1} is {0} {2}, {1}", "a good", "this", "test")
> * the formatted message returned is: "this is a good test, this" > * @param messageFormat the message format string > * @param varArgs optional extra parameters for formatting the message > * @return the formatted message > */ > function mbFormatMessage(messageFormat) > { > var message = messageFormat; > var varArgs = [].slice.call(arguments, mbFormatMessage.length); > for (var i in varArgs) { > var parm = new RegExp("\\{" + i + "\\}", "g"); > message = message.replace(parm, varArgs[i]); > } > return message; > } > > > Regards, > -- Cameron Shorter Systems Architect, http://lisasoft.com.au Tel: +61 (0)2 8570 5050 Mob: +61 (0)419 142 254 From erik.uzureau at metacarta.com Fri Oct 19 18:27:02 2007 From: erik.uzureau at metacarta.com (Erik Uzureau) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jslint cleanup committed In-Reply-To: <3F98421E-0C63-4763-B2C2-3F727AE79D11@dmsolutions.ca> References: <20071018130448.GB14361@metacarta.com> <4718DD96.8000102@openplans.org> <4718E9B2.9010308@openplans.org> <3F98421E-0C63-4763-B2C2-3F727AE79D11@dmsolutions.ca> Message-ID: <6ae3fb590710191527h3acb8fdcn29493fdf1856418c@mail.gmail.com> Another wicked patch. Also tim's complementary patch http://trac.openlayers.org/changeset/5027 ...which clears the jslint complaints for the 'OpenLayers' variable. Great job with this. Let's hope we can keep the OL codebase free of these slips in the future. :-) E On 10/19/07, Paul Spencer wrote: > For those of you that are not on IRC, Tim and I worked this out this > afternoon and I committed: > > http://trac.openlayers.org/changeset/5028 > > which cleans up some 29 instances of local scope variables used > without 'var', making them global in scope. This is now part of my > regularly scheduled routine :) > > Cheers > > Paul > > On 19-Oct-07, at 1:30 PM, Tim Schaub wrote: > > > Hey- > > > > Paul Spencer wrote: > >> > >> for the record, I did fix the above warnings in Rico/* and did NOT > >> touch > >> firebug. As Tim indicates, it doesn't affect the final build. As of > >> now, the final full build is 'lint free' with my current settings, > >> which > >> makes me very happy :) I may look at some of the other settings > >> to see > >> if there is anything else we could potentially benefit from in the > >> newer > >> version of jslint which I discovered last night ... > >> > > > > Paul, I'd also be interested in the result of disallowing use of > > undefined variables. Perhaps you already talked about this and I > > missed it. > > > > What would be really cool is if jslint could be tweaked to look for > > variable assignment outside of the OpenLayers namespace - something we > > occasionally stumble across and have no way of testing for. (I know a > > case where window.i == 2 and I cannot figure out why.) This might > > even > > be caught by the "no undefined variables" filter if it is initialized > > without the var keyword (which is likely). > > > > Tim > > > >> Cheers > >> > >> Paul > >> > >> +-----------------------------------------------------------------+ > >> |Paul Spencer pspencer@dmsolutions.ca | > >> +-----------------------------------------------------------------+ > >> |Chief Technology Officer | > >> |DM Solutions Group Inc http://www.dmsolutions.ca/ | > >> +-----------------------------------------------------------------+ > >> > >> > >> > >> > >> > >> > >> !DSPAM:4033,4718de8133771961014482! > >> > > > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev > > +-----------------------------------------------------------------+ > |Paul Spencer pspencer@dmsolutions.ca | > +-----------------------------------------------------------------+ > |Chief Technology Officer | > |DM Solutions Group Inc http://www.dmsolutions.ca/ | > +-----------------------------------------------------------------+ > > > > > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > From veen at science.uva.nl Sat Oct 20 10:00:15 2007 From: veen at science.uva.nl (Lourens Veen) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <1192713830.8316.9.camel@mellon> References: <1192713830.8316.9.camel@mellon> Message-ID: <200710201600.15839.veen@science.uva.nl> On Thursday 18 October 2007 15:23:50 Jachym Cepicky wrote: > Hi, > > I would like to ask about the status of OpenLayers language support. > There are not many strings, which could be translated, however, some > are there and IMHO we should (de)find a way, how to make this > possible. Was there some discussion to this topic? I haven't followed this thread that carefully so maybe it has come up already, but what about number formats? OL's current scale display uses K and M for thousands and millions, which is not how it's commonly done on paper maps. I'd really rather have it write out the full number, but of course it would then have to use a number format that is appropriate to the locale. In Dutch, I would like it to say 1:50.000, while in English that would be 1:50,000. If you're designing i18n anyway... Lourens -- ? ? ? ? ?ir. Lourens Veen ? ? ? ? University of Amsterdam - IBED ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? Nieuwe Achtergracht 166 ? ?Scientific Software Engineer ? 1018 WV Amsterdam, The Netherlands ? ? ?? ? ? ? EcoGRID-GAN ? ? ? ? ? t: +31 20-5257453 ?f: +31 20-5257431 From gertjan at West.NL Sat Oct 20 13:02:15 2007 From: gertjan at West.NL (Gertjan van Oosten) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <200710201600.15839.veen@science.uva.nl> References: <1192713830.8316.9.camel@mellon> <200710201600.15839.veen@science.uva.nl> Message-ID: <52090.82.93.165.204.1192899735.squirrel@www.west.nl> As quoted from Lourens Veen: > I haven't followed this thread that carefully so maybe it has come up > already, but what about number formats? OL's current scale display uses > K and M for thousands and millions, which is not how it's commonly done > on paper maps. I'd really rather have it write out the full number, but > of course it would then have to use a number format that is appropriate > to the locale. In Dutch, I would like it to say 1:50.000, while in > English that would be 1:50,000. > > If you're designing i18n anyway... Well, strictly speaking you're talking about L10n (Localisation). That's currently beyond MapBuilder's reach, as you would need locale-specific formatters for that (date, numbers, money). It is doable; some JavaScript classes to do this may be available out there. I haven't taken the effort to integrate them into MapBuilder though (not enough demand). Once you have these L10n classes, fitting them into the i18n framework is easy, as I will show below. Let's say we start out with: var msg = "Scale 1:" + map.scale; which (if map.scale==25000) would give us "Scale 1:25000", in any language. Say the message key "mapScale" maps to the message format string "Scale 1:{0}" in English or "Schaal 1:{0}" in Dutch. After the I18n step it would be: var msg = OpenLayers.String.format("mapScale", msp.scale); which would give "Scale 1:25000" in English and "Schaal 1:25000" in Dutch. If we then add L10n, e.g.: var formattedNumber = OpenLayers.Something.DecimalFormat(map.scale); would return the number formatted according to the proper decimal format for the current locale, e.g. #,### for UK or #.### for NL. Combining the two: var msg = OpenLayers.String.format("mapScale", formattedNumber); would return the I18n'ed message, and since the formattedNumber is L10n'ed, the resulting message would be say "Scale 1:25,000" in English and "Schaal 1:25.000" in Dutch. As you see, I18n and L10n are quite complementary. So the two could be implemented separately (and could even have separate RFC's, which has my preference). Regards, -- -- Gertjan van Oosten, gertjan@West.NL, West Consulting B.V., +31 15 2191 600 From veen at science.uva.nl Sat Oct 20 15:28:52 2007 From: veen at science.uva.nl (Lourens Veen) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <52090.82.93.165.204.1192899735.squirrel@www.west.nl> References: <1192713830.8316.9.camel@mellon> <200710201600.15839.veen@science.uva.nl> <52090.82.93.165.204.1192899735.squirrel@www.west.nl> Message-ID: <200710202128.52985.veen@science.uva.nl> On Saturday 20 October 2007 19:02:15 Gertjan van Oosten wrote: > As quoted from Lourens Veen: > > I haven't followed this thread that carefully so maybe it has come > > up already, but what about number formats? OL's current scale > > display uses K and M for thousands and millions, which is not how > > it's commonly done on paper maps. I'd really rather have it write > > out the full number, but of course it would then have to use a > > number format that is appropriate to the locale. In Dutch, I would > > like it to say 1:50.000, while in English that would be 1:50,000. > > > > If you're designing i18n anyway... > > Well, strictly speaking you're talking about L10n (Localisation). That's not quite how I thought the terms were used, and Wikipedia seems to agree with me. Adding the capability for numbers to be formatted in different ways is i18n, adding specific number formats for certain locales would be L10n. Not that it affects the technical argument :-). > That's currently beyond MapBuilder's reach, as you would need > locale-specific formatters for that (date, numbers, money). It is > doable; some JavaScript classes to do this may be available out > there. I haven't taken the effort to integrate them into MapBuilder > though (not enough demand). Once you have these L10n classes, fitting > them into the i18n framework is easy, as I will show below. > > Let's say we start out with: > var msg = "Scale 1:" + map.scale; > which (if map.scale==25000) would give us "Scale 1:25000", in any > language. > > Say the message key "mapScale" maps to the message format string > "Scale 1:{0}" in English or "Schaal 1:{0}" in Dutch. After the I18n > step it would be: var msg = OpenLayers.String.format("mapScale", > msp.scale); which would give "Scale 1:25000" in English and "Schaal > 1:25000" in Dutch. > > If we then add L10n, e.g.: > var formattedNumber = > OpenLayers.Something.DecimalFormat(map.scale); would return the > number formatted according to the proper decimal format for the > current locale, e.g. #,### for UK or #.### for NL. > > Combining the two: > var msg = OpenLayers.String.format("mapScale", formattedNumber); > would return the I18n'ed message, and since the formattedNumber is > L10n'ed, the resulting message would be say "Scale 1:25,000" in > English and "Schaal 1:25.000" in Dutch. > > As you see, I18n and L10n are quite complementary. So the two could > be implemented separately (and could even have separate RFC's, which > has my preference). Sounds good to me. Unfortunately I have no time to work on it myself at the moment, but I thought I'd mention it in case your tuits were round enough. Lourens -- ? ? ? ? ?ir. Lourens Veen ? ? ? ? University of Amsterdam - IBED ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? Nieuwe Achtergracht 166 ? ?Scientific Software Engineer ? 1018 WV Amsterdam, The Netherlands ? ? ?? ? ? ? EcoGRID-GAN ? ? ? ? ? t: +31 20-5257453 ?f: +31 20-5257431 From cameron.shorter at gmail.com Sat Oct 20 16:03:28 2007 From: cameron.shorter at gmail.com (Cameron Shorter) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] OSGeo access to OWS Context specs Message-ID: <471A5F10.60201@gmail.com> OSGeo Standards, Mapbuilder stores state information inside an OWS Context document, and consequently we may be the largests tester of OWS Context. Every few months we have debates over various parts of the OWS Context spec as developers implement new things. The latest is a discussion shared with Openlayers over how Layer groups should be handled. The current OWS Context spec/schema is globally available at: * http://www.ogcnetwork.net/schemas/owc/0.2.0/ A wiki list of identified issues is at: * https://portal.opengeospatial.org/wiki/twiki/bin/view/CGDIPilot/CgdiipOwsContext Unfortunately, this is only available to OGC members. My next email will be to lobby to make it visible. As part of OGC's mass market strategy, can we please open up access to the OWS Context issue list. This should ensure that specs would benefit from greater testing and increase adoption. An alternative I've considered is setting up an issue tracker as part of Mapbuilder to address OWS Context. (I'd prefer an issue tracker over a wiki as it is easier to manage issues through to completion). -- Cameron Shorter Systems Architect, http://lisasoft.com.au Tel: +61 (0)2 8570 5050 Mob: +61 (0)419 142 254 From jachym.cepicky at gmail.com Sat Oct 20 16:54:09 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] status of translation possibilities In-Reply-To: <200710201600.15839.veen@science.uva.nl> References: <1192713830.8316.9.camel@mellon> <200710201600.15839.veen@science.uva.nl> Message-ID: <1192913649.6858.0.camel@kocour> hi Lourens Veen p??e v So 20. 10. 2007 v 16:00 +0200: > I haven't followed this thread that carefully so maybe it has come up > already, but what about number formats? OL's current scale display uses > K and M for thousands and millions, which is not how it's commonly done > on paper maps. I'd really rather have it write out the full number, but > of course it would then have to use a number format that is appropriate > to the locale. In Dutch, I would like it to say 1:50.000, while in > English that would be 1:50,000. good point j -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071020/1026dccc/attachment.bin From cameron.shorter at gmail.com Sun Oct 21 06:44:22 2007 From: cameron.shorter at gmail.com (Cameron Shorter) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Developement of an external vector symbol In-Reply-To: <4718CAC8.7020705@camptocamp.com> References: <4718CAC8.7020705@camptocamp.com> Message-ID: <471B2D86.3020705@gmail.com> Barbara, This sounds like a great initiative and I'd be surprised if there isn't already a number of projects which transform between SVG and VML. I know that there has been work converting GML to SVG, and hence there shouldn't be much work transforming GeoRSS to SVG. So I suggest: 1. Before you do any development, research available tools that you can use. 2. Create one vector library (either SVG or VML) and use a translation tool to automatically convert from one to the other. 3. Precompile the alternative library. Ie, don't convert SVG to VML on the fly in Javascript. It will slow the browser down. 3. I suspect that SVG would be a better format to support than VML. I believe SVG is used by a number of other projects, eg the Linux Icon libraries. 4. Search for existing icon libraries and consider extending the existing icon libraries rather than creating a new project. Maybe make a geographic version of the library. 