From Ross.Searle at dnr.qld.gov.au Wed Dec 1 00:15:58 1999 From: Ross.Searle at dnr.qld.gov.au (Searle Ross) Date: Wed, 1 Dec 1999 18:15:58 +1000 Subject: Query map Message-ID: Has anyone worked out how to get the querymap stuff working yet. I am using V3.3 (downloaded today) I have tried to copy the info at http://arachnid.dnr.state.mn.us/ms_test/itemquery/test.html but I must be missing something. My map file is attached below. When I query the map a reference to an image is substituted into the returned template but no image is actually created. I have no idea how this is meant to be setup so any hints would be appreciated On another topic, has anyone had any luck with relational joins. I can get a join to work no worries (as per attached map file - the relevant section is commented out) but it is only returning the first row of the joined table (ie one to one) even though the type is set to multiple. Thanks Ross # # Start of map file # NAME NRmeta STATUS ON SIZE 600 600 SHADESET /data4/WWW/mapserver/symbols/shade.sym MARKERSET /data4/WWW/mapserver/symbols/marker.sym LINESET /data4/WWW/mapserver/symbols/line.sym EXTENT 224000 7012000 445400 7310000 UNITS METERS SHAPEPATH "/data4/WWW/develop/NRmeta/gis" IMAGECOLOR 255 255 255 # # Start of web interface definition # WEB IMAGEPATH /data4/WWW/tmp/ IMAGEURL /tmp/ TEMPLATE nrmain.html #LOG test.log footer "sel.html" END QUERYMAP status on STYLE HILITE COLOR 255 255 0 END # # Start of layer definitions # LAYER NAME test2 TYPE POLYGON STATUS DEFAULT DATA test2 CLASS SYMBOL 0 OUTLINECOLOR 0 0 0 COLOR 248 248 95 END HEADER "qry_test_header.html" FOOTER "qry_test_footer.html" QUERY TEMPLATE qry_test_main.html #TEMPLATE jointest2.html # JOIN # type multiple # from ID # name test1 # table jointable.dbf # TEMPLATE jointest.html # to TEST # END end # END END END END Ross Searle Natural Resource Information Management Officer Department of Natural Resources PO Box 1143 Bundaberg QLD 4670. Australia Ph : 0741 537 888 Fax : 0741 537 823 From pgallen at randomlogic.com Wed Dec 1 01:20:18 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Wed, 01 Dec 1999 01:20:18 -0800 Subject: MapServer Running on Alpha Message-ID: <3844E852.FD1370BC@randomlogic.com> OK, I've finally got the MapServer running on my Alpha. It took a little doing as follows: 1. I had to fix a few places in the code where integers were cast to pointers and vice versa. Alpha pointers are 64 bit, integers are 32 bit, whereas on x86 they are both 32 bit. This turned out to be trivial (fortunately there weren't too many cases of this :) 2. I had to fix the damn egcs compiler. It kept reporting type mismatch errors with memcpy(), strcpy(), and other such functions. I removed then reloaded the compiler and all libraries. (one of these days I'll D/L gcc 2.95.1 and try it out :) 3. Since the Alpha has only a 4GB HDD, and the map data is well over 4GB uncompressed, I had to create symlinks and mount NFS drives from the K6 to access the map data. 4. Being new to Apache (and any web server for that matter), I had to figure out how to make Apache respond to an alternate port (instead of the usual port 80). The Alpha now responds to 24.5.14.144:3000, with the K6 still at port 80, though I'm still not sure if my firewall is allowing external access to this port. Works great internally though, and configuring Apache was actually quite simple (considering the httpd.conf file has a comment telling exactly how to do this, and Comanche has a field for it :) The Alpha is quite a bit faster at making the maps than the K6 is (duh) with the bottleneck now being the slow 10Mbps ethernet. I really need to upgrade to either fiber or 100Mbps. PGA From pgallen at randomlogic.com Wed Dec 1 06:00:42 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Wed, 01 Dec 1999 06:00:42 -0800 Subject: Labeling Problems Message-ID: <38452A0A.F31A0372@randomlogic.com> I seem to be having problems with labels, mainly street labels. I am having two problems, but only on the Alpha system. 1. Street labels only appear when zoomed way in on a map, about 5-10 city blocks. Anything else and the CGI crashes with a "premature end of script headers" in the Apache error log. 2. Turning on any other labels - cities, landmarks, etc. - also causes a crash. Everything seems to work fine on the K6 however. My make files for the two systems are identical, and so is the directory tree as well as the .map file. Any ideas of where to look for what's causing this problem? PGA From steve.lime at dnr.state.mn.us Wed Dec 1 10:36:18 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Wed, 01 Dec 1999 12:36:18 -0600 Subject: Query map Message-ID: Hi Ross: QueryMaps are actually the last detail I'm hammering out with 3.3 before the beta. They are fully implemented for ITEMQUERY, ITEMNQUERY and QUERY mode. I've attached the test suite files for the itemquery example. Hopefully that will help. I'll fill out that section of the mapfile docs on the main website tonite. In general QueryMaps are mean't as a way to visualize a set of query results. Here are the basic params: QUERYMAP STATUS ON|OFF - the usual STYLE SELECTED|HILITE|NORMAL - style of map, see below COLOR r g b - hilite color END Here's what the styles mean: SELECTED - draws layers normally but only draws selected features in queried layers HILITE - same as SELECTED but selected features are drawn with the hilite color NORMAL - draws all layers normally. this would be used for generating a map based on the extents of the selected features. To use the extent(s) of the selected features to create the map you would set mapext=shapes. This causes a normal query with the query map generated for the area covering only the selected shapes. I should also note that all of the query modes also/will support a map only mode that will allow you to generate maps based solely on, say item values. These modes are ITEMQUERYMAP, QUERYMAP, ITEMNQUERYMAP and so on... They are just like mode=MAP and return just an image. I believe case 3 or 4 in the included file shows this functionality. Steve BTW I'm looking into the JOIN problem... Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 >>> Searle Ross 12/01/99 02:15AM >>> Has anyone worked out how to get the querymap stuff working yet. I am using V3.3 (downloaded today) I have tried to copy the info at http://arachnid.dnr.state.mn.us/ms_test/itemquery/test.html but I must be missing something. My map file is attached below. When I query the map a reference to an image is substituted into the returned template but no image is actually created. I have no idea how this is meant to be setup so any hints would be appreciated On another topic, has anyone had any luck with relational joins. I can get a join to work no worries (as per attached map file - the relevant section is commented out) but it is only returning the first row of the joined table (ie one to one) even though the type is set to multiple. Thanks Ross # # Start of map file # NAME NRmeta STATUS ON SIZE 600 600 SHADESET /data4/WWW/mapserver/symbols/shade.sym MARKERSET /data4/WWW/mapserver/symbols/marker.sym LINESET /data4/WWW/mapserver/symbols/line.sym EXTENT 224000 7012000 445400 7310000 UNITS METERS SHAPEPATH "/data4/WWW/develop/NRmeta/gis" IMAGECOLOR 255 255 255 # # Start of web interface definition # WEB IMAGEPATH /data4/WWW/tmp/ IMAGEURL /tmp/ TEMPLATE nrmain.html #LOG test.log footer "sel.html" END QUERYMAP status on STYLE HILITE COLOR 255 255 0 END # # Start of layer definitions # LAYER NAME test2 TYPE POLYGON STATUS DEFAULT DATA test2 CLASS SYMBOL 0 OUTLINECOLOR 0 0 0 COLOR 248 248 95 END HEADER "qry_test_header.html" FOOTER "qry_test_footer.html" QUERY TEMPLATE qry_test_main.html #TEMPLATE jointest2.html # JOIN # type multiple # from ID # name test1 # table jointable.dbf # TEMPLATE jointest.html # to TEST # END end # END END END END Ross Searle Natural Resource Information Management Officer Department of Natural Resources PO Box 1143 Bundaberg QLD 4670. Australia Ph : 0741 537 888 Fax : 0741 537 823 -------------- next part -------------- A non-text attachment was scrubbed... Name: itemquery.tar.gz Type: application/octet-stream Size: 2752 bytes Desc: not available URL: From pgallen at randomlogic.com Thu Dec 2 03:42:45 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Thu, 02 Dec 1999 03:42:45 -0800 Subject: MapServer and C math functions Message-ID: <38465B35.E68D5501@randomlogic.com> This is actually a question for Steve, but I thought both the question and answer might be informative for all users. I'd like to know, if you can tell me or at least give me an idea, what math functions mapserv and all associated utilities use. I've D/L and installed Compaq's math library for Alpha Linux which is up to 8 times faster than the standard Linux libm library. I have not compiled using it yet, but I'm curious what performance gains (if any) I might see when/if I do. Also as an update to my previous post. Both my servers have been confirmed available on the Internet: Alpha 433MHz: http://24.5.14.144:3000 K6-180MHz: http://24.5.14.144 The fastest map will be the San Diego map on the Alpha. It is using a UW SCSI HDD that is local to the Alpha to serve data. The other Alpha maps are using NFS drives and are bandwidth limited to 10Mbps. You will notice the difference with detailed maps and there are still some development issues. Thanks, PGA From pgallen at randomlogic.com Thu Dec 2 12:14:15 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Thu, 02 Dec 1999 12:14:15 -0800 Subject: MapServer and C math functions References: <38465B35.E68D5501@randomlogic.com> <38466753.EFF73D0A@here.is> Message-ID: <3846D317.C2017D22@randomlogic.com> Ma wrote: > Hi Pall, > > I tried your servers, but I got some errors: > for http://24.5.14.144/ > by applications: > An error has occurred in msLoadMap(): Unknown identifier. (LABELOVERLAP):(14) > An error has occurred in msLoadMap(): Unknown identifier. (LABELOVERLAP):(15) > An error has occurred in msLoadMap(): Unknown identifier. (LABELOVERLAP):(15) > An error has occurred in msLoadMap(): Unknown identifier. (LABELOVERLAP):(15) > An error has occurred in msOpenSHPFile(): Unable to access file. (mgt6073) - with > all layers > and the same for the other one (port 3000) > > Could you check them out. > > Best rgrds, Damn. I forgot to remove the LABELOVERLAP line from all the .map files. Even I am not perfect. ;) Thanks for letting me know. PGA From pgallen at randomlogic.com Fri Dec 3 14:21:23 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Fri, 03 Dec 1999 14:21:23 -0800 Subject: MapServer and C math functions References: <3846D5F7.BA44EEF3@randomlogic.com> Message-ID: <38484263.9E99F190@randomlogic.com> "Paul G. Allen" wrote: > Stephen Lime wrote: > > > Hmmm... Probably just pow, cos/acos, and sin/asin. Probably have more > > effect in the Proj and freetype libraries. > > > > Then the faster math library will definitely help these portions of mapserv. These > are some of the slowest functions listed (and as a rule should always be avoided if > possibly when code speed is an issue). I couldn't get PROJ.4 to compile with the Compaq math library, but everything else did. PROJ.4 wants to use __atan2() and the Compaq library only contains atan2(). I noticed that libm (the Linux math library) contains both __atan2() and atan2(). I thought there was a compiler switch that would remove leading "_" from library function names but I couldn't find it. Maybe I was thinking of another compiler? > > > > > > BTW Did you try recompiling mapserv without your pointer mods after fixing > > egcs. Seems strange not to allow int *, yet char * still works. I did talk to the > > shapelib author and he had compiled his stuff on an alpha, but not with linux > > worked without any mods. > > > > I could easily do this. I will be recompiling anyway to try the Compaq math library > and when you release another version of mapserv. It's simple enough to modify the > header file to force POINTER = (int). Tonight I will look at compiling again without the POINTER mod and also see if I can track down the labeling problem. I was thinking, the labeling problem is similar to the one that you fixed a version or two ago (another 3.3 alpha) where my Field Unit labels would crash the CGI if I tried to display too many of them. If the map only has a couple streets, streams, or whatever displayed, then the labels work fine. Any more and it crashes with an error. Again, this is only on the Alpha. PGA From steve.lime at dnr.state.mn.us Fri Dec 3 14:26:00 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Fri, 03 Dec 1999 16:26:00 -0600 Subject: Greyscale JPEG's Message-ID: Greetings: Anybody wanting greyscale JPEG support, particularly useful with DOQ images, will find it added in 3.3 beta which will be available soon... s. From pgallen at randomlogic.com Sat Dec 4 17:17:19 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Sat, 04 Dec 1999 17:17:19 -0800 Subject: Alpha labeling problems Message-ID: <3849BD1F.4A7B795B@randomlogic.com> I've been trying to narrow down the source of the problem a little more. If I change from TT fonts to bitmap, I can display more labels on a map before it crashes, but it does still crash. Since the problem seems not to be limited to TrueType fonts, I'll see if I can use the shp2img interface and bitmap fonts to try and isolate the problem. PGA From kames at erols.com Sat Dec 4 17:53:30 1999 From: kames at erols.com (Kieran Ames) Date: Sat, 04 Dec 1999 20:53:30 -0500 Subject: Compiling for NT Message-ID: <3849C59A.EB6920A3@erols.com> Okay, I guess I'll go ahead and ask the question... Is anyone planning on compiling version 3.3 (including the Utility Programs and MapScript), for NT 4.0? Thanks much. Kieran From bfraser at geoanalytic.com Sun Dec 5 21:00:41 1999 From: bfraser at geoanalytic.com (Brent Fraser) Date: Sun, 5 Dec 1999 22:00:41 -0700 Subject: Compiling for NT References: <3849C59A.EB6920A3@erols.com> Message-ID: <000701bf3fa6$e12f2eb0$030102c0@cghh1.ab.wave.home.com> I've compiled the mapserver executable, but not MapScript yet. I'm having some problems with the make files and the compiler. Brent ----- Original Message ----- From: "Kieran Ames" To: Sent: Saturday, December 04, 1999 6:53 PM Subject: Compiling for NT > Okay, I guess I'll go ahead and ask the question... > Is anyone planning on compiling version 3.3 (including the Utility > Programs and MapScript), for NT 4.0? > > Thanks much. > > Kieran > > From praveen at regencyinfotech.com Mon Dec 6 10:27:00 1999 From: praveen at regencyinfotech.com (praveen) Date: Mon, 6 Dec 1999 23:57:00 +0530 Subject: response Message-ID: <001e01bf4017$84606ac0$ba653fca@regency> response to the mailing list. Am in interested in Web GIS. Praveen -------------- next part -------------- An HTML attachment was scrubbed... URL: From steve.lime at dnr.state.mn.us Tue Dec 7 07:18:52 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Tue, 07 Dec 1999 09:18:52 -0600 Subject: how to compile ms_3.3 with freetype Message-ID: Edit the mapserver Makefile and change the Freetype section to point to where you installed it. It should look something like: TTF_LIB=-L./gdft -lgdft -L/path/to/dir/containing/library -lttf TTF_INC=-I./gdft -I/path/to/dir/containing/headers TTF=-DUSE_TTF TTF_GD=gdft Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 >>> Peter Jacobs 12/07/99 08:13AM >>> Hello, For my mapserv I use a host where I'm not able to install freetype on the normal directories like /usr/bin because I don't have the privilliges. So I installed the libraries of freetype with ./configure --prefix and --exec-prefix. When I type make in the ms_3.3 directory I get all sorts of errors like missing freetype.h Who can please help me with the makefile of mapserv Regards Peter Jacobs From pcameron at gis.umn.edu Tue Dec 7 11:26:08 1999 From: pcameron at gis.umn.edu (Paul Cameron) Date: Tue, 7 Dec 1999 13:26:08 -0600 (CST) Subject: mapserv/mapscript 3.3 compilation probs Message-ID: Hello, I don't know a whole lot about Makefiles so I hope this makes some sense. I'm having some problems getting the newest version of mapserv/mapscript to compile on IRIX 5.3. The problem seems to be related to the removal of the regex directory from the distribution. In map.h, line 46 there is the line: #include The compilation would crash when it tried to find that file. I got around this problem by copying a regex directory from an old distribution over and adding these lines in the Makefile (pulled from older Makefile): REGEXP_LIB=-L./regex -lregex REGEXP_INC=-I./regex That fixed the problem for compiling the main mapserv program but now I'm having the same problem with mapscript and I'm not sure how to modify the Makefile so it can find the regex.h file. The output from compiling looks like cc -c -I../shapelib -I../gd-1.2 -I../gdft -D_POSIX_SOURCE -ansiposix -D_BSD_TYPES -Olimit 3000 -I/usr/local/include -DLANGUAGE_C -O -DVERSION=\"0.10\" -DXS_VERSION=\"0.10\" -KPIC -I/usr/local/lib/perl5/IP22-irix/5.00404/CORE -DUSE_TTF -DUSE_TIFF -DUSE_EPPL -DUSE_REGEXP mapscript_wrap.c cfe: Error: ./../map.h: 46: Cannot open file regex.h for #include make: *** [mapscript_wrap.o] Error 1 Any help would be much appreciated, Paul Cameron From steve.lime at dnr.state.mn.us Tue Dec 7 19:44:38 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Tue, 07 Dec 1999 21:44:38 -0600 Subject: Need an algorithm... Message-ID: In order to get scalable symbols working I