5. Think bigger than just OpenLayers. You start the free icon library used by all Open Source GIS projects. It might be a candidate project for: http://wiki.osgeo.org/index.php/OSGeo_Labs (Although you may want to seed your project within Openlayers) Barbara Philippot wrote: > Ian Mayo a ?crit : > >> Barbara, >> I'm afraid I can't offer an opinion on your languages challenge. It >> would be great if there was a pre-existing file-format for vector >> symbols, then we just specify filenames rather than hard-code the >> vectored shapes. >> >> > Yes, this is why we are trying to do. But people would have to specify > two filenames (one in SVG, the other in VML) to replace the pre-existing > symbol catalogue files and it's not still the ideal solution but perhaps > it's the only one. > >> I'm currenlty working on provision of per-entry icons (images) for >> georss feeds. [1]. I've made good progress, but may have to pause my >> efforts for the next 7 days, see my demo at: [2]. Had you considered >> using the symbols on a per-entry fashion in a georss feed? I suspect >> it's fortunate I've got a self-induced development pause: it would >> seem to be worth offering per-entry symbols as well as icons. >> >> Appreciate your thoughts/opinions. >> >> > Well, I hadn't considered using symbols in a georss feed and I don't > know well Georss format. For the moment, symbols are associated with > points geometry with a style.use="symbol" in the same way as images were > associated with style.externalGraphic. So, I haven't looked in detail to > the way you attach icons to GeoRss nodes but it seems that you would be > able to use SVG and VML symbols as well as images. > > I hope to make progress before your development pause ends :-) > >> Ian. >> >> 1. http://trac.openlayers.org/ticket/1073 >> 2. http://dev.openlayers.org/sandbox/ianmayo/branches/rss/openlayers/examples/georss.html >> >> >> > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > -- Cameron Shorter Systems Architect, http://lisasoft.com.au Tel: +61 (0)2 8570 5050 Mob: +61 (0)419 142 254 From cameron.shorter at gmail.com Sun Oct 21 06:48:07 2007 From: cameron.shorter at gmail.com (Cameron Shorter) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] [OSGeo-Standards] OSGeo access to OWS Context specs In-Reply-To: <002701c8135e$76a75330$6401a8c0@SusieandCarl> References: <471A5F10.60201@gmail.com> <002701c8135e$76a75330$6401a8c0@SusieandCarl> Message-ID: <471B2E67.1090208@gmail.com> Forwarding Carl's email, as I'm aware he is not subscribed to all lists he replied to. Carl Reed OGC Account wrote: > I will be out of the office for two weeks. However, as the proposal is > a good one, I would suggest working with Raj, Greg, and Mark Reichardt > to make this happen. > > Also, FYI, as soon as we get out from under all the changes to > documents and the web site for the new OGC IPR policy, we will be > standing up a public OGC Change Request Proposal system whereby anyone > can submit change requests and track their progress. > > Regards > > Carl > > ----- Original Message ----- From: "Cameron Shorter" > > To: ; "Kralidis,Tom [Burlington]" > > Cc: ; "openlayers" > ; > Sent: Saturday, October 20, 2007 2:03 PM > Subject: [OSGeo-Standards] OSGeo access to OWS Context specs > > >> OSGeo Standards, >> >> Mapbuilder stores state information inside an OWS Context document, >> and consequently we may be the largests tester of OWS Context. >> >> Every few months we have debates over various parts of the OWS >> Context spec as developers implement new things. The latest is a >> discussion shared with Openlayers over how Layer groups should be >> handled. >> >> The current OWS Context spec/schema is globally available at: >> * http://www.ogcnetwork.net/schemas/owc/0.2.0/ >> >> A wiki list of identified issues is at: >> * >> https://portal.opengeospatial.org/wiki/twiki/bin/view/CGDIPilot/CgdiipOwsContext >> >> Unfortunately, this is only available to OGC members. My next email >> will be to lobby to make it visible. >> >> >> As part of OGC's mass market strategy, can we please open up access >> to the OWS Context issue list. This should ensure that specs would >> benefit from greater testing and increase adoption. >> >> An alternative I've considered is setting up an issue tracker as part >> of Mapbuilder to address OWS Context. (I'd prefer an issue tracker >> over a wiki as it is easier to manage issues through to completion). >> >> -- >> Cameron Shorter >> Systems Architect, http://lisasoft.com.au >> Tel: +61 (0)2 8570 5050 >> Mob: +61 (0)419 142 254 >> >> _______________________________________________ >> Standards mailing list >> Standards@lists.osgeo.org >> http://lists.osgeo.org/mailman/listinfo/standards > > -- Cameron Shorter Systems Architect, http://lisasoft.com.au Tel: +61 (0)2 8570 5050 Mob: +61 (0)419 142 254 From crschmidt at metacarta.com Sun Oct 21 12:12:15 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Developement of an external vector symbol In-Reply-To: <471B2D86.3020705@gmail.com> References: <4718CAC8.7020705@camptocamp.com> <471B2D86.3020705@gmail.com> Message-ID: <20071021161215.GA22608@metacarta.com> On Sun, Oct 21, 2007 at 08:44:22PM +1000, Cameron Shorter wrote: > 5. Think bigger than just OpenLayers. You start the free icon library > used by all Open Source GIS projects. Or you could use the existing project for all open clipart. http://openclipart.org/ Regards, -- Christopher Schmidt MetaCarta From bluecarto at gmail.com Sun Oct 21 15:18:00 2007 From: bluecarto at gmail.com (Pierre GIRAUD) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Developement of an external vector symbol In-Reply-To: <20071021161215.GA22608@metacarta.com> References: <4718CAC8.7020705@camptocamp.com> <471B2D86.3020705@gmail.com> <20071021161215.GA22608@metacarta.com> Message-ID: Here is another interesting page with already existing svg symbols catalogue. http://wiki.openstreetmap.org/index.php/Osmarender/Symbols Regards, On 10/21/07, Christopher Schmidt wrote: > On Sun, Oct 21, 2007 at 08:44:22PM +1000, Cameron Shorter wrote: > > 5. Think bigger than just OpenLayers. You start the free icon library > > used by all Open Source GIS projects. > > Or you could use the existing project for all open clipart. > > http://openclipart.org/ > > Regards, > -- > Christopher Schmidt > MetaCarta > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > From domenico.febbo at gmail.com Tue Oct 23 04:28:29 2007 From: domenico.febbo at gmail.com (Domenico Febbo) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] proxyHost cgi issue Message-ID: <97236bbb0710230128y928bbe5r66f72cc4abef4e03@mail.gmail.com> I need to use proxy.cgi script to forward "get" request on XMLHttpRequest object to a external host, but I have same problem with the path to add at "url=" field. Moreover I've tried the online script http://dev.openlayers.org/sandbox/docs/examples/proxy.cgi but the http://openlayers.org redirect seems to have a issue because the OpenLayers object isn't found (try with Firebug enabled) and all the relative path are wrong. Somebody have other proxy script for Apache webserver? thanks, Domenico From ahhughes at gmail.com Tue Oct 23 07:19:58 2007 From: ahhughes at gmail.com (Andrew Hughes) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] proxyHost cgi issue In-Reply-To: <97236bbb0710230128y928bbe5r66f72cc4abef4e03@mail.gmail.com> References: <97236bbb0710230128y928bbe5r66f72cc4abef4e03@mail.gmail.com> Message-ID: <5f528cf40710230419h4da059aduaec8bc9465a303f9@mail.gmail.com> http://developer.yahoo.com/javascript/howto-proxy.html On 10/23/07, Domenico Febbo wrote: > > I need to use proxy.cgi script to forward "get" request on > XMLHttpRequest object to a external host, but I have same problem with > the path to add at "url=" field. > Moreover I've tried the online script > http://dev.openlayers.org/sandbox/docs/examples/proxy.cgi > but the http://openlayers.org redirect seems to have a issue because > the OpenLayers object isn't found (try with Firebug enabled) and all > the relative path are wrong. > Somebody have other proxy script for Apache webserver? > thanks, > Domenico > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071023/28fc6745/attachment.html From lynn.forrester at mottmac.com Tue Oct 23 09:41:04 2007 From: lynn.forrester at mottmac.com (LynnF) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Vector Layer and Popups Message-ID: <13364682.post@talk.nabble.com> Hi, I am currently working on trying to show a border around an area of interest on my map. The boundary is a list of points that I read and then load into a LinearRing and then load into a VectorFeature which is then finally placed onto a Vector Layer. The problem I am having is that if this boundary layer is on then the popups from other layers do not work (using the Text Layer). Can anyone suggest a solution to this problem? I have tried adding them in different orders but didn't solve anything. Thanks, Lynn. -- View this message in context: http://www.nabble.com/Vector-Layer-and-Popups-tf4677572.html#a13364682 Sent from the OpenLayers Dev mailing list archive at Nabble.com. From crschmidt at metacarta.com Tue Oct 23 09:51:40 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Vector Layer and Popups In-Reply-To: <13364682.post@talk.nabble.com> References: <13364682.post@talk.nabble.com> Message-ID: <20071023135140.GA11127@metacarta.com> On Tue, Oct 23, 2007 at 06:41:04AM -0700, LynnF wrote: > > Hi, > > I am currently working on trying to show a border around an area of interest > on my map. The boundary is a list of points that I read and then load into > a LinearRing and then load into a VectorFeature which is then finally placed > onto a Vector Layer. > The problem I am having is that if this boundary layer is on then the popups > from other layers do not work (using the Text Layer). I'm assuming you mean "The markers are not clickable." If you add the vector layer before your text layer, this hsould not be the case -- the markers should end up on top. Are you using a SelectFeature control? If so, that will present a problem, since it moves the selected vector layer up to the top of the map, above other elements, which would result in what you are describing. Regards, -- Christopher Schmidt MetaCarta From val.cartei at gmail.com Tue Oct 23 10:19:00 2007 From: val.cartei at gmail.com (Val Cartei) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] overviewMap viewport size to show entire overview map Message-ID: I am trying to modify the size of the overviewMap control from its original settings (180,90) to display the whole overview map. I seem to have achieved this, however when I drag the rectangle to browse the overview, its map scrolls, whereas it should stay fixed in the overview viewport. If I restrict the extents of the map to its size it doesn't work either. Any suggestions? Here's the code: //overviewMap object to resize to match the size of the overview map ovmap_size=new OpenLayers.Size ($("img#OpenLayersDiv43").width(),$("img#OpenLayersDiv43").height()); scroller.update(ovmap_size); //restrict the extent of the overview map scroller.ovmap.restrictedExtent=new OpenLayers.Bounds.fromSize(ovmap_size); -- Best Regards, Valentina Cartei UK Tel. +44 (0)796 6882820 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071023/d742cbae/attachment.html From jing at pelmorex.com Tue Oct 23 13:28:53 2007 From: jing at pelmorex.com (Jim Ing) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] proxyHost cgi issue (Domenico Febbo) In-Reply-To: References: Message-ID: Domenico, Here's one that I'm using: http://www.jmarshall.com/tools/cgiproxy/ Just make sure your Apache httpd.conf has these settings: LoadModule cgi_module modules/mod_cgi.so "> Options Indexes FollowSymLinks ExecCGI AddHandler cgi-script .cgi Also, you might need to edit the first line of nph-proxy.cgi to point to your Perl #!/usr/bin/perl Jim -----Original Message----- Date: Tue, 23 Oct 2007 10:28:29 +0200 From: "Domenico Febbo" Subject: [OpenLayers-Dev] proxyHost cgi issue To: dev@openlayers.org I need to use proxy.cgi script to forward "get" request on XMLHttpRequest object to a external host, but I have same problem with the path to add at "url=" field. Moreover I've tried the online script http://dev.openlayers.org/sandbox/docs/examples/proxy.cgi but the http://openlayers.org redirect seems to have a issue because the OpenLayers object isn't found (try with Firebug enabled) and all the relative path are wrong. Somebody have other proxy script for Apache webserver? thanks, Domenico From jachym.cepicky at gmail.com Tue Oct 23 13:58:42 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Layers into Groups Message-ID: <1193162322.7428.12.camel@kocour> Hi, could anybody of you, guys, have a look at new "custom LayerSwitcher" example [1] ? I hope, it works on all platforms Comments are more then welcomed Jachym P.S. You will have to probably zoom to Canada, go get all layers work [1] http://dev.openlayers.org/sandbox/jachym/openlayers/examples/layerswitcher.html Jachym Cepicky p??e v Po 15. 10. 2007 v 13:20 +0200: > Hi developers, > I have implemented ordering map layers into groups [1]. You have to > specify new layer attribute group:"Group Name", e.g.: > > var dm = new OpenLayers.Layer.WMS( "DM Solutions Transit", > "http://www2.dmsolutions.ca/cgi-bin/mswms_gmap", > {layers: "rail,road", > transparent: "true", format: "image/png" > }, > { > group:"Transit" > }); > > You can define group hierarchy using "/" mark: > > .... > group: "Transit/Roads" > .... > > Only changed file is LayerSwitcher.js > > For example see [2] > > Hope, you'll like it > > Jachym > > [1] http://trac.openlayers.org/ticket/1086 > [2] > http://dev.openlayers.org/sandbox/jachym/openlayers/examples/layer-groups.html > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071023/3d716e5d/attachment.bin From ludwigbrinckmann at gmail.com Tue Oct 23 16:06:38 2007 From: ludwigbrinckmann at gmail.com (Ludwig M Brinckmann) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Layers into Groups In-Reply-To: <1193162322.7428.12.camel@kocour> References: <1193162322.7428.12.camel@kocour> Message-ID: <3f7a6e1c0710231306q5821daeex21517c8d582cf2ee@mail.gmail.com> Hi Jachym, for me on XP and Firefox 2.0.0.8 this appears quite unresponsive, even though after a bit of clicking something eventually happens. I am not sure about the graphical solution as I do not really know when a layer is switched on or off. A while ago I experimented a bit with a custom layer switcher outside the map window, it still seems to be working at http://www.yunnanexplorer.com/mars.html There were several ideas in it: First of all ignore the tick-boxes -- they flicker a layer on and off. This is not very intuitive. I thought once having a special effects div hidden with every layer that could be turned on and off. Flickering a layer, moving a sliding window over a layer, inverting colours, image stretching could all go in there. Never followed through. The sliders allow to adjust transparency of a layer: red all the way to the left is off, green is 100% opaque. The rest somewhere in the middle: good way if you want to look at data in more than one map at a time. Furthermore it is possible to drag the layers up and down, changing the stacking/drawing order. I know that conflicts with the baselayer concept of OL. These effects were done with scriptaculous and do not really work that well under IE, but someone with more patience/time than me could probably make them work. It is now over a year old and I have not done any work on it, so some of the layers have ceased to function. I think building better/different layer switchers, particularly with grouping, is quite important... Ludwig On 23/10/2007, Jachym Cepicky wrote: > > Hi, > > could anybody of you, guys, have a look at new "custom LayerSwitcher" > example [1] ? > > I hope, it works on all platforms > > Comments are more then welcomed > > Jachym > > P.S. You will have to probably zoom to Canada, go get all layers work > > [1] > > http://dev.openlayers.org/sandbox/jachym/openlayers/examples/layerswitcher.html > > Jachym Cepicky p??e v Po 15. 10. 2007 v 13:20 +0200: > > Hi developers, > > I have implemented ordering map layers into groups [1]. You have to > > specify new layer attribute group:"Group Name", e.g.: > > > > var dm = new OpenLayers.Layer.WMS( "DM Solutions Transit", > > "http://www2.dmsolutions.ca/cgi-bin/mswms_gmap", > > {layers: "rail,road", > > transparent: "true", format: "image/png" > > }, > > { > > group:"Transit" > > }); > > > > You can define group hierarchy using "/" mark: > > > > .... > > group: "Transit/Roads" > > .... > > > > Only changed file is LayerSwitcher.js > > > > For example see [2] > > > > Hope, you'll like it > > > > Jachym > > > > [1] http://trac.openlayers.org/ticket/1086 > > [2] > > > http://dev.openlayers.org/sandbox/jachym/openlayers/examples/layer-groups.html > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev > -- > Jachym Cepicky > e-mail: jachym.cepicky@gmail.com > URL: http://les-ejk.cz > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071023/da06f8ec/attachment.html From lorenzo at ominiverdi.com Tue Oct 23 18:21:10 2007 From: lorenzo at ominiverdi.com (Lorenzo Becchi) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Layers into Groups In-Reply-To: <3f7a6e1c0710231306q5821daeex21517c8d582cf2ee@mail.gmail.com> References: <1193162322.7428.12.camel@kocour> <3f7a6e1c0710231306q5821daeex21517c8d582cf2ee@mail.gmail.com> Message-ID: <471E73D6.8020506@ominiverdi.com> Jachym, I see that all legend layouts can vary a lot. what about Tim's idea to use templates? I'm pretty scared about it but probably is a good solution. Tim, do you think templates could really apply to Layer switcher without getting mad? if yes can you "enlighten our way"? ciao lorenzo > > > A while ago I experimented a bit with a custom layer switcher outside > the map window, it still seems to be working at > > http://www.yunnanexplorer.com/mars.html > you've never tried it with Safari. Layers names disappears if you click on them. :-| > > There were several ideas in it: > First of all ignore the tick-boxes -- they flicker a layer on and off. > This is not very intuitive. I thought once having a special effects > div hidden with every layer that could be turned on and off. > Flickering a layer, moving a sliding window over a layer, inverting > colours, image stretching could all go in there. Never followed through. > > The sliders allow to adjust transparency of a layer: red all the way > to the left is off, green is 100% opaque. The rest somewhere in the > middle: good way if you want to look at data in more than one map at a > time. > > Furthermore it is possible to drag the layers up and down, changing > the stacking/drawing order. I know that conflicts with the baselayer > concept of OL. > > These effects were done with scriptaculous and do not really work that > well under IE, but someone with more patience/time than me could > probably make them work. It is now over a year old and I have not done > any work on it, so some of the layers have ceased to function. > > I think building better/different layer switchers, particularly with > grouping, is quite important... > > Ludwig > > > > > > > On 23/10/2007, * Jachym Cepicky* > wrote: > > Hi, > > could anybody of you, guys, have a look at new "custom LayerSwitcher" > example [1] ? > > I hope, it works on all platforms > > Comments are more then welcomed > > Jachym > > P.S. You will have to probably zoom to Canada, go get all layers work > > [1] > http://dev.openlayers.org/sandbox/jachym/openlayers/examples/layerswitcher.html > > Jachym Cepicky p??e v Po 15. 10. 2007 v 13:20 +0200: > > Hi developers, > > I have implemented ordering map layers into groups [1]. You have to > > specify new layer attribute group:"Group Name", e.g.: > > > > var dm = new OpenLayers.Layer.WMS ( "DM Solutions Transit", > > "http://www2.dmsolutions.ca/cgi-bin/mswms_gmap", > > {layers: "rail,road", > > transparent: "true", format: "image/png" > > }, > > { > > group:"Transit" > > }); > > > > You can define group hierarchy using "/" mark: > > > > .... > > group: "Transit/Roads" > > .... > > > > Only changed file is LayerSwitcher.js > > > > For example see [2] > > > > Hope, you'll like it > > > > Jachym > > > > [1] http://trac.openlayers.org/ticket/1086 > > [2] > > > http://dev.openlayers.org/sandbox/jachym/openlayers/examples/layer-groups.html > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev > -- > Jachym Cepicky > e-mail: jachym.cepicky@gmail.com > URL: http://les-ejk.cz > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > From jachym.cepicky at gmail.com Wed Oct 24 03:59:17 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Layers into Groups In-Reply-To: <3f7a6e1c0710231306q5821daeex21517c8d582cf2ee@mail.gmail.com> References: <1193162322.7428.12.camel@kocour> <3f7a6e1c0710231306q5821daeex21517c8d582cf2ee@mail.gmail.com> Message-ID: <1193212758.6297.21.camel@kocour> Hi, Ludwig M Brinckmann p??e v ?t 23. 10. 2007 v 21:06 +0100: > Hi Jachym, > > for me on XP and Firefox 2.0.0.8 this appears quite unresponsive, even > though after a bit of clicking something eventually happens. I am not > sure about the graphical solution as I do not really know when a layer > is switched on or off. Some layers were active only, if you zoomed to region around Canada (fixed). I completely agree with you, that the "design" of the page "could" be more intuitive -- sorry, no graphical designer. > > A while ago I experimented a bit with a custom layer switcher outside > the map window, it still seems to be working at > > http://www.yunnanexplorer.com/mars.html > > There were several ideas in it: > First of all ignore the tick-boxes -- they flicker a layer on and off. > This is not very intuitive. I thought once having a special effects > div hidden with every layer that could be turned on and off. > Flickering a layer, moving a sliding window over a layer, inverting > colours, image stretching could all go in there. Never followed > through. nice demo > > The sliders allow to adjust transparency of a layer: red all the way > to the left is off, green is 100% opaque. The rest somewhere in the > middle: good way if you want to look at data in more than one map at a > time. this is somehow more intuitive, than my previous layer-buttons.html example. Sliders are better for opacity settings, but I did not know, how to get them in current LayerSwitcher dimensions. > > > Furthermore it is possible to drag the layers up and down, changing > the stacking/drawing order. I know that conflicts with the baselayer > concept of OL. Dragging is definitely more intuitive, than buttons (again layer-buttons.html in my sandbox), when it is working. Together with layer groups (layer-groups.html), the code would get pretty complicated.. > > These effects were done with scriptaculous and do not really work that > well under IE, but someone with more patience/time than me could > probably make them work. It is now over a year old and I have not done > any work on it, so some of the layers have ceased to function. > > I think building better/different layer switchers, particularly with > grouping, is quite important... I'm working on it. If others like it, I hope, it would appear in OpenLayers trunk - I'm not avare of any other impediment. Jachym > > Ludwig > > > > > > > On 23/10/2007, Jachym Cepicky wrote: > Hi, > > could anybody of you, guys, have a look at new "custom > LayerSwitcher" > example [1] ? > > I hope, it works on all platforms > > Comments are more then welcomed > > Jachym > > P.S. You will have to probably zoom to Canada, go get all > layers work > > [1] > http://dev.openlayers.org/sandbox/jachym/openlayers/examples/layerswitcher.html > > Jachym Cepicky p??e v Po 15. 10. 2007 v 13:20 +0200: > > Hi developers, > > I have implemented ordering map layers into groups [1]. You > have to > > specify new layer attribute group:"Group Name", e.g.: > > > > var dm = new OpenLayers.Layer.WMS ( "DM Solutions Transit", > > > "http://www2.dmsolutions.ca/cgi-bin/mswms_gmap", > > {layers: "rail,road", > > transparent: "true", format: "image/png" > > }, > > { > > group:"Transit" > > }); > > > > You can define group hierarchy using "/" mark: > > > > .... > > group: "Transit/Roads" > > .... > > > > Only changed file is LayerSwitcher.js > > > > For example see [2] > > > > Hope, you'll like it > > > > Jachym > > > > [1] http://trac.openlayers.org/ticket/1086 > > [2] > > > http://dev.openlayers.org/sandbox/jachym/openlayers/examples/layer-groups.html > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev > -- > Jachym Cepicky > e-mail: jachym.cepicky@gmail.com > URL: http://les-ejk.cz > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > > -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071024/c76f73ae/attachment.bin From jachym.cepicky at gmail.com Wed Oct 24 04:18:07 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Layers into Groups In-Reply-To: <471E73D6.8020506@ominiverdi.com> References: <1193162322.7428.12.camel@kocour> <3f7a6e1c0710231306q5821daeex21517c8d582cf2ee@mail.gmail.com> <471E73D6.8020506@ominiverdi.com> Message-ID: <1193213887.6297.27.camel@kocour> Hi, Lorenzo Becchi p??e v St 24. 10. 2007 v 00:21 +0200: > Jachym, I see that all legend layouts can vary a lot. > what about Tim's idea to use templates? > I'm pretty scared about it but probably is a good solution. > Tim, do you think templates could really apply to Layer switcher without > getting mad? if yes can you "enlighten our way"? > > ciao > lorenzo How should the templates look like? My approach is, have everything (layers, groups, etc.) in separate divs, so they are now customisable via CSS styles. The example should be fixed for IE as well now. Jachym > > > > > > > > A while ago I experimented a bit with a custom layer switcher outside > > the map window, it still seems to be working at > > > > http://www.yunnanexplorer.com/mars.html > > > you've never tried it with Safari. Layers names disappears if you click > on them. > :-| > > > > There were several ideas in it: > > First of all ignore the tick-boxes -- they flicker a layer on and off. > > This is not very intuitive. I thought once having a special effects > > div hidden with every layer that could be turned on and off. > > Flickering a layer, moving a sliding window over a layer, inverting > > colours, image stretching could all go in there. Never followed through. > > > > The sliders allow to adjust transparency of a layer: red all the way > > to the left is off, green is 100% opaque. The rest somewhere in the > > middle: good way if you want to look at data in more than one map at a > > time. > > > > Furthermore it is possible to drag the layers up and down, changing > > the stacking/drawing order. I know that conflicts with the baselayer > > concept of OL. > > > > These effects were done with scriptaculous and do not really work that > > well under IE, but someone with more patience/time than me could > > probably make them work. It is now over a year old and I have not done > > any work on it, so some of the layers have ceased to function. > > > > I think building better/different layer switchers, particularly with > > grouping, is quite important... > > > > Ludwig > > > > > > > > > > > > > > On 23/10/2007, * Jachym Cepicky* > > wrote: > > > > Hi, > > > > could anybody of you, guys, have a look at new "custom LayerSwitcher" > > example [1] ? > > > > I hope, it works on all platforms > > > > Comments are more then welcomed > > > > Jachym > > > > P.S. You will have to probably zoom to Canada, go get all layers work > > > > [1] > > http://dev.openlayers.org/sandbox/jachym/openlayers/examples/layerswitcher.html > > > > Jachym Cepicky p??e v Po 15. 10. 2007 v 13:20 +0200: > > > Hi developers, > > > I have implemented ordering map layers into groups [1]. You have to > > > specify new layer attribute group:"Group Name", e.g.: > > > > > > var dm = new OpenLayers.Layer.WMS ( "DM Solutions Transit", > > > "http://www2.dmsolutions.ca/cgi-bin/mswms_gmap", > > > {layers: "rail,road", > > > transparent: "true", format: "image/png" > > > }, > > > { > > > group:"Transit" > > > }); > > > > > > You can define group hierarchy using "/" mark: > > > > > > .... > > > group: "Transit/Roads" > > > .... > > > > > > Only changed file is LayerSwitcher.js > > > > > > For example see [2] > > > > > > Hope, you'll like it > > > > > > Jachym > > > > > > [1] http://trac.openlayers.org/ticket/1086 > > > [2] > > > > > http://dev.openlayers.org/sandbox/jachym/openlayers/examples/layer-groups.html > > > _______________________________________________ > > > Dev mailing list > > > Dev@openlayers.org > > > http://openlayers.org/mailman/listinfo/dev > > -- > > Jachym Cepicky > > e-mail: jachym.cepicky@gmail.com > > URL: http://les-ejk.cz > > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > > > > > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev > > > > > > > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev > > -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071024/3c2f9534/attachment.bin From domenico.febbo at gmail.com Wed Oct 24 09:10:34 2007 From: domenico.febbo at gmail.com (Domenico Febbo) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] proxyHost cgi issue Message-ID: <97236bbb0710240610x194c352dlac9919436488150f@mail.gmail.com> Hi, with same modify I've configured the OpenLayers script "proxy.cgi" for my lan (adding http_proxy configuration). thanks, Jim for that url. I've this question: the same cgi is configurable for IIS web server or there is other script to workaround the "samePolicy" restriction? Domenico > Date: Tue, 23 Oct 2007 13:28:53 -0400 > From: "Jim Ing" > Subject: Re: [OpenLayers-Dev] proxyHost cgi issue (Domenico Febbo) > To: > Message-ID: > > Content-Type: text/plain; charset="us-ascii" > > Domenico, > > Here's one that I'm using: > http://www.jmarshall.com/tools/cgiproxy/ > > Just make sure your Apache httpd.conf has these settings: > > LoadModule cgi_module modules/mod_cgi.so > > "> > Options Indexes FollowSymLinks ExecCGI > > > AddHandler cgi-script .cgi > > Also, you might need to edit the first line of nph-proxy.cgi to point to > your Perl > #!/usr/bin/perl > > > Jim > > > -----Original Message----- > Date: Tue, 23 Oct 2007 10:28:29 +0200 > From: "Domenico Febbo" > Subject: [OpenLayers-Dev] proxyHost cgi issue > To: dev@openlayers.org > > I need to use proxy.cgi script to forward "get" request on > XMLHttpRequest object to a external host, but I have same problem with > the path to add at "url=" field. > Moreover I've tried the online script > http://dev.openlayers.org/sandbox/docs/examples/proxy.cgi > but the http://openlayers.org redirect seems to have a issue because > the OpenLayers object isn't found (try with Firebug enabled) and all > the relative path are wrong. > Somebody have other proxy script for Apache webserver? > thanks, > Domenico > > > ------------------------------ From bartvde at osgis.nl Thu Oct 25 04:57:37 2007 From: bartvde at osgis.nl (Bart van den Eijnden (OSGIS)) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] deactivating other handlers Message-ID: Hi devs, I've written a FeatureInfo Handler, which only uses click. However, the DragPan handler for instance is still active. What is the best way to deactivate these other handlers? Ofcourse I can also implement mousedown etc. and not let the event propagate, but that seems a bit cumbersome. Is there a way to make a Handler exclusive, or to remove all other handlers from the map? Best regards, Bart -- Bart van den Eijnden OSGIS, Open Source GIS http://www.osgis.nl From jachym.cepicky at gmail.com Thu Oct 25 07:57:45 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg Message-ID: <1193313465.6482.6.camel@kocour> Hi, in OpenLayers.Layer.Tilecache, around line 55, there is this.extension = (this.extension == 'jpeg') ? 'jpg' : this.extension; in tilecache/TileCache/Layer.py, around line 80, there is if extension.lower() == 'jpg': extension = 'jpeg' # MIME IMHO, it does not make sense, if I understand this well, you can never access JPEG TileCache from OpenLayers Do I understand it well? Thanks Jachym -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071025/93eb6131/attachment.bin From pspencer at dmsolutions.ca Thu Oct 25 08:49:06 2007 From: pspencer at dmsolutions.ca (Paul Spencer) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg In-Reply-To: <1193313465.6482.6.camel@kocour> References: <1193313465.6482.6.camel@kocour> Message-ID: <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> Jachym, if I understand it correctly, the image saved on disk has an extension of .jpg while the mime type returned by the web server with a JPEG image is image/jpeg. So the code is fine and is doing the right thing. You just need to specify the correct extension in OpenLayers (either 'jpg' or 'jpeg' will do) and in TileCache config (again, probably either will do). Paul On 25-Oct-07, at 7:57 AM, Jachym Cepicky wrote: > Hi, > > in OpenLayers.Layer.Tilecache, around line 55, there is > > this.extension = (this.extension == 'jpeg') ? 