need to be able to apply multiple parallel brushes with different dash intervals. Results are *very* nice. However I need an algorithm to draw parallel lines where a line is a series of connected segments. Is anyone aware of such a beast? Steve From steve.lime at dnr.state.mn.us Mon Dec 13 14:28:43 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Mon, 13 Dec 1999 16:28:43 -0600 Subject: Version 3.3 beta - soon... Message-ID: Just so folks don't think I'm slacking off I wanted to let people know that I plan to release 3.3 beta early this week. There are ton's of new features and I've added a couple of new ones just in the last few days. Here are just a few: - support for gd versions 1.2 and up - support for greyscale JPEG images - font drop shadows - automatic font rotation - new labeling controls including offsets, duplicate control, label buffers, etc... - expanded class/query expressions to 3 types * string compare * regular expressions * logical expressions (shapefiles only) > now you can write range queries like ([AREA] > 20000 and [AREA] <= 100000) > or do calculations like ([AREA]/[PERIMETER] > 0.7) - introducing querymaps- a mechanism to map query results - form access to virtually all map file parameters - expanded symbol sets - concatenated annotation (eg. TEXT "[FIRSTNAME], [LASTNAME]") - inline features (eg. features defined IN the map file) - mapscript - perl based access to the core mapserver API (java,tk/tcl,python versions possible as well) - raster classes - reclassify raster images on-the-fly - unique temporary image names (pid + time) - one-to-many joins Lots of bug fixes (and of course new bugs) etc... Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 From bfraser at geoanalytic.com Tue Dec 14 11:14:46 1999 From: bfraser at geoanalytic.com (Brent Fraser) Date: Tue, 14 Dec 1999 12:14:46 -0700 Subject: FYI: press release: OGC SEEKS INPUT ON WEB MAPPING STANDARD Message-ID: <001e01bf4667$7c12dcd0$390002c0@servo> MapServers, I know some have seen this press release, but for those not in the loop, here it is... Brent Fraser GeoAnalytic Inc. 1150 , 633 - 6th Avenue SW Calgary, AB T2P 2Y5 Tel: (403)213-2700 Fax: (403)213-2707 bfraser at geoanalytic.com www.geoanalytic.com ============================================= OGC SEEKS INPUT ON WEB MAPPING STANDARD Wayland, MA, USA, December 13, 1999: The Open GIS Consortium, Inc. (OGC) today issued Requests for Comment on its pending OpenGIS Web Map Server Interface Specification and OpenGIS Geography Markup Language (GML) Specification. The OpenGIS Web Map Server Interface Specification provides a set of open protocols that will make it possible for users of ordinary web browsers to obtain and automatically overlay "map layers" of the same geographic region, where each layer is a "view" into geographic data potentially held by a different web server. Map servers can be from different vendors, and they can vary widely in terms of processing capabilities and data type (including multiple vector and raster types). The three protocols are relatively simple, and it is not difficult in most cases to add them to data servers. GetCapabilities provides information about what a server can do. GetMap gets maps, which might be georeferenced images (such as JPEG and PNG), simple display graphics (such as SVG or WebCGM), or more universally displayable "picture element" data (Simple Features XML, also known as GML). GetFeature_info provides information about features, so that users can query to get attributes of specific geographic features in a map layer. GML is a method for encoding OpenGIS Simple Features (basic vector-based geographic data) in XML. XSLT "stylesheets" can then be used to render this "smart data" in application-specific ways. The GML draft document will be submitted to W3C as a Note. In the OpenGIS Web Mapping Testbed, the Web Map Server was used with catalog services defined in the OpenGIS Catalog Services Specification, which provides a common architecture for online automated directories or registries of web-based geospatial data and geoprocessing services, rather like "spatial search engines." As vendors implement these open standards in software, and as online geodata suppliers make their data available, web users will easily find, view, overlay, and combine different thematic maps for a given region. Geospatial data includes digital maps (including individual map elements such as roads, lakes and political boundaries), Earth images, and database records with street addresses or other fields that indicate physical location. Much geospatial data is currently available on the web, but users must posses considerable expertise and special geographic information system (GIS) software to overlay or otherwise combine different "map layers" of the same geographic region. Now, consensus among GIS software vendors in OGC has made it possible for such overlays and combinations of complex and essentially different kinds of geographic information to happen automatically over the Internet, despite differences in the vendors' geographic data storage, analysis, and display systems. The OpenGIS Technical Document 99-077 titled, "Web Map Server Interface Specification" was submitted by BBN Technologies (Cambridge MA), Cadcorp Ltd. (UK), CubeWerx Inc. (Canada), Ionic Software s.a. (Belgium), Laser-Scan Ltd. (UK), SICAD Geomatics GmbH & Co. oHG. (Germany), Social Change Online Pty Ltd. (Australia), and the US Army Engineer Research and Development Center (Alexandria, VA). The OpenGIS Technical Document 99-082r1 titled "Geography Markup Language (GML)" was submitted by Galdos Systems, Inc. (Canada), Oracle Corp. (Redwood Shores, CA), MapInfo Corp. (Troy, NY), Compusult, Ltd (Canada), and CubeWerx (Canada). Both documents are available for comment at