'jpg' : this.extension; > > in tilecache/TileCache/Layer.py, around line 80, there is > > if extension.lower() == 'jpg': extension = 'jpeg' # MIME > > IMHO, it does not make sense, if I understand this well, you can never > access JPEG TileCache from OpenLayers > > Do I understand it well? > > Thanks > > Jachym > -- > Jachym Cepicky > e-mail: jachym.cepicky@gmail.com > URL: http://les-ejk.cz > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev +-----------------------------------------------------------------+ |Paul Spencer pspencer@dmsolutions.ca | +-----------------------------------------------------------------+ |Chief Technology Officer | |DM Solutions Group Inc http://www.dmsolutions.ca/ | +-----------------------------------------------------------------+ From jachym.cepicky at gmail.com Thu Oct 25 08:51:17 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] alt or title text on Layer.Marker Message-ID: <1193316677.6482.12.camel@kocour> Hi, maybe, it would be good to have option for OpenLayers.Layer.Marker, which would enable displaying alt text of icons on mouse over. The alt attribute could be e.g. Title of OL.Layer.Text Is there such possibility, which would display alt or other attribute, while moving with mouse over icon? thanks jachym -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071025/3f8603cc/attachment.bin From jachym.cepicky at gmail.com Thu Oct 25 08:53:04 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg In-Reply-To: <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> References: <1193313465.6482.6.camel@kocour> <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> Message-ID: <1193316784.6482.15.camel@kocour> Hi, no, if I understand it right, TileCache will always save the files with extension "jpeg", while OpenLayers will always request "jpg" files. Jachym Paul Spencer p??e v ?t 25. 10. 2007 v 08:49 -0400: > Jachym, > > if I understand it correctly, the image saved on disk has an > extension of .jpg while the mime type returned by the web server with > a JPEG image is image/jpeg. So the code is fine and is doing the > right thing. You just need to specify the correct extension in > OpenLayers (either 'jpg' or 'jpeg' will do) and in TileCache config > (again, probably either will do). > > Paul > > On 25-Oct-07, at 7:57 AM, Jachym Cepicky wrote: > > > Hi, > > > > in OpenLayers.Layer.Tilecache, around line 55, there is > > > > this.extension = (this.extension == 'jpeg') ? 'jpg' : this.extension; > > > > in tilecache/TileCache/Layer.py, around line 80, there is > > > > if extension.lower() == 'jpg': extension = 'jpeg' # MIME > > > > IMHO, it does not make sense, if I understand this well, you can never > > access JPEG TileCache from OpenLayers > > > > Do I understand it well? > > > > Thanks > > > > Jachym > > -- > > Jachym Cepicky > > e-mail: jachym.cepicky@gmail.com > > URL: http://les-ejk.cz > > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > > > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev > > +-----------------------------------------------------------------+ > |Paul Spencer pspencer@dmsolutions.ca | > +-----------------------------------------------------------------+ > |Chief Technology Officer | > |DM Solutions Group Inc http://www.dmsolutions.ca/ | > +-----------------------------------------------------------------+ > > > > > -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071025/d0ca51d4/attachment.bin From pspencer at dmsolutions.ca Thu Oct 25 08:57:02 2007 From: pspencer at dmsolutions.ca (Paul Spencer) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg In-Reply-To: <1193316784.6482.15.camel@kocour> References: <1193313465.6482.6.camel@kocour> <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> <1193316784.6482.15.camel@kocour> Message-ID: <720A0DE4-3A47-47F7-9908-0389C15CC16A@dmsolutions.ca> No. TileCache saves the image with a .jpg extension but uses 'jpeg' for the mime type of the returned content. The underlying library, PIL, will do that automagically for you. Cheers Paul On 25-Oct-07, at 8:53 AM, Jachym Cepicky wrote: > Hi, > > no, if I understand it right, TileCache will always save the files > with > extension "jpeg", while OpenLayers will always request "jpg" files. > > Jachym > > > Paul Spencer p??e v ?t 25. 10. 2007 v 08:49 -0400: >> Jachym, >> >> if I understand it correctly, the image saved on disk has an >> extension of .jpg while the mime type returned by the web server with >> a JPEG image is image/jpeg. So the code is fine and is doing the >> right thing. You just need to specify the correct extension in >> OpenLayers (either 'jpg' or 'jpeg' will do) and in TileCache config >> (again, probably either will do). >> >> Paul >> >> On 25-Oct-07, at 7:57 AM, Jachym Cepicky wrote: >> >>> Hi, >>> >>> in OpenLayers.Layer.Tilecache, around line 55, there is >>> >>> this.extension = (this.extension == 'jpeg') ? 'jpg' : >>> this.extension; >>> >>> in tilecache/TileCache/Layer.py, around line 80, there is >>> >>> if extension.lower() == 'jpg': extension = 'jpeg' # MIME >>> >>> IMHO, it does not make sense, if I understand this well, you can >>> never >>> access JPEG TileCache from OpenLayers >>> >>> Do I understand it well? >>> >>> Thanks >>> >>> Jachym >>> -- >>> Jachym Cepicky >>> e-mail: jachym.cepicky@gmail.com >>> URL: http://les-ejk.cz >>> GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub >>> >>> _______________________________________________ >>> Dev mailing list >>> Dev@openlayers.org >>> http://openlayers.org/mailman/listinfo/dev >> >> +-----------------------------------------------------------------+ >> |Paul Spencer pspencer@dmsolutions.ca | >> +-----------------------------------------------------------------+ >> |Chief Technology Officer | >> |DM Solutions Group Inc http://www.dmsolutions.ca/ | >> +-----------------------------------------------------------------+ >> >> >> >> >> > -- > Jachym Cepicky > e-mail: jachym.cepicky@gmail.com > URL: http://les-ejk.cz > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > +-----------------------------------------------------------------+ |Paul Spencer pspencer@dmsolutions.ca | +-----------------------------------------------------------------+ |Chief Technology Officer | |DM Solutions Group Inc http://www.dmsolutions.ca/ | +-----------------------------------------------------------------+ From crschmidt at metacarta.com Thu Oct 25 08:59:32 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] deactivating other handlers In-Reply-To: References: Message-ID: <20071025125932.GC3540@metacarta.com> On Thu, Oct 25, 2007 at 10:57:37AM +0200, Bart van den Eijnden (OSGIS) wrote: > Hi devs, > > I've written a FeatureInfo Handler, which only uses click. However, the > DragPan handler for instance is still active. > > What is the best way to deactivate these other handlers? > > Ofcourse I can also implement mousedown etc. and not let the event > propagate, but that seems a bit cumbersome. > > Is there a way to make a Handler exclusive, or to remove all other handlers > from the map? Controls should use registerPriority when registering events on activate -- registerPriority says 'me first!'. So, even though dragPan would get the event, if you registerPriority, you'll get it first, and then stop it so that dragPan doesn't get it. If that doesn't work somewhere, then it would be good to know, but I think it should. Regards, -- Christopher Schmidt MetaCarta From crschmidt at metacarta.com Thu Oct 25 09:02:48 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg In-Reply-To: <1193316784.6482.15.camel@kocour> References: <1193313465.6482.6.camel@kocour> <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> <1193316784.6482.15.camel@kocour> Message-ID: <20071025130248.GE3540@metacarta.com> On Thu, Oct 25, 2007 at 02:53:04PM +0200, Jachym Cepicky wrote: > Hi, > > no, if I understand it right, TileCache will always save the files with > extension "jpeg", while OpenLayers will always request "jpg" files. You're correct. It's a bug in the TileCache layer. I've bumped into a couple times, but haven't fixed it. Please feel free to file a ticket with a patch. Regards, -- Christopher Schmidt MetaCarta From gonrial at gmail.com Thu Oct 25 09:03:39 2007 From: gonrial at gmail.com (Ezequiel Gonzalez Rial) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] alt or title text on Layer.Marker In-Reply-To: <1193316677.6482.12.camel@kocour> References: <1193316677.6482.12.camel@kocour> Message-ID: Jachym, I've developed something arround that (Ticket #751) wich is not yet on the trunk, it's a label that apperars next to the marker (by css you can set it above, below or wherever you want), as it inheritance from marker you don't loose anything you have already coded. -- Best Regards, Ezequiel Gonzalez Rial (gonrial@gmail.com) 2007/10/25, Jachym Cepicky : > Hi, > > maybe, it would be good to have option for OpenLayers.Layer.Marker, > which would enable displaying alt text of icons on mouse over. The alt > attribute could be e.g. Title of OL.Layer.Text > > Is there such possibility, which would display alt or other attribute, > while moving with mouse over icon? > > thanks > > jachym > > -- > Jachym Cepicky > e-mail: jachym.cepicky@gmail.com > URL: http://les-ejk.cz > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev From crschmidt at metacarta.com Thu Oct 25 09:08:21 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg In-Reply-To: <720A0DE4-3A47-47F7-9908-0389C15CC16A@dmsolutions.ca> References: <1193313465.6482.6.camel@kocour> <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> <1193316784.6482.15.camel@kocour> <720A0DE4-3A47-47F7-9908-0389C15CC16A@dmsolutions.ca> Message-ID: <20071025130821.GF3540@metacarta.com> On Thu, Oct 25, 2007 at 08:57:02AM -0400, Paul Spencer wrote: > No. TileCache saves the image with a .jpg extension but uses 'jpeg' > for the mime type of the returned content. The underlying library, > PIL, will do that automagically for you. No. 1. PIL is not neccesarily involved in transactions. It is only used with meta-tiling, so invoking the name of PIL here is irrelevant. 2. The last parts of the diskcache key are: "%03d.%s" % (int(tile.y) % 1000, tile.layer.extension) 3. The Layer class converts the extension: if extension.lower() == 'jpg': extension = 'jpeg' # MIME self.extension = extension.lower() Jachym is right. The Layer.TileCache is wrong. Regards, -- Christopher Schmidt MetaCarta From pspencer at dmsolutions.ca Thu Oct 25 09:11:19 2007 From: pspencer at dmsolutions.ca (Paul Spencer) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg In-Reply-To: <20071025130821.GF3540@metacarta.com> References: <1193313465.6482.6.camel@kocour> <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> <1193316784.6482.15.camel@kocour> <720A0DE4-3A47-47F7-9908-0389C15CC16A@dmsolutions.ca> <20071025130821.GF3540@metacarta.com> Message-ID: Ok, I take that back :) I was looking at the PIL code and it takes the extension (jpeg) as a format option but I hadn't considered the case where you were not using PIL. But in either case, I now realize the final image name is being created incorrectly regardless of whether PIL is used or not. Paul On 25-Oct-07, at 9:08 AM, Christopher Schmidt wrote: > On Thu, Oct 25, 2007 at 08:57:02AM -0400, Paul Spencer wrote: >> No. TileCache saves the image with a .jpg extension but uses 'jpeg' >> for the mime type of the returned content. The underlying library, >> PIL, will do that automagically for you. > > No. > > 1. PIL is not neccesarily involved in transactions. It is only used > with meta-tiling, so invoking the name of PIL here is irrelevant. > 2. The last parts of the diskcache key are: > "%03d.%s" % (int(tile.y) % 1000, tile.layer.extension) > 3. The Layer class converts the extension: > > if extension.lower() == 'jpg': extension = 'jpeg' # MIME > self.extension = extension.lower() > > Jachym is right. The Layer.TileCache is wrong. > > Regards, > -- > Christopher Schmidt > MetaCarta +-----------------------------------------------------------------+ |Paul Spencer pspencer@dmsolutions.ca | +-----------------------------------------------------------------+ |Chief Technology Officer | |DM Solutions Group Inc http://www.dmsolutions.ca/ | +-----------------------------------------------------------------+ From jachym.cepicky at gmail.com Thu Oct 25 10:11:50 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg In-Reply-To: <20071025130248.GE3540@metacarta.com> References: <1193313465.6482.6.camel@kocour> <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> <1193316784.6482.15.camel@kocour> <20071025130248.GE3540@metacarta.com> Message-ID: <1193321510.6482.20.camel@kocour> Hi, http://dev.tilecache.org/ticket/555 Cheers Jachym Christopher Schmidt p??e v ?t 25. 10. 2007 v 09:02 -0400: > On Thu, Oct 25, 2007 at 02:53:04PM +0200, Jachym Cepicky wrote: > > Hi, > > > > no, if I understand it right, TileCache will always save the files with > > extension "jpeg", while OpenLayers will always request "jpg" files. > > You're correct. It's a bug in the TileCache layer. I've bumped into a > couple times, but haven't fixed it. Please feel free to file a ticket > with a patch. > > Regards, -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071025/317ac2bc/attachment.bin From crschmidt at metacarta.com Thu Oct 25 11:17:47 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg In-Reply-To: <1193321510.6482.20.camel@kocour> References: <1193313465.6482.6.camel@kocour> <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> <1193316784.6482.15.camel@kocour> <20071025130248.GE3540@metacarta.com> <1193321510.6482.20.camel@kocour> Message-ID: <20071025151747.GA12250@metacarta.com> On Thu, Oct 25, 2007 at 04:11:50PM +0200, Jachym Cepicky wrote: > Hi, > > http://dev.tilecache.org/ticket/555 Actually, it needs to be fixed on the client side, not the server side. The server side WMS uses the extension to request data from the WMS server as "image/"+layer.extension -- this clearly won't work with image/jpg for most WMS servers. Instead, the TileCache Layer in openlayers needs to be changed to reverse its current logic. Regards, -- Christopher Schmidt MetaCarta From bluecarto at gmail.com Thu Oct 25 11:26:35 2007 From: bluecarto at gmail.com (Pierre GIRAUD) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] patched ticket to review Message-ID: Hello all, As proposed by Chris on the IRC, here's a list of ticket I would like to see reviewed : #819 : Linestring draws filled in IE #965 : rendered geometries blink sometimes in the VML renderer #1042 : renderer flashes black #1107 : handlers have to check if layer exists before destroying it Regards Pierre From jachym.cepicky at gmail.com Thu Oct 25 11:38:58 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg In-Reply-To: <20071025151747.GA12250@metacarta.com> References: <1193313465.6482.6.camel@kocour> <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> <1193316784.6482.15.camel@kocour> <20071025130248.GE3540@metacarta.com> <1193321510.6482.20.camel@kocour> <20071025151747.GA12250@metacarta.com> Message-ID: <1193326738.6482.22.camel@kocour> Hmm, this makes sense is http://trac.openlayers.org/ticket/1110 all right for you? Jachym Christopher Schmidt p??e v ?t 25. 10. 2007 v 11:17 -0400: > On Thu, Oct 25, 2007 at 04:11:50PM +0200, Jachym Cepicky wrote: > > Hi, > > > > http://dev.tilecache.org/ticket/555 > > Actually, it needs to be fixed on the client side, not the server side. > The server side WMS uses the extension to request data from the WMS > server as "image/"+layer.extension -- this clearly won't work with > image/jpg for most WMS servers. Instead, the TileCache Layer in > openlayers needs to be changed to reverse its current logic. > > Regards, -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071025/10027318/attachment.bin From crschmidt at metacarta.com Thu Oct 25 11:48:24 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] jpeg vs jpg In-Reply-To: <1193326738.6482.22.camel@kocour> References: <1193313465.6482.6.camel@kocour> <82C4DC2C-755D-4B0A-AD0E-4C942938EF14@dmsolutions.ca> <1193316784.6482.15.camel@kocour> <20071025130248.GE3540@metacarta.com> <1193321510.6482.20.camel@kocour> <20071025151747.GA12250@metacarta.com> <1193326738.6482.22.camel@kocour> Message-ID: <20071025154824.GA13839@metacarta.com> On Thu, Oct 25, 2007 at 05:38:58PM +0200, Jachym Cepicky wrote: > Hmm, this makes sense > > is http://trac.openlayers.org/ticket/1110 all right for you? Yep. Thanks. WIll take care of it sometime soon. Regards, -- Christopher Schmidt MetaCarta From erik.uzureau at metacarta.com Thu Oct 25 20:16:33 2007 From: erik.uzureau at metacarta.com (Erik Uzureau) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] OL Vector Layer: What are the Limitations? Message-ID: <6ae3fb590710251716k559d42a1nb24993ec75bd7a28@mail.gmail.com> Dear users, dev, Someone asked me today what the max number of coordinates could be drawn for a geometry with the OL Vector Layer and I did not know how to really answer. I asked cr5 and he gave me some off the cuff estimates, which I have added to the FAQ here: http://trac.openlayers.org/wiki/FrequentlyAskedQuestions#WhatisthemaximumnumberofCoordinatesFeaturesIcandrawwithaVectorlayer Basically, he has suggested the following as upper bounds of reasonable use: * ~2500 Coordinates * ~100-200 Features (Since each geometry is rendered as a separate DOM object, dragging and the like get seriously slowed down the more features you have on the map.) If you've been using the vector layers in different browsers or with different limits... please let us know by either replying to this email... or just going right in and directly ediitng the wiki itself. The more input we have the better we know our own limits. Thanks for anyone who can chime in --Erik From rdewit at users.sourceforge.net Fri Oct 26 00:13:21 2007 From: rdewit at users.sourceforge.net (Roald de Wit) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Empty geometries causing errors in IE Message-ID: <1193372001.25343.12.camel@bender> Hi list, Creating an empty geometry is fine in Firefox (SVG), but fails in IE (VML). It goes wrong in VML.js in setNodeDimension. This method tries to create a bbox, then tests for its existence, but not if there is something in there. I'd like to start a discussion on whether OL should allow for empty geometries or not. The least that should come out of the discussion is a solution where the user doesn't get a javascript error for trying. See ticket #1111 for a simple 'fix'. Chris, Tim, anyone else? -- Roald de Wit Software Engineer roald.dewit@lisasoft.com Commercial Support for Open Source GIS Software http://lisasoft.com/LISAsoft/SupportedProducts/ From eider at fh-bingen.de Fri Oct 26 01:54:12 2007 From: eider at fh-bingen.de (Dipl. Inf. Carsten Eider) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] [OpenLayers-Users] OL Vector Layer: What are the Limitations? In-Reply-To: <6ae3fb590710251716k559d42a1nb24993ec75bd7a28@mail.gmail.com> References: <6ae3fb590710251716k559d42a1nb24993ec75bd7a28@mail.gmail.com> Message-ID: <47218104.5080807@fh-bingen.de> Skipped content of type multipart/alternative-------------- next part -------------- A non-text attachment was scrubbed... Name: eider.vcf Type: text/x-vcard Size: 330 bytes Desc: not available Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071026/045cd298/eider.vcf From jachym.cepicky at gmail.com Fri Oct 26 04:37:23 2007 From: jachym.cepicky at gmail.com (Jachym Cepicky) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] alt or title text on Layer.Marker In-Reply-To: References: <1193316677.6482.12.camel@kocour> Message-ID: <1193387843.7094.6.camel@kocour> Hi, not so bad, the labels are working for me (+1 for adding it to trunk), but I would need something, which would appear on mouse over I'll try to update your Labels.js, if you would not mind cheers Jachym Ezequiel Gonzalez Rial p??e v ?t 25. 10. 2007 v 10:03 -0300: > Jachym, > I've developed something arround that (Ticket #751) wich is not yet on > the trunk, it's a label that apperars next to the marker (by css you > can set it above, below or wherever you want), as it inheritance from > marker you don't loose anything you have already coded. > > -- > Best Regards, > Ezequiel Gonzalez Rial (gonrial@gmail.com) > > 2007/10/25, Jachym Cepicky : > > Hi, > > > > maybe, it would be good to have option for OpenLayers.Layer.Marker, > > which would enable displaying alt text of icons on mouse over. The alt > > attribute could be e.g. Title of OL.Layer.Text > > > > Is there such possibility, which would display alt or other attribute, > > while moving with mouse over icon? > > > > thanks > > > > jachym > > > > -- > > Jachym Cepicky > > e-mail: jachym.cepicky@gmail.com > > URL: http://les-ejk.cz > > GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub > > > > > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev -- Jachym Cepicky e-mail: jachym.cepicky@gmail.com URL: http://les-ejk.cz GPG: http://www.les-ejk.cz/pgp/jachym_cepicky-gpg.pub -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: Toto je =?UTF-8?Q?digit=C3=A1ln=C4=9B?= =?ISO-8859-1?Q?_podepsan=E1?= =?UTF-8?Q?_=C4=8D=C3=A1st?= =?ISO-8859-1?Q?_zpr=E1vy?= Url : http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071026/1f0296dd/attachment.bin From toncho11 at abv.bg Fri Oct 26 09:34:26 2007 From: toncho11 at abv.bg (__ __) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] colors a specific country Message-ID: <590357296.935721193405666921.JavaMail.nobody@mail51.abv.bg> Hi, I need a solution which colors a specific country with specific color. Any suggestion will be greatly appreciated!!! Cheers, Anton From val.cartei at gmail.com Fri Oct 26 10:26:54 2007 From: val.cartei at gmail.com (Val Cartei) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] How to make overview map resizable? Message-ID: Hi, I am trying to make the overview map resizable. I suppose this can be done by setting the scale of the overview map. Is there a method to update it dynamically? -- Best Regards, Valentina Cartei UK Tel. +44 (0)796 6882820 -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071026/67caf1ad/attachment.html From tschaub at openplans.org Fri Oct 26 11:28:11 2007 From: tschaub at openplans.org (Tim Schaub) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] deactivating other handlers In-Reply-To: References: Message-ID: <4722078B.4010207@openplans.org> Hey- Bart van den Eijnden (OSGIS) wrote: > Hi devs, > > I've written a FeatureInfo Handler, which only uses click. However, the > DragPan handler for instance is still active. > Not sure exactly what your FeatureInfo handler does, but it feels like it might be more of a control than a handler. Perhaps it could be a control that used the Feature handler (which can be configured to respond to a click)? Or maybe it does something like GetFeatureInfo (from WMS)? In this case, it should almost certainly be a control (because it doesn't represent a complex sequence of browser events). > What is the best way to deactivate these other handlers? Handlers are designed to abstract out sequences of browser events. They are typically sent to controls to modify the behavior of a control. When you activate a control, the appropriate handlers are activated. Deactivate, and the handlers are deactivated. We want there to be many controls in a given mapping app. Perhaps many will be active at one time. In this case, any handlers associated with these controls need to play well together. The order of activation is important. Controls (and their handlers) that are activated later get priority in event handling. > Is there a way to make a Handler exclusive, or to remove all other handlers > from the map? It wouldn't make sense necessarily to silence (or plug the ears of maybe) all handlers on the map. This would be akin to unregistering all event listeners just because you want to register a new one. You're better off having your application control the order of control (and hence handler) activation. This is a somewhat tedious task. Because of this, we have a control panel class. If you add a tool type control to a control panel, when it is activated (and it's handler is activated), it will deactivate all other tool type controls (and their handlers) on the same panel. I think this structure gives enough to work with in most situations. Perhaps I've totally missed what you're trying to do though. If you give more detail, I can try again. Tim > Best regards, > Bart > > -- > Bart van den Eijnden > OSGIS, Open Source GIS > http://www.osgis.nl > > > > > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > !DSPAM:4033,47205aa1282033327367457! > From barbara.philippot at camptocamp.com Fri Oct 26 11:32:51 2007 From: barbara.philippot at camptocamp.com (Barbara Philippot) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Developement of an external vector symbol In-Reply-To: <471B2D86.3020705@gmail.com> References: <4718CAC8.7020705@camptocamp.com> <471B2D86.3020705@gmail.com> Message-ID: <472208A3.7050806@camptocamp.com> Thanks for your advice Cameron! I researched available tools to transform SVG into VML. I found that only one project was usable without too many modifications : Mark Finkle's opensource project : Implementation of simple static SVG rendering in Internet Explorer using VML, which is under the terms of MIT license. http://starkravingfinkle.org/blog/2006/03/svg-in-ie/ After a certain number of modifications (symbol and viewbox management, minor corrections), I was able to transform SVG symbols into VML groups. Some problems are still there but most of them are well transformed. So, I inserted the needed functions into Openlayers codebase. For the moment, they are independant functions put into VML.js (one main function and 4 linked functions). We need to put these functions elsewhere : methods still in VML.js, in Util.js?, or integrate them into a new class and a new file? Since VML shapetype elements can't be as complex as SVG symbol elements, I transformed SVG symbols into VML groups with a visibility put to hidden. The specified group matching the point to draw is then cloned, made visible and put at the right place. So when creating a new vector layer, we can now specify an option with the URL of a single SVG symbols library which will be loaded in the rendererRoot node. Symbols of this library can be used on Firefox (directly in SVG) or on IE (transformed on the fly in Javascript into VML). It slows a little the browser down but loading time is satisfactory. For the moment, we have used this catalogue : http://wiki.openstreetmap.org/index.php/Osmarender/Symbols but people can add their own symbols or symbols from the Open Clip Art library provided that SVG symbols are put inside a element with the correct viewBox attribute associated. See : http://dev.openlayers.org/sandbox/camptocamp/vectorSymbols/examples/vector-symbols.html Regards, Barbara Cameron Shorter a ?crit : > Barbara, > This sounds like a great initiative and I'd be surprised if there > isn't already a number of projects which transform between SVG and VML. > I know that there has been work converting GML to SVG, and hence there > shouldn't be much work transforming GeoRSS to SVG. > > So I suggest: > 1. Before you do any development, research available tools that you > can use. > 2. Create one vector library (either SVG or VML) and use a translation > tool to automatically convert from one to the other. > 3. Precompile the alternative library. Ie, don't convert SVG to VML on > the fly in Javascript. It will slow the browser down. > 3. I suspect that SVG would be a better format to support than VML. I > believe SVG is used by a number of other projects, eg the Linux Icon > libraries. > 4. Search for existing icon libraries and consider extending the > existing icon libraries rather than creating a new project. Maybe make > a geographic version of the library. > 5. Think bigger than just OpenLayers. You start the free icon library > used by all Open Source GIS projects. It might be a candidate project > for: http://wiki.osgeo.org/index.php/OSGeo_Labs > (Although you may want to seed your project within Openlayers) > > > Barbara Philippot wrote: >> Ian Mayo a ?crit : >> >>> Barbara, >>> I'm afraid I can't offer an opinion on your languages challenge. It >>> would be great if there was a pre-existing file-format for vector >>> symbols, then we just specify filenames rather than hard-code the >>> vectored shapes. >>> >> Yes, this is why we are trying to do. But people would have to specify >> two filenames (one in SVG, the other in VML) to replace the pre-existing >> symbol catalogue files and it's not still the ideal solution but perhaps >> it's the only one. >> >>> I'm currenlty working on provision of per-entry icons (images) for >>> georss feeds. [1]. I've made good progress, but may have to pause my >>> efforts for the next 7 days, see my demo at: [2]. Had you considered >>> using the symbols on a per-entry fashion in a georss feed? I suspect >>> it's fortunate I've got a self-induced development pause: it would >>> seem to be worth offering per-entry symbols as well as icons. >>> >>> Appreciate your thoughts/opinions. >>> >> Well, I hadn't considered using symbols in a georss feed and I don't >> know well Georss format. For the moment, symbols are associated with >> points geometry with a style.use="symbol" in the same way as images were >> associated with style.externalGraphic. So, I haven't looked in detail to >> the way you attach icons to GeoRss nodes but it seems that you would be >> able to use SVG and VML symbols as well as images. >> >> I hope to make progress before your development pause ends :-) >> >>> Ian. >>> >>> 1. http://trac.openlayers.org/ticket/1073 >>> 2. >>> http://dev.openlayers.org/sandbox/ianmayo/branches/rss/openlayers/examples/georss.html >>> >>> >>> >> >> _______________________________________________ >> Dev mailing list >> Dev@openlayers.org >> http://openlayers.org/mailman/listinfo/dev >> >> > > From tschaub at openplans.org Fri Oct 26 11:52:50 2007 From: tschaub at openplans.org (Tim Schaub) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] Empty geometries causing errors in IE In-Reply-To: <1193372001.25343.12.camel@bender> References: <1193372001.25343.12.camel@bender> Message-ID: <47220D52.6020904@openplans.org> Hey- Roald de Wit wrote: > Hi list, > > Creating an empty geometry is fine in Firefox (SVG), but fails in IE > (VML). > It goes wrong in VML.js in setNodeDimension. This method tries to create > a bbox, then tests for its existence, but not if there is something in > there. > > I'd like to start a discussion on whether OL should allow for empty > geometries or not. The least that should come out of the discussion is a > solution where the user doesn't get a javascript error for trying. > I'm entirely in favor of this. I don't think the renderer is the right place to bomb (or hey maybe even throw an exception!) on an empty geometry. I think it makes sense to have our format classes throw exceptions on empty geometries. They could even have an option to allow empty geometry. In any