http://www.opengis.org/techno/request.htm. After the OGC Technical Committee reviews and incorporates comments, these specifications are likely to be formally adopted (as a single specification) at the OGC Technical Committee and Management Committee meetings to be held in Vancouver, BC in February, 2000. OGC is an international, not-for-profit organization working toward integration of geospatial capabilities into the world's information systems. Lance McKee Vice President, Corporate Communications and Public Sector Programs Open GIS Consortium, Inc. 35 Main Street Wayland, MA 01778 USA Tel. 508-655-5858 Fax 508-655-2237 http://www.opengis.org lmckee at opengis.org www.opengis.org From steve.lime at dnr.state.mn.us Tue Dec 14 12:16:10 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Tue, 14 Dec 1999 14:16:10 -0600 Subject: FYI: press release: OGC SEEKS INPUT ON WEB MAPPING STANDARD Message-ID: I'd like hear what folks involved with the OGC think about these proposed standards and whether or not the MapServer should strive to support them. I've skimmed the rfc's and in many cases the MapServer could easily be extended to support them. I do have concerns about things like layer symbolization and annotation especially with regards to label collisions. Stick maps are not what I'd call progress given what might be possible within an individual system. The earthquake mapping MapScript demo shows what can happen when an app is dependent on remote data. I shudder at the thought of trying to tie remote datasets of any reasonable size into a robust "public" application. GML may well be suited to non-realtime exchange of geodata but seems a bit bloated (ascii + xml tags) for high performance apps. Will GML be well suited to streaming - an implementation question? Still, good topics to discuss and possibly move forward on within the context of the MapServer. Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 >>> "Brent Fraser" 12/14/99 01:14PM >>> MapServers, I know some have seen this press release, but for those not in the loop, here it is... Brent Fraser GeoAnalytic Inc. 1150 , 633 - 6th Avenue SW Calgary, AB T2P 2Y5 Tel: (403)213-2700 Fax: (403)213-2707 bfraser at geoanalytic.com www.geoanalytic.com ============================================= OGC SEEKS INPUT ON WEB MAPPING STANDARD Wayland, MA, USA, December 13, 1999: The Open GIS Consortium, Inc. (OGC) today issued Requests for Comment on its pending OpenGIS Web Map Server Interface Specification and OpenGIS Geography Markup Language (GML) Specification. The OpenGIS Web Map Server Interface Specification provides a set of open protocols that will make it possible for users of ordinary web browsers to obtain and automatically overlay "map layers" of the same geographic region, where each layer is a "view" into geographic data potentially held by a different web server. Map servers can be from different vendors, and they can vary widely in terms of processing capabilities and data type (including multiple vector and raster types). The three protocols are relatively simple, and it is not difficult in most cases to add them to data servers. GetCapabilities provides information about what a server can do. GetMap gets maps, which might be georeferenced images (such as JPEG and PNG), simple display graphics (such as SVG or WebCGM), or more universally displayable "picture element" data (Simple Features XML, also known as GML). GetFeature_info provides information about features, so that users can query to get attributes of specific geographic features in a map layer. GML is a method for encoding OpenGIS Simple Features (basic vector-based geographic data) in XML. XSLT "stylesheets" can then be used to render this "smart data" in application-specific ways. The GML draft document will be submitted to W3C as a Note. In the OpenGIS Web Mapping Testbed, the Web Map Server was used with catalog services defined in the OpenGIS Catalog Services Specification, which provides a common architecture for online automated directories or registries of web-based geospatial data and geoprocessing services, rather like "spatial search engines." As vendors implement these open standards in software, and as online geodata suppliers make their data available, web users will easily find, view, overlay, and combine different thematic maps for a given region. Geospatial data includes digital maps (including individual map elements such as roads, lakes and political boundaries), Earth images, and database records with street addresses or other fields that indicate physical location. Much geospatial data is currently available on the web, but users must posses considerable expertise and special geographic information system (GIS) software to overlay or otherwise combine different "map layers" of the same geographic region. Now, consensus among GIS software vendors in OGC has made it possible for such overlays and combinations of complex and essentially different kinds of geographic information to happen automatically over the Internet, despite differences in the vendors' geographic data storage, analysis, and display systems. The OpenGIS Technical Document 99-077 titled, "Web Map Server Interface Specification" was submitted by BBN Technologies (Cambridge MA), Cadcorp Ltd. (UK), CubeWerx Inc. (Canada), Ionic Software s.a. (Belgium), Laser-Scan Ltd. (UK), SICAD Geomatics GmbH & Co. oHG. (Germany), Social Change Online Pty Ltd. (Australia), and the US Army Engineer Research and Development Center (Alexandria, VA). The OpenGIS Technical Document 99-082r1 titled "Geography Markup Language (GML)" was submitted by Galdos Systems, Inc. (Canada), Oracle Corp. (Redwood Shores, CA), MapInfo Corp. (Troy, NY), Compusult, Ltd (Canada), and CubeWerx (Canada). Both documents are available for comment at http://www.opengis.org/techno/request.htm. After the OGC Technical Committee reviews and incorporates comments, these specifications are likely to be formally adopted (as a single specification) at the OGC Technical Committee and Management Committee meetings to be held in Vancouver, BC in February, 2000. OGC is an international, not-for-profit organization working toward integration of geospatial capabilities into the world's information systems. Lance McKee Vice President, Corporate Communications and Public Sector Programs Open GIS Consortium, Inc. 35 Main Street Wayland, MA 01778 USA Tel. 508-655-5858 Fax 508-655-2237 http://www.opengis.org lmckee at opengis.org www.opengis.org From steve.lime at dnr.state.mn.us Tue Dec 14 12:24:18 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Tue, 14 Dec 1999 14:24:18 -0600 Subject: Load balancing Message-ID: Anyone using Apache might be interested in a load balancing module called backhand. Seems kinda cool - see www.backhand.org for more information. Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 From steve.lime at dnr.state.mn.us Tue Dec 14 14:39:50 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Tue, 14 Dec 1999 16:39:50 -0600 Subject: FYI: press release: OGC SEEKS INPUT ON WEB MAPPING STANDARD Message-ID: I would probably look at building conformity using mapscript rather than changing the base C code. That is write perl wrappers to send/recieve GML or answer the "Get" type requests. Most of this is just undertanding a different syntax either via a URL or using XML and Perl is exceptionally well suited to those tasks. MapServer pretty much already understands the OGC features since they are all allowed within a shapefile anyway. Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 >>> "Sullivan, James R." 12/14/99 03:48PM >>> Believe the OGC folks have some very good ideas. But it is part of their job to be 'tea leaf readers' to some extent. Believe Mapserver should strive to be as close to OGC as possible without getting 'painted into a corner'. How would Mapserver conform to the Object wrappers being prescirbed by OGC? Jim Sullivan -----Original Message----- From: Stephen Lime [SMTP:steve.lime at dnr.state.mn.us] Sent: Tuesday, December 14, 1999 3:16 PM To: bfraser at geoanalytic.com; mapserver-users at lists.gis.umn.edu Subject: Re: FYI: press release: OGC SEEKS INPUT ON WEB MAPPING STANDARD I'd like hear what folks involved with the OGC think about these proposed standards and whether or not the MapServer should strive to support them. I've skimmed the rfc's and in many cases the MapServer could easily be extended to support them. I do have concerns about things like layer symbolization and annotation especially with regards to label collisions. Stick maps are not what I'd call progress given what might be possible within an individual system. The earthquake mapping MapScript demo shows what can happen when an app is dependent on remote data. I shudder at the thought of trying to tie remote datasets of any reasonable size into a robust "public" application. GML may well be suited to non-realtime exchange of geodata but seems a bit bloated (ascii + xml tags) for high performance apps. Will GML be well suited to streaming - an implementation question? Still, good topics to discuss and possibly move forward on within the context of the MapServer. Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 >>> "Brent Fraser" 12/14/99 01:14PM >>> MapServers, I know some have seen this press release, but for those not in the loop, here it is... Brent Fraser GeoAnalytic Inc. 1150 , 633 - 6th Avenue SW Calgary, AB T2P 2Y5 Tel: (403)213-2700 Fax: (403)213-2707 bfraser at geoanalytic.com www.geoanalytic.com ============================================= OGC SEEKS INPUT ON WEB MAPPING STANDARD Wayland, MA, USA, December 13, 1999: The Open GIS Consortium, Inc. (OGC) today issued Requests for Comment on its pending OpenGIS Web Map Server Interface Specification and OpenGIS Geography Markup Language (GML) Specification. The OpenGIS Web Map Server Interface Specification provides a set of open protocols that will make it possible for users of ordinary web browsers to obtain and automatically overlay "map layers" of the same geographic region, where each layer is a "view" into geographic data potentially held by a different web server. Map servers can be from different vendors, and they can vary widely in terms of processing capabilities and data type (including multiple vector and raster types). The three protocols are relatively simple, and it is not difficult in most cases to add them to data servers. GetCapabilities provides information about what a server can do. GetMap gets maps, which might be georeferenced images (such as JPEG and PNG), simple display graphics (such as SVG or WebCGM), or more universally displayable "picture element" data (Simple Features XML, also known as GML). GetFeature_info provides information about features, so that users can query to get attributes of specific geographic features in a map layer. GML is a method for encoding OpenGIS Simple Features (basic vector-based geographic data) in XML. XSLT "stylesheets" can then be used to render this "smart data" in application-specific ways. The GML draft document will be submitted to W3C as a Note. In the OpenGIS Web Mapping Testbed, the Web Map Server was used with catalog services defined in the OpenGIS Catalog Services Specification, which provides a common architecture for online automated directories or registries of web-based geospatial data and geoprocessing services, rather like "spatial search engines." As vendors implement these open standards in software, and as online geodata suppliers make their data available, web users will easily find, view, overlay, and combine different thematic maps for a given region. Geospatial data includes