case, it should be up to the application designer to decide what to do with empty geometry. As library developers, I don't think we should impose our beliefs about empty geometries on all others. Breaking in only one browser, and only when rendering, is a bit late and a bit inconsistent. > See ticket #1111 for a simple 'fix'. See also http://trac.openlayers.org/ticket/968 I'd like to see some tests for the renderers really. I haven't looked closely at your patch, but it would be nice to make sure we don't slip once we decide to make a change (like make browser behavior consistent). Tim > > Chris, Tim, anyone else? > From tschaub at openplans.org Fri Oct 26 11:56:51 2007 From: tschaub at openplans.org (Tim Schaub) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] colors a specific country In-Reply-To: <590357296.935721193405666921.JavaMail.nobody@mail51.abv.bg> References: <590357296.935721193405666921.JavaMail.nobody@mail51.abv.bg> Message-ID: <47220E43.9070906@openplans.org> __ __ wrote: > Hi, > > I need a solution which colors a specific country with specific color. > Any suggestion will be greatly appreciated!!! Red. > > Cheers, > Anton > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > !DSPAM:4033,4721ecee301671628642973! > From tschaub at openplans.org Fri Oct 26 12:44:01 2007 From: tschaub at openplans.org (Tim Schaub) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] colors a specific country In-Reply-To: <47220E43.9070906@openplans.org> References: <590357296.935721193405666921.JavaMail.nobody@mail51.abv.bg> <47220E43.9070906@openplans.org> Message-ID: <47221951.1050905@openplans.org> Excuse me. I meant to ask for more detail. The correct solution depends on where your data resides and what renders it. Do you have the geometries representing countries client side (in the browser)? If you are using a WMS that you are in control of, the solution will be different than if you are requesting data from a server over which you have no control. Anyway, more detail would help. Tim Tim Schaub wrote: > __ __ wrote: >> Hi, >> >> I need a solution which colors a specific country with specific color. >> Any suggestion will be greatly appreciated!!! > > Red. > >> Cheers, >> Anton >> _______________________________________________ >> Dev mailing list >> Dev@openlayers.org >> http://openlayers.org/mailman/listinfo/dev >> >> >> > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > !DSPAM:4033,47220e8b65681030819293! > From cameron.shorter at gmail.com Fri Oct 26 16:19:56 2007 From: cameron.shorter at gmail.com (Cameron Shorter) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] OL Vector Layer: What are the Limitations? In-Reply-To: <6ae3fb590710251716k559d42a1nb24993ec75bd7a28@mail.gmail.com> References: <6ae3fb590710251716k559d42a1nb24993ec75bd7a28@mail.gmail.com> Message-ID: <47224BEC.1020101@gmail.com> An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071027/74ac9d78/attachment.html From crschmidt at metacarta.com Sun Oct 28 15:09:31 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] [Incubator] OpenLayers graduation time In-Reply-To: <4724AA54.9020403@pobox.com> References: <1C0D69C7-2EFB-4516-AE9F-2134A2837A35@gmail.com> <4724AA54.9020403@pobox.com> Message-ID: <20071028190931.GA8355@metacarta.com> On Sun, Oct 28, 2007 at 11:27:16AM -0400, Frank Warmerdam wrote: > Howard Butler wrote: > >All, > > > >I would like to motion to graduate OpenLayers from incubation. > >OpenLayers has had the advantage of tailoring their governance, > >infrastructure, and licensing/provenance informed by the expectations of > >OSGeo incubation as the project has grown, and they have completed all > >required activities. > > Folks, > > I declare this motion passed with support from HowardB, DanielM, CameronS, > FrankW, PaulS, ArnulfC, and JodyG, and no dissent. > > The recommendation now goes to the board for final ratification this coming > Friday, but for our purposes I think we can consider it passed, and thus > consider a new incubation candidate for it's slot. But we will hold off on > formal news items and stuff till after Friday's final approval. > > Howard, thank you very much for your work as mentor for the OpenLayers > project, > and to the OpenLayers team - great work! You lapped several founding > projects > still in incubation. :-) Frank, Thank you for your, and OSGeo's support, and we look forward to our continued participation in the OSGeo community! Regards, -- Christopher Schmidt MetaCarta From bartvde at osgis.nl Mon Oct 29 08:10:39 2007 From: bartvde at osgis.nl (Bart van den Eijnden (OSGIS)) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] defaultIcon: calculateOffset Message-ID: <8fd5abc07c19bbb884892310e4b18471@145.50.39.11> Hi list, can anybody explain why the calculateOffset from the defaultIcon places the icon on top of the point, and not in the center of the point? OpenLayers.Marker.defaultIcon = function() { var url = OpenLayers.Util.getImagesLocation() + "marker.png"; var size = new OpenLayers.Size(21, 25); var calculateOffset = function(size) { return new OpenLayers.Pixel(-(size.w/2), -size.h); }; return new OpenLayers.Icon(url, size, null, calculateOffset); Best regards, Bart -- Bart van den Eijnden OSGIS, Open Source GIS http://www.osgis.nl From ahhughes at gmail.com Mon Oct 29 09:35:53 2007 From: ahhughes at gmail.com (Andrew Hughes) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] defaultIcon: calculateOffset In-Reply-To: <8fd5abc07c19bbb884892310e4b18471@145.50.39.11> References: <8fd5abc07c19bbb884892310e4b18471@145.50.39.11> Message-ID: <5f528cf40710290635n7be8ad18q779332f5df35d1d6@mail.gmail.com> IMHO, the intersection point between the two longest of the 4-sides is clearly the focal point..this is the bottom point and... hence the intersection point bottom is aligned over the position. Perhaps what you are really raising is that the default icon does not clearly focus its attention on the bottom of the icon. An arrow / teardrop, flag or triangle may be a clearer option? I would agree the default is not as good as it could be. On 10/29/07, Bart van den Eijnden (OSGIS) wrote: > > Hi list, > > can anybody explain why the calculateOffset from the defaultIcon places > the > icon on top of the point, and not in the center of the point? > > OpenLayers.Marker.defaultIcon = function() { > var url = OpenLayers.Util.getImagesLocation() + "marker.png"; > var size = new OpenLayers.Size(21, 25); > var calculateOffset = function(size) { > return new OpenLayers.Pixel(-(size.w/2), -size.h); > }; > > return new OpenLayers.Icon(url, size, null, calculateOffset); > > Best regards, > Bart > > -- > Bart van den Eijnden > OSGIS, Open Source GIS > http://www.osgis.nl > > > > > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.osgeo.org/pipermail/openlayers-dev/attachments/20071030/2e9dad11/attachment.html From bartvde at osgis.nl Mon Oct 29 11:13:00 2007 From: bartvde at osgis.nl (Bart van den Eijnden (OSGIS)) Date: Wed Sep 1 17:23:36 2010 Subject: [OpenLayers-Dev] prototype cleanWhitespace Message-ID: <29d2ec6b4c44c4234091c40607d410b3@145.50.39.11> Hi list, would it be an idea to include prototype's cleanWhitespace into OpenLayers Util? It's kind of handy when dealing with Mapserver's GML responses which have whitespace in there. Or do people use other techniques? Best regards, Bart -- Bart van den Eijnden OSGIS, Open Source GIS http://www.osgis.nl From erik.uzureau at metacarta.com Mon Oct 29 11:41:27 2007 From: erik.uzureau at metacarta.com (Erik Uzureau) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] defaultIcon: calculateOffset In-Reply-To: <8fd5abc07c19bbb884892310e4b18471@145.50.39.11> References: <8fd5abc07c19bbb884892310e4b18471@145.50.39.11> Message-ID: <6ae3fb590710290841h43b0f9f1id40ecc17a997104c@mail.gmail.com> Basically this is because the general markers (icons) that we've been working with are sort of like the gmaps one: the point of focus is on the bottom line center of the icon itself. e On 10/29/07, Bart van den Eijnden (OSGIS) wrote: > Hi list, > > can anybody explain why the calculateOffset from the defaultIcon places the > icon on top of the point, and not in the center of the point? > > OpenLayers.Marker.defaultIcon = function() { > var url = OpenLayers.Util.getImagesLocation() + "marker.png"; > var size = new OpenLayers.Size(21, 25); > var calculateOffset = function(size) { > return new OpenLayers.Pixel(-(size.w/2), -size.h); > }; > > return new OpenLayers.Icon(url, size, null, calculateOffset); > > Best regards, > Bart > > -- > Bart van den Eijnden > OSGIS, Open Source GIS > http://www.osgis.nl > > > > > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > From bartvde at osgis.nl Mon Oct 29 11:47:30 2007 From: bartvde at osgis.nl (Bart van den Eijnden (OSGIS)) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] defaultIcon: calculateOffset Message-ID: <2be719a94da38b12d88ef5ac372832fb@145.50.39.11> But with this particular icon as a user I would expect the focal point to be in the dot which is approximately in the center of the icon. But I get the logic behind this now more clearly at least, thanks for the explanation (also Andrew ofcourse). Best regards, Bart -- Bart van den Eijnden OSGIS, Open Source GIS http://www.osgis.nl --------- Oorspronkelijk bericht -------- Van: Erik Uzureau Naar: Bart van den Eijnden OSGIS Cc: dev@openlayers.org Onderwerp: Re: [OpenLayers-Dev] defaultIcon: calculateOffset Datum: 29/10/07 13:42 > Basically this is because the general markers (icons) that we've been > working with are sort of like the gmaps one: the point of focus is on > the bottom line center of the icon itself. > > e > > On 10/29/07, Bart van den Eijnden (OSGIS) <bartvde@osgis.nl> wrote: > > Hi list, > > > > can anybody explain why the calculateOffset from the defaultIcon places the > > icon on top of the point, and not in the center of the point? > > > > OpenLayers.Marker.defaultIcon = function() { > > var url = OpenLayers.Util.getImagesLocation() + "marker.png"; > > var size = new OpenLayers.Size(21, 25); > > var calculateOffset = function(size) { > > return new OpenLayers.Pixel(-(size.w/2), -size.h); > > }; > > > > return new OpenLayers.Icon(url, size, null, calculateOffset); > > > > Best regards, > > Bart > > > > -- > > Bart van den Eijnden > > OSGIS, Open Source GIS > > http://www.osgis.nl > > > > > > > > > > > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev > > > > From eric.c2c at gmail.com Mon Oct 29 15:07:46 2007 From: eric.c2c at gmail.com (Eric Lemoine) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish Message-ID: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> Hi OpenLayers dev! The Camptocamp team will soon be releasing the first version of MapFish (previously known as CartoWeb 4). Before the official (user) release, we want to make MapFish available to developers, and in particular to the OpenLayers dev community. Keep in mind that this isn't a release, there's still things to be done, including documentation and bug fixing. MapFish aims to be an easy-to-use and extensible web GIS application. MapFish is composed of two parts: MapFish Client and MapFish Server. MapFish Client is based on OpenLayers. For MapFish Client we've decided to move away from Dojo and use Ext [1] instead as the GUI framework. We're quite pleased with this change, Ext gives us satistaction in many respects: features, look and feel, documentation. Our move to Ext isn't complete yet. MapFish Server currently provides Python code based on gispython.org [2] Shapely and GeoJSON. The MapFishSample code base provides examples on how to use MapFish Server's Python code. MapFishSample uses the Pylons framework [3], with SQLAlchemy [4] as the ORM. With MapFish Server, we plan to build a a complete webmapping framework based on Pylons. Note also that MapFish Server aims to be RESTful and as compatible as possible with FeatureServer. You can visit to know more about MapFish, check out our demos, access the download page, register to the trac, register to the mailing lists, etc. We're looking forward to your feedback, [1] [2] [3] [4] -- The MapFish dev team From tschaub at openplans.org Mon Oct 29 17:30:44 2007 From: tschaub at openplans.org (Tim Schaub) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> References: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> Message-ID: <47265104.5080708@openplans.org> Hey- This looks great! My only other comment is that the "Why MapFish instead of OpenLayers?" question is a bit odd. Seems sort of like asking "Why sandwich instead of bread?" Tim Eric Lemoine wrote: > Hi OpenLayers dev! > > The Camptocamp team will soon be releasing the first version of MapFish > (previously known as CartoWeb 4). Before the official (user) release, we want > to make MapFish available to developers, and in particular to the OpenLayers > dev community. > > Keep in mind that this isn't a release, there's still things to be done, > including documentation and bug fixing. > > MapFish aims to be an easy-to-use and extensible web GIS application. MapFish > is composed of two parts: MapFish Client and MapFish Server. MapFish Client is > based on OpenLayers. > > For MapFish Client we've decided to move away from Dojo and use Ext [1] instead > as the GUI framework. We're quite pleased with this change, Ext gives us > satistaction in many respects: features, look and feel, documentation. > Our move to Ext isn't complete yet. > > MapFish Server currently provides Python code based on gispython.org [2] > Shapely and GeoJSON. The MapFishSample code base provides examples on > how to use MapFish Server's Python code. MapFishSample uses the Pylons > framework [3], with SQLAlchemy [4] as the ORM. With MapFish Server, we > plan to build a a complete webmapping framework based on Pylons. Note > also that MapFish Server aims to be RESTful and as compatible as > possible with FeatureServer. > > You can visit to know more about MapFish, check out > our demos, access the download page, register to the trac, register to the > mailing lists, etc. > > We're looking forward to your feedback, > > [1] > [2] > [3] > [4] > > -- > The MapFish dev team > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > !DSPAM:4033,47262f97161502092453641! > From rdewit at users.sourceforge.net Mon Oct 29 20:00:34 2007 From: rdewit at users.sourceforge.net (Roald de Wit) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] Empty geometries causing errors in IE In-Reply-To: <47220D52.6020904@openplans.org> References: <1193372001.25343.12.camel@bender> <47220D52.6020904@openplans.org> Message-ID: <1193702434.25343.81.camel@bender> On Fri, 2007-10-26 at 09:52 -0600, Tim Schaub wrote: > I'm entirely in favor of this. I don't think the renderer is the right > place to bomb (or hey maybe even throw an exception!) on an empty geometry. > > I think it makes sense to have our format classes throw exceptions on > empty geometries. They could even have an option to allow empty > geometry. In any case, it should be up to the application designer to > decide what to do with empty geometry. As library developers, I don't > think we should impose our beliefs about empty geometries on all others. > > Breaking in only one browser, and only when rendering, is a bit late and > a bit inconsistent. > > > > See ticket #1111 for a simple 'fix'. > > See also http://trac.openlayers.org/ticket/968 > > I'd like to see some tests for the renderers really. I haven't looked > closely at your patch, but it would be nice to make sure we don't slip > once we decide to make a change (like make browser behavior consistent). That fix in the patch I mentioned is probably not the best solution (although it worked for us). A few places where I can imagine we could prevent the renderer from trying to render the empty geometry is: - Renderer/Elements.js: drawGeometryNode() - Renderer.js: drawFeature() - Format/*.js: parseGeometry() It would be nice if we can come up with a clean solution for this. Roald -- Roald de Wit Software Engineer roald.dewit@lisasoft.com Commercial Support for Open Source GIS Software http://lisasoft.com/LISAsoft/SupportedProducts/ From lancelot at inetnebr.com Mon Oct 29 21:51:31 2007 From: lancelot at inetnebr.com (Lance Dyas) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> References: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> Message-ID: <47268E23.9020108@inetnebr.com> While it took a while to load I do love the ExtJS interface its down right hot.hmmm liked the old name better than MapFish... (more dignified, guess somebody is shooting for fun instead ;-)) From cameron.shorter at gmail.com Mon Oct 29 22:12:43 2007 From: cameron.shorter at gmail.com (Cameron Shorter) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> References: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> Message-ID: <4726931B.8040007@gmail.com> Eric, Looks very interesting. Quick questions. I couldn't find a reference to a license on the home page. What will you be using? My first thoughts when I look between server and client is "Should standard protocols be used". Do you see use of OGC Standards to be appropriate here? Or are some of your interfaces appropriate to put forward as a standard? Eric Lemoine wrote: > Hi OpenLayers dev! > > The Camptocamp team will soon be releasing the first version of MapFish > (previously known as CartoWeb 4). Before the official (user) release, we want > to make MapFish available to developers, and in particular to the OpenLayers > dev community. > > Keep in mind that this isn't a release, there's still things to be done, > including documentation and bug fixing. > > MapFish aims to be an easy-to-use and extensible web GIS application. MapFish > is composed of two parts: MapFish Client and MapFish Server. MapFish Client is > based on OpenLayers. > > For MapFish Client we've decided to move away from Dojo and use Ext [1] instead > as the GUI framework. We're quite pleased with this change, Ext gives us > satistaction in many respects: features, look and feel, documentation. > Our move to Ext isn't complete yet. > > MapFish Server currently provides Python code based on gispython.org [2] > Shapely and GeoJSON. The MapFishSample code base provides examples on > how to use MapFish Server's Python code. MapFishSample uses the Pylons > framework [3], with SQLAlchemy [4] as the ORM. With MapFish Server, we > plan to build a a complete webmapping framework based on Pylons. Note > also that MapFish Server aims to be RESTful and as compatible as > possible with FeatureServer. > > You can visit to know more about MapFish, check out > our demos, access the download page, register to the trac, register to the > mailing lists, etc. > > We're looking forward to your feedback, > > [1] > [2] > [3] > [4] > > -- > The MapFish dev team > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > -- Cameron Shorter Geospatial Systems Architect Tel: +61 (0)2 8570 5050 Mob: +61 (0)419 142 254 Think Globally, Fix Locally Commercial Support for Geospatial Open Source Software http://www.lisasoft.com/LISAsoft/SupportedProducts.html From cameron.shorter at gmail.com Tue Oct 30 01:34:35 2007 From: cameron.shorter at gmail.com (Cameron Shorter) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] Empty geometries causing errors in IE In-Reply-To: <1193702434.25343.81.camel@bender> References: <1193372001.25343.12.camel@bender> <47220D52.6020904@openplans.org> <1193702434.25343.81.camel@bender> Message-ID: <4726C26B.5040504@gmail.com> Roald de Wit wrote: > On Fri, 2007-10-26 at 09:52 -0600, Tim Schaub wrote: > >> I'm entirely in favor of this. I don't think the renderer is the right >> place to bomb (or hey maybe even throw an exception!) on an empty geometry. >> >> I think it makes sense to have our format classes throw exceptions on >> empty geometries. They could even have an option to allow empty >> geometry. In any case, it should be up to the application designer to >> decide what to do with empty geometry. As library developers, I don't >> think we should impose our beliefs about empty geometries on all others. >> >> Breaking in only one browser, and only when rendering, is a bit late and >> a bit inconsistent. >> >> >> >>> See ticket #1111 for a simple 'fix'. >>> >> See also http://trac.openlayers.org/ticket/968 >> >> I'd like to see some tests for the renderers really. I haven't looked >> closely at your patch, but it would be nice to make sure we don't slip >> once we decide to make a change (like make browser behavior consistent). >> > > That fix in the patch I mentioned is probably not the best solution > (although it worked for us). > > A few places where I can imagine we could prevent the renderer from > trying to render the empty geometry is: > > - Renderer/Elements.js: drawGeometryNode() > - Renderer.js: drawFeature() > - Format/*.js: parseGeometry() > > It would be nice if we can come up with a clean solution for this. > > Roald > > Chris has noted on IRC that he doesn't think VML.setNodeDimension() is the right place to catch this error and after digging around I agree. The problem is that a point geometry is created which has an x,y point of (null, null). This eventually causes an exception in the VML renderer. I propose we fix in Geometry/Point.initialize(x,y) where we should throw an exception if x or y is null. Reason: 1. Exceptions are already raised by Format/KML.js if it needs to create a (null,null) point, so I believe we have a precedence for handling errors by raising exceptions. 2. Our error was caused because Format/GML.js allows the creation of (null,null) points, probably from an empty tag. 3. I don't expect to break any legacy code by forcing Geometry to have valid values because old code would have caused exceptions in VML.js as we discovered. -- Cameron Shorter Geospatial Systems Architect Tel: +61 (0)2 8570 5050 Mob: +61 (0)419 142 254 Think Globally, Fix Locally Commercial Support for Geospatial Open Source Software http://www.lisasoft.com/LISAsoft/SupportedProducts.html From bartvde at osgis.nl Tue Oct 30 03:37:55 2007 From: bartvde at osgis.nl (Bart van den Eijnden (OSGIS)) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish Message-ID: <219375735ea16c3c7546e6a7ac989ad8@145.50.39.11> Wrt the license, the source code says GPL: /* * Copyright (C) 2007 Camptocamp * * This file is part of MapFish * * MapFish is free software: you can redistribute it and/or modify * it under the terms of the GNU Lesser General Public License as published by * the Free Software Foundation, either version 3 of the License, or * (at your option) any later version. * * MapFish is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public License * along with MapFish. If not, see . */ Best regards, Bart -- Bart van den Eijnden OSGIS, Open Source GIS http://www.osgis.nl --------- Oorspronkelijk bericht -------- Van: Cameron Shorter Naar: Eric Lemoine Cc: dev@openlayers ML Onderwerp: Re: [OpenLayers-Dev] MapFish Datum: 30/10/07 00:13 > Eric, > Looks very interesting. > > Quick questions. > I couldn't find a reference to a license on the home page. What will you > be using? > > My first thoughts when I look between server and client is "Should > standard protocols be used". Do you see use of OGC Standards to be > appropriate here? > Or are some of your interfaces appropriate to put forward as a standard? > > Eric Lemoine wrote: > > Hi OpenLayers dev! > > > > The Camptocamp team will soon be releasing the first version of MapFish > > (previously known as CartoWeb 4). Before the official (user) release, we want > > to make MapFish available to developers, and in particular to the OpenLayers > > dev community. > > > > Keep in mind that this isn't a release, there's still things to be done, > > including documentation and bug fixing. > > > > MapFish aims to be an easy-to-use and extensible web GIS application. MapFish > > is composed of two parts: MapFish Client and MapFish Server. MapFish Client is > > based on OpenLayers. > > > > For MapFish Client we've decided to move away from Dojo and use Ext [1] instead > > as the GUI framework. We're quite pleased with this change, Ext gives us > > satistaction in many respects: features, look and feel, documentation. > > Our move to Ext isn't complete yet. > > > > MapFish Server currently provides Python code based on gispython.org [2] > > Shapely and GeoJSON. The MapFishSample code base provides examples on > > how to use MapFish Server's Python code. MapFishSample uses the Pylons > > framework [3], with SQLAlchemy [4] as the ORM. With MapFish Server, we > > plan to build a a complete webmapping framework based on Pylons. Note > > also that MapFish Server aims to be RESTful and as compatible as > > possible with FeatureServer. > > > > You can visit <http://www.mapfish.org> to know more about MapFish, check out > > our demos, access the download page, register to the trac, register to the > > mailing lists, etc. > > > > We're looking forward to your feedback, > > > > [1] <http://extjs.com/> > > [2] <http://gispython.org/> > > [3] <http://pylonshq.com/> > > [4] <http://www.sqlalchemy.org/> > > > > -- > > The MapFish dev team > > _______________________________________________ > > Dev mailing list > > Dev@openlayers.org > > http://openlayers.org/mailman/listinfo/dev > > > > > > > -- > Cameron Shorter > Geospatial Systems Architect > Tel: +61 (0)2 8570 5050 > Mob: +61 (0)419 142 254 > > Think Globally, Fix Locally > Commercial Support for Geospatial Open Source Software > http://www.lisasoft.com/LISAsoft/SupportedProducts.html > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev > > From eric.c2c at gmail.com Tue Oct 30 03:57:00 2007 From: eric.c2c at gmail.com (Eric Lemoine) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <219375735ea16c3c7546e6a7ac989ad8@145.50.39.11> References: <219375735ea16c3c7546e6a7ac989ad8@145.50.39.11> Message-ID: <5ec103de0710300057o5270405ei3392d8e6ff034d90@mail.gmail.com> On 10/30/07, Bart van den Eijnden (OSGIS) wrote: > Wrt the license, the source code says GPL: It's actually LGPLv3. -- Eric From eric.c2c at gmail.com Tue Oct 30 05:22:32 2007 From: eric.c2c at gmail.com (Eric Lemoine) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <47265104.5080708@openplans.org> References: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> <47265104.5080708@openplans.org> Message-ID: <5ec103de0710300222j635c2d70o97c801ec1e43bfab@mail.gmail.com> On 10/29/07, Tim Schaub wrote: > Hey- > > This looks great! > > My only other comment is that the "Why MapFish instead of OpenLayers?" > question is a bit odd. Seems sort of like asking "Why sandwich instead > of bread?" Hi Tim Indeed, that question did not make any sense. I removed that question from the wiki (and changed things here and there in the project description). Thanks for your feedback, -- Eric From roch.sebastien at gmail.com Tue Oct 30 10:19:26 2007 From: roch.sebastien at gmail.com (=?ISO-8859-1?Q?S=E9bastien_Roch?=) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] FixedZoomLevels Message-ID: Hi Dev, I had a problem with some Google Layers in my map : I had some layers with min and maxScale above a google layer (the baselayer). At first it worked well, but after having zoomed 3 times, the layers don't appear anymore (the google one is ok) If at this moment I eval in the console "map.getScale()", it returns "NaN" I found that the resolutions array of the google layer was not properly set : this is done in FixedZoomLevels.js, in the "initResolutions" function : if (this.RESOLUTIONS != null) { var resolutionsIndex = 0; this.resolutions = []; for(var i= this.minZoomLevel; i < this.numZoomLevels; i++) { this.resolutions[resolutionsIndex++] = this.RESOLUTIONS[i]; } } I changed this part with : if (this.RESOLUTIONS != null) { var resolutionsIndex = 0; this.resolutions = []; for(var i= this.minZoomLevel; i <= this.maxZoomLevel; i++) { // HERE this.resolutions[resolutionsIndex++] = this.RESOLUTIONS[i]; } } And this works for me. Can anyone tell me if this is right ? if not, I'll post my code with my layers, maybe I'm wrong with something. Thanks Sebastien From eric.c2c at gmail.com Tue Oct 30 12:27:49 2007 From: eric.c2c at gmail.com (Eric Lemoine) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <4726931B.8040007@gmail.com> References: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> <4726931B.8040007@gmail.com> Message-ID: <5ec103de0710300927m44dddb30m93db64d1fecdf7d6@mail.gmail.com> On 10/30/07, Cameron Shorter wrote: > Eric, > Looks very interesting. Hi Cameron > > Quick questions. > I couldn't find a reference to a license on the home page. What will you > be using? Already answered. > My first thoughts when I look between server and client is "Should > standard protocols be used". Do you see use of OGC Standards to be > appropriate here? On the client-side there are components, such as the layer widget, that don't rely on any service. Some others, such as geostat, do rely on services that serve features using GeoJSON. The services support filtering, and we're trying to conform to the parameters specified in OpenSearch Geo. The application developer can also add his own app-specific filter parameters, but the protocol becomes totally custom. As you've probably noted we could have used WFS and Filter Encoding for this. But our goal being to provide an extensible framework on the server side, where the application developer can provide his own code to respond to query of any complexity. I feel that achieving this same level of functionality with WFS and Filter Encoding would much more complex. I'm very interested to hear from you regarding all this. > Or are some of your interfaces appropriate to put forward as a standard? We'll be interested in pushing REST-based interfaces in the realm of feature services and filtering. I often read Sean Gillies' blog (CC:ed to this email) and he also seems to be interested in that area. Thanks, -- Eric From sgillies at frii.com Tue Oct 30 15:41:20 2007 From: sgillies at frii.com (Sean Gillies) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <5ec103de0710300927m44dddb30m93db64d1fecdf7d6@mail.gmail.com> References: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> <4726931B.8040007@gmail.com> <5ec103de0710300927m44dddb30m93db64d1fecdf7d6@mail.gmail.com> Message-ID: <472788E0.3000002@frii.com> Eric Lemoine wrote: > On 10/30/07, Cameron Shorter wrote: >> Eric, >> Looks very interesting. > > Hi Cameron > >> Quick questions. >> I couldn't find a reference to a license on the home page. What will you >> be using? > > Already answered. > >> My first thoughts when I look between server and client is "Should >> standard protocols be used". Do you see use of OGC Standards to be >> appropriate here? > > On the client-side there are components, such as the layer widget, > that don't rely on any service. Some others, such as geostat, do rely > on services that serve features using GeoJSON. The services support > filtering, and we're trying to conform to the parameters specified in > OpenSearch Geo. The application developer can also add his own > app-specific filter parameters, but the protocol becomes totally > custom. > > As you've probably noted we could have used WFS and Filter Encoding > for this. But our goal being to provide an extensible framework on the > server side, where the application developer can provide his own code > to respond to query of any complexity. I feel that achieving this same > level of functionality with WFS and Filter Encoding would much more > complex. > > I'm very interested to hear from you regarding all this. > >> Or are some of your interfaces appropriate to put forward as a