digital maps (including individual map elements such as roads, lakes and political boundaries), Earth images, and database records with street addresses or other fields that indicate physical location. Much geospatial data is currently available on the web, but users must posses considerable expertise and special geographic information system (GIS) software to overlay or otherwise combine different "map layers" of the same geographic region. Now, consensus among GIS software vendors in OGC has made it possible for such overlays and combinations of complex and essentially different kinds of geographic information to happen automatically over the Internet, despite differences in the vendors' geographic data storage, analysis, and display systems. The OpenGIS Technical Document 99-077 titled, "Web Map Server Interface Specification" was submitted by BBN Technologies (Cambridge MA), Cadcorp Ltd. (UK), CubeWerx Inc. (Canada), Ionic Software s.a. (Belgium), Laser-Scan Ltd. (UK), SICAD Geomatics GmbH & Co. oHG. (Germany), Social Change Online Pty Ltd. (Australia), and the US Army Engineer Research and Development Center (Alexandria, VA). The OpenGIS Technical Document 99-082r1 titled "Geography Markup Language (GML)" was submitted by Galdos Systems, Inc. (Canada), Oracle Corp. (Redwood Shores, CA), MapInfo Corp. (Troy, NY), Compusult, Ltd (Canada), and CubeWerx (Canada). Both documents are available for comment at http://www.opengis.org/techno/request.htm. After the OGC Technical Committee reviews and incorporates comments, these specifications are likely to be formally adopted (as a single specification) at the OGC Technical Committee and Management Committee meetings to be held in Vancouver, BC in February, 2000. OGC is an international, not-for-profit organization working toward integration of geospatial capabilities into the world's information systems. Lance McKee Vice President, Corporate Communications and Public Sector Programs Open GIS Consortium, Inc. 35 Main Street Wayland, MA 01778 USA Tel. 508-655-5858 Fax 508-655-2237 http://www.opengis.org lmckee at opengis.org www.opengis.org From steve.lime at dnr.state.mn.us Thu Dec 16 23:33:42 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Fri, 17 Dec 1999 01:33:42 -0600 Subject: MapServer 3.3 Beta Message-ID: It's up for grabs. Most everything seems to be working just fine. Still updating a few things on the docs and will be adding the demo suite over the next week. Enjoy. Any folks interested in helping with demos, documentation or an FAQ please let me know. The biggest problem I see is getting people up to speed on how to use it. You can do some very cool stuff with the new version and I want very much to get the word out. But for now I'm tired and am going to bed. Start finding those bugs! Steve From pgallen at randomlogic.com Fri Dec 17 04:27:26 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Fri, 17 Dec 1999 04:27:26 -0800 Subject: MapServer 3.3 Beta References: Message-ID: <385A2C2E.AB56B048@randomlogic.com> Stephen Lime wrote: > It's up for grabs. Most everything seems to be working just fine. Still updating a few things on the docs and will be adding the demo suite over the next week. Enjoy. Any folks interested in helping with demos, documentation or an FAQ please let me know. The biggest problem I see is getting people up to speed on how to use it. You can do some very cool stuff with the new version and I want very much to get the word out. But for now I'm tired and am going to bed. Start finding those bugs! > In the next week I will be attempting many things and putting Mapscript to the test. You can bet I'll be taking notes as I go, and once I'm happy with what I have, I'll be making it public. :-) PGA From pgallen at randomlogic.com Fri Dec 17 06:04:17 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Fri, 17 Dec 1999 06:04:17 -0800 Subject: Mapscript demo, what's the URL? Message-ID: <385A42E1.943215C6@randomlogic.com> Since I'm not currently at home, I can't look at my mail archives to find the answer. So, can someone tell me where that Mapscript example/demo is located? Thanks, PGA From steve.lime at dnr.state.mn.us Fri Dec 17 08:36:32 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Fri, 17 Dec 1999 10:36:32 -0600 Subject: MapServer 3.3 Beta Message-ID: Wow, that didn't take long. It figures I'd screw something up that was easy. I forgot to re-SWIG the interface file after the last few structure edits. I re-posted the initial beta with the correct Perl code. Grab it again and try it. Note the MapScript interface has changed from the initial demos. Actually I expect this MapScript to change pretty rapidly. The docs are up-to-date with regards to MapScript. Hmmm. A powered by mapserver/mapscript logo. Have to think about that. Any artists out there? Steve >>> Robert Chavez 12/17/99 10:35AM >>> Steve, I'd be happy to help out with demos and FAQs for mapscript. The demo I've been working on is actually nearing completion and has some basic perl-mapscript interactions in it the might be worth documenting. I'm in the final bug fixing stage now, but I'll pass the URL along when it's finished. Once the atlas is finished it will run off the Perseus Project web site which sees a fair amount of action (about 300,000 unique hits a day). Is there some standard boiler-plate text or graphic I could attach at the bottom of the page (Powered by MapServer, for example) that would give credit where credit is due and direct interested users back to the source? if not I'll just make something up. You've done an awesome job with all this. Just tried compiling mapscript from the new beta. Got this error message: mapscript_wrap.c: In function `_wrap_labelCacheMemberObj_layer_set': mapscript_wrap.c:3312: structure has no member named `layer' mapscript_wrap.c: In