standard? > > We'll be interested in pushing REST-based interfaces in the realm of > feature services and filtering. I often read Sean Gillies' blog (CC:ed > to this email) and he also seems to be interested in that area. > > Thanks, > > -- > Eric > What could be more standard than REST and HTTP? Cheers, Sean From pspencer at dmsolutions.ca Tue Oct 30 16:45:56 2007 From: pspencer at dmsolutions.ca (Paul Spencer) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] safari bug on window resize Message-ID: <33FA5E70-3F68-4C7A-A187-E9E6F8C57F9E@dmsolutions.ca> Devs, there seems to be a bug in Safari 3 (not sure about 2) with trunk where a window resize causes the tiles to disappear. Zak and I tracked this down to the fact that Safari does not trigger an onload event on the image when you set the .src of the image and it is not different from the current src. Our fix was this patch: Index: /ms4w/apps/openlayers/trunk/openlayers/lib/OpenLayers/Tile/ Image.js =================================================================== --- /ms4w/apps/openlayers/trunk/openlayers/lib/OpenLayers/Tile/ Image.js (revision 5069) +++ /ms4w/apps/openlayers/trunk/openlayers/lib/OpenLayers/Tile/ Image.js (working copy) @@ -119,6 +119,7 @@ OpenLayers.Util.modifyAlphaImageDiv(this.imgDiv, null, null, imageSize, this.url); } else { + this.imgDiv.src = ''; this.imgDiv.src = this.url; OpenLayers.Util.modifyDOMElement(this.imgDiv, null, null, imageSize) ; By assigning an empty string, this causes Safari to behave properly when you assign the real source. An alternate patch would be to set the imgDiv.src = '' in the clear function. An alternate patch would be to check if the urls are identical (not sure if checking == is sufficient or if the isEquivalent function should be used) and just show the imgDiv if they are the same: But all this leads me to question what is happening if we are setting the source to the same image? It appears that a window resize is causing the map to redraw as if the zoom level has changed. This causes all the tiles to get cleared. I don't understand why this is necessary. There is code in setCenter to just move the layer container if the zoom hasn't changed and this seems like a reasonable thing, so why is updateSize forcing the zoom to change? I am willing to put together a patch if I can get some feedback on what the right solution would be: 1. clear src just before setting it 2. clear src in the clear function when we hide the imgDiv 3. only set the src if the new url is different from the old url, otherwise just show the imgDiv again 4. change updateSize in Map to not force a zoom level change Cheers Paul +-----------------------------------------------------------------+ |Paul Spencer pspencer@dmsolutions.ca | +-----------------------------------------------------------------+ |Chief Technology Officer | |DM Solutions Group Inc http://www.dmsolutions.ca/ | +-----------------------------------------------------------------+ From crschmidt at metacarta.com Tue Oct 30 17:19:32 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] safari bug on window resize In-Reply-To: <33FA5E70-3F68-4C7A-A187-E9E6F8C57F9E@dmsolutions.ca> References: <33FA5E70-3F68-4C7A-A187-E9E6F8C57F9E@dmsolutions.ca> Message-ID: <20071030211932.GA20272@metacarta.com> On Tue, Oct 30, 2007 at 04:45:56PM -0400, Paul Spencer wrote: > Devs, > > there seems to be a bug in Safari 3 (not sure about 2) with trunk > where a window resize causes the tiles to disappear. Zak and I > tracked this down to the fact that Safari does not trigger an onload > event on the image when you set the .src of the image and it is not > different from the current src. Yes, I've discussed this on the list before, a couple weeks ago: http://trac.openlayers.org/ticket/758 > But all this leads me to question what is happening if we are setting > the source to the same image? > > It appears that a window resize is causing the map to redraw as if > the zoom level has changed. This causes all the tiles to get > cleared. I don't understand why this is necessary. There is code in > setCenter to just move the layer container if the zoom hasn't changed > and this seems like a reasonable thing, so why is updateSize forcing > the zoom to change? > > I am willing to put together a patch if I can get some feedback on > what the right solution would be: > > 1. clear src just before setting it > > 2. clear src in the clear function when we hide the imgDiv > > 3. only set the src if the new url is different from the old url, > otherwise just show the imgDiv again > > 4. change updateSize in Map to not force a zoom level change This won't fix the problem in its entirity -- base layer switching will still cause the problem. Regards, -- Christopher Schmidt MetaCarta From cameron.shorter at gmail.com Tue Oct 30 18:20:27 2007 From: cameron.shorter at gmail.com (Cameron Shorter) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <5ec103de0710300927m44dddb30m93db64d1fecdf7d6@mail.gmail.com> References: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> <4726931B.8040007@gmail.com> <5ec103de0710300927m44dddb30m93db64d1fecdf7d6@mail.gmail.com> Message-ID: <4727AE2B.2040506@gmail.com> Re Open Specs: Assuming specifications are of similar complexity to alternatives, a developer should pick specs as they offer an element of future proofing to an application. Unfortunately some of the OGC specs don't fall into the "similar complexity" category, as evidenced by the level of uptake things like the WFS spec. Encouragingly the OGC has shown signs of recognizing this with their recent "Mass Market" focus. So you are likely to find OGC receptive to suggestions to simple specs, as evidenced by OGC move toward Simple GML and GeoRSS. If you are looking into routing, I suggest you start by looking at the OGC's OpenLS spec. I am not familiar enough with it to make a call on whether a simpler protocol would be more appropriate. Eric Lemoine wrote: > On 10/30/07, Cameron Shorter wrote: > >> Eric, >> Looks very interesting. >> > > Hi Cameron > > >> Quick questions. >> I couldn't find a reference to a license on the home page. What will you >> be using? >> > > Already answered. > > >> My first thoughts when I look between server and client is "Should >> standard protocols be used". Do you see use of OGC Standards to be >> appropriate here? >> > > On the client-side there are components, such as the layer widget, > that don't rely on any service. Some others, such as geostat, do rely > on services that serve features using GeoJSON. The services support > filtering, and we're trying to conform to the parameters specified in > OpenSearch Geo. The application developer can also add his own > app-specific filter parameters, but the protocol becomes totally > custom. > > As you've probably noted we could have used WFS and Filter Encoding > for this. But our goal being to provide an extensible framework on the > server side, where the application developer can provide his own code > to respond to query of any complexity. I feel that achieving this same > level of functionality with WFS and Filter Encoding would much more > complex. > > I'm very interested to hear from you regarding all this. > > >> Or are some of your interfaces appropriate to put forward as a standard? >> > > We'll be interested in pushing REST-based interfaces in the realm of > feature services and filtering. I often read Sean Gillies' blog (CC:ed > to this email) and he also seems to be interested in that area. > > Thanks, > > -- > Eric > > -- Cameron Shorter Geospatial Systems Architect Tel: +61 (0)2 8570 5050 Mob: +61 (0)419 142 254 Think Globally, Fix Locally Commercial Support for Geospatial Open Source Software http://www.lisasoft.com/LISAsoft/SupportedProducts.html From cameron.shorter at gmail.com Tue Oct 30 18:24:47 2007 From: cameron.shorter at gmail.com (Cameron Shorter) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <472788E0.3000002@frii.com> References: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> <4726931B.8040007@gmail.com> <5ec103de0710300927m44dddb30m93db64d1fecdf7d6@mail.gmail.com> <472788E0.3000002@frii.com> Message-ID: <4727AF2F.3070002@gmail.com> Sean Gillies wrote: > > What could be more standard than REST and HTTP? > > Sean, nothing. I was referring to the OGC specs which sit on top of REST and HTTP. -- Cameron Shorter Geospatial Systems Architect Tel: +61 (0)2 8570 5050 Mob: +61 (0)419 142 254 Think Globally, Fix Locally Commercial Support for Geospatial Open Source Software http://www.lisasoft.com/LISAsoft/SupportedProducts.html From crschmidt at metacarta.com Tue Oct 30 20:13:04 2007 From: crschmidt at metacarta.com (Christopher Schmidt) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <4727AF2F.3070002@gmail.com> References: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> <4726931B.8040007@gmail.com> <5ec103de0710300927m44dddb30m93db64d1fecdf7d6@mail.gmail.com> <472788E0.3000002@frii.com> <4727AF2F.3070002@gmail.com> Message-ID: <20071031001304.GA15115@metacarta.com> On Wed, Oct 31, 2007 at 09:24:47AM +1100, Cameron Shorter wrote: > > Sean Gillies wrote: > > > > What could be more standard than REST and HTTP? > > > > > Sean, nothing. I was referring to the OGC specs which sit on top of REST > and HTTP. I haven't yet seen any OGC specs that 'sit on top of' REST or HTTP. OpenLS definitely doesn't apply, nor do WMS, WFS, or any of the other protocols that OpenLayers supports, as far as I'm aware. Which ones are you thinking about? Regards, -- Christopher Schmidt MetaCarta From pspencer at dmsolutions.ca Tue Oct 30 20:29:00 2007 From: pspencer at dmsolutions.ca (Paul Spencer) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] safari bug on window resize In-Reply-To: <20071030211932.GA20272@metacarta.com> References: <33FA5E70-3F68-4C7A-A187-E9E6F8C57F9E@dmsolutions.ca> <20071030211932.GA20272@metacarta.com> Message-ID: <9214765B-7959-4C92-B019-3541F4509F42@dmsolutions.ca> Chris, On 30-Oct-07, at 5:19 PM, Christopher Schmidt wrote: > On Tue, Oct 30, 2007 at 04:45:56PM -0400, Paul Spencer wrote: >> Devs, >> >> there seems to be a bug in Safari 3 (not sure about 2) with trunk >> where a window resize causes the tiles to disappear. Zak and I >> tracked this down to the fact that Safari does not trigger an onload >> event on the image when you set the .src of the image and it is not >> different from the current src. > > Yes, I've discussed this on the list before, a couple weeks ago: > > http://trac.openlayers.org/ticket/758 I've added a note to the ticket to the effect that the bug also occurs on resize since the ticket only talks about baseLayer switching > >> >> I am willing to put together a patch if I can get some feedback on >> what the right solution would be: >> >> 1. clear src just before setting it >> >> 2. clear src in the clear function when we hide the imgDiv >> >> 3. only set the src if the new url is different from the old url, >> otherwise just show the imgDiv again >> >> 4. change updateSize in Map to not force a zoom level change > > This won't fix the problem in its entirity -- base layer switching > will > still cause the problem. Right. The patch in the above ticket implements option 2 above. Assuming that the patch is eventually reviewed and committed to trunk, this will fix the problem that I am observing. However, it doesn't answer the broader question of why we force a redraw of the layers on a window resize. On Safari, the resize event is emitted continuously. The cost of a redraw is high enough that the window really chunks along during a resize (at least on my machine). If the reason for doing a redraw of the layers is to catch the case where the window grows and the grid has to expand, then I think there could be a lighter weight test that is done on resize before forcing a redraw. Is there another reason for the current behaviour? Cheers Paul +-----------------------------------------------------------------+ |Paul Spencer pspencer@dmsolutions.ca | +-----------------------------------------------------------------+ |Chief Technology Officer | |DM Solutions Group Inc http://www.dmsolutions.ca/ | +-----------------------------------------------------------------+ From cameron.shorter at gmail.com Tue Oct 30 21:45:12 2007 From: cameron.shorter at gmail.com (Cameron Shorter) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <20071031001304.GA15115@metacarta.com> References: <5ec103de0710291207k295f4fc1oe6efad3900d02f2f@mail.gmail.com> <4726931B.8040007@gmail.com> <5ec103de0710300927m44dddb30m93db64d1fecdf7d6@mail.gmail.com> <472788E0.3000002@frii.com> <4727AF2F.3070002@gmail.com> <20071031001304.GA15115@metacarta.com> Message-ID: <4727DE28.3020109@gmail.com> Christopher Schmidt wrote: > On Wed, Oct 31, 2007 at 09:24:47AM +1100, Cameron Shorter wrote: > >> Sean Gillies wrote: >> >>> >>> What could be more standard than REST and HTTP? >>> >>> >>> >> Sean, nothing. I was referring to the OGC specs which sit on top of REST >> and HTTP. >> > > I haven't yet seen any OGC specs that 'sit on top of' REST or HTTP. > OpenLS definitely doesn't apply, nor do WMS, WFS, or any of the other > protocols that OpenLayers supports, as far as I'm aware. > > Which ones are you thinking about? > > Regards, > I stand corrected. -- Cameron Shorter Geospatial Systems Architect Tel: +61 (0)2 8570 5050 Mob: +61 (0)419 142 254 Think Globally, Fix Locally Commercial Support for Geospatial Open Source Software http://www.lisasoft.com/LISAsoft/SupportedProducts.html From pat at cappelaere.com Wed Oct 31 08:18:56 2007 From: pat at cappelaere.com (Pat Cappelaere) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] MapFish In-Reply-To: <4727AF2F.3070002@gmail.com> Message-ID: I would really disagree with that statement. A REST-RPC style interface is not RESTful and appropriate for the Mass Market. So the current OGC specs are not quite there yet. But the OGC has recognized that and is working on this for OWS-5. There is still a challenge on the table for a RESTful WMS (which has been picked up by Arnulf) I am currently working hard at pushing those issues to the surface. We are having good success with the Workflow Chaining Service and Pubusb. I need to start the documentation process. Our version of the RESTful WFS is falling a little behind because of lack of funding... Hopefully, the SoCal fires are bringing some attention to the issues... http://blog.geobliki.com Pat. > From: Cameron Shorter > Organization: LISAsoft > Date: Wed, 31 Oct 2007 09:24:47 +1100 > To: Sean Gillies > Cc: "dev@openlayers ML" > Subject: Re: [OpenLayers-Dev] MapFish > > > Sean Gillies wrote: >> >> What could be more standard than REST and HTTP? >> >> > Sean, nothing. I was referring to the OGC specs which sit on top of REST > and HTTP. > > -- > Cameron Shorter > Geospatial Systems Architect > Tel: +61 (0)2 8570 5050 > Mob: +61 (0)419 142 254 > > Think Globally, Fix Locally > Commercial Support for Geospatial Open Source Software > http://www.lisasoft.com/LISAsoft/SupportedProducts.html > > > _______________________________________________ > Dev mailing list > Dev@openlayers.org > http://openlayers.org/mailman/listinfo/dev From richard.thurbin at cloudamber.com Wed Oct 31 20:51:34 2007 From: richard.thurbin at cloudamber.com (Richard Thurbin) Date: Wed Sep 1 17:23:37 2010 Subject: [OpenLayers-Dev] Sliverlight Message-ID: <47292316.3080808@cloudamber.com> Following on from the flash based mapping client posted a few days ago, I stumbled across this Microsoft Silverlight based version: http://silverlight.net/samples/1.1/virtualearthviewer/ I very much like the fade effects, smooth pan & zoom and speed wise it seems bang on. The only thing it needs navigation wise is a "window" tool to zoom in and out very quickly. Thoughts? -- Richard Thurbin www.cloudamber.com +44 7917 704 145