function `_wrap_labelCacheMemberObj_layer_get': mapscript_wrap.c:3333: structure has no member named `layer' mapscript_wrap.c: In function `_wrap_labelCacheMemberObj_class_set': mapscript_wrap.c:3356: structure has no member named `class' mapscript_wrap.c: In function `_wrap_labelCacheMemberObj_class_get': mapscript_wrap.c:3377: structure has no member named `class' mapscript_wrap.c: In function `_wrap_labelCacheMemberObj_tile_set': mapscript_wrap.c:3400: structure has no member named `tile' mapscript_wrap.c: In function `_wrap_labelCacheMemberObj_tile_get': mapscript_wrap.c:3421: structure has no member named `tile' mapscript_wrap.c: In function `_wrap_labelCacheMemberObj_shape_set': mapscript_wrap.c:3444: structure has no member named `shape' mapscript_wrap.c: In function `_wrap_labelCacheMemberObj_shape_get': mapscript_wrap.c:3465: structure has no member named `shape' make: *** [mapscript_wrap.o] Error 1 Looks like something is missing somewhere. Rob Stephen Lime wrote: > It's up for grabs. Most everything seems to be working just fine. Still updating a few things on the docs and will be adding the demo suite over the next week. Enjoy. Any folks interested in helping with demos, documentation or an FAQ please let me know. The biggest problem I see is getting people up to speed on how to use it. You can do some very cool stuff with the new version and I want very much to get the word out. But for now I'm tired and am going to bed. Start finding those bugs! > > Steve From steve.lime at dnr.state.mn.us Fri Dec 17 11:01:00 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Fri, 17 Dec 1999 13:01:00 -0600 Subject: MapServer 3.3 App Message-ID: Just to get the creative juices going... Here's a mapserver 3.3 app for viewing 1:24K - 1:250K DRG files. Files exist as geotiffs for each quad and are tiled for use within the mapserver. Total volume is 9.1 Gb. Maps are served via NFS from a raid and it seems pretty snappy. This example uses querymaps to provide a map AND a list of corresponding quads. Will share configs with anyone interested, http://www.dnr.state.mn.us/cgi-bin/mapserv3.3?map=/usr/local/www/docs/mapping/quads/view.map&mapxy=527907.773870+5042917.947352&layer=quads24k&buffer=1219&mode=nquery Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 From pgallen at randomlogic.com Sat Dec 18 23:14:05 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Sat, 18 Dec 1999 23:14:05 -0800 Subject: Mapscript demo, what's the URL? References: <385A42E1.943215C6@randomlogic.com> Message-ID: <385C85BD.278A7137@randomlogic.com> I got the URL, Thanks. PGA From pgallen at randomlogic.com Mon Dec 20 06:27:52 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Mon, 20 Dec 1999 06:27:52 -0800 Subject: Mapscript (Perl) question Message-ID: <385E3CE7.C5A46B76@randomlogic.com> I'm still a bit new to Perl, so please bear with me. I'm a little confused about 'new'. When a line such as: $map = new mapObj(""); Is this creating a new file named or does the file already exist and it simple allows the use of this file in the script? PGA From bfraser at geoanalytic.com Mon Dec 20 06:58:02 1999 From: bfraser at geoanalytic.com (Brent Fraser) Date: Mon, 20 Dec 1999 07:58:02 -0700 Subject: Mapscript (Perl) question References: <385E3CE7.C5A46B76@randomlogic.com> Message-ID: <002001bf4afa$9cfeffe0$390002c0@servo> Paul, It creates a "map object" with properties set according to the values in your already-existing "filename.map". These properties are then available to the rest of your Perl script. Brent ----- Original Message ----- From: "Paul G. Allen" To: "MapServer Users List" Sent: Monday, December 20, 1999 7:27 AM Subject: Mapscript (Perl) question > I'm still a bit new to Perl, so please bear with me. > > I'm a little confused about 'new'. When a line such as: > > $map = new mapObj(""); > > Is this creating a new file named or does the file already exist > and it simple allows the use of this file in the script? > > PGA From steve.lime at dnr.state.mn.us Mon Dec 20 07:01:47 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Mon, 20 Dec 1999 09:01:47 -0600 Subject: convert shapefiles Message-ID: I wrote a simple shapefile dumping program using MapScript. It's distributed in the mapscript/examples directory in the new 3.3 version. Dump.pl shows you how to access the spatial side of a shapefile. You can use DBI and the DBD::Xbase module to access the .dbf side. The is also a 'create.pl' script you can modify to take data from just about any perl accessible location and create shapefiles. Be careful with polygon data as the shapefile specification talks about vertex ordering for inner (i.e. holes) vs outer rings. MapServer does not require that ordering (at this time) but other packages may. Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 >>> "Peter H.M. Jacobs" 12/20/99 05:07AM >>> Does anyone know how to convert shapefiles to comma separated value files, or these CSV files to Shapefiles? Regards, peter jacobs From rchavez at perseus.tufts.edu Mon Dec 20 07:16:32 1999 From: rchavez at perseus.tufts.edu (Robert Chavez) Date: Mon, 20 Dec 1999 10:16:32 -0500 Subject: Mapscript (Perl) question References: <385E3CE7.C5A46B76@randomlogic.com> Message-ID: <385E4850.AAE85D53@perseus.tufts.edu> It's creating a new mapObj out of "filename.map" (which is your already existing map file). I'm actually having a problem with this in the latest mapscript implementation. Mapscript dies on the $map = new mapObj call where it did not in the previous implementation. My map file hasn't changed and runs fine in mapserv, but not mapscript. Has anyone run across this problem? Rob Chavez Programmer Perseus Project "Paul G. Allen" wrote: > I'm still a bit new to Perl, so please bear with me. > > I'm a little confused about 'new'. When a line such as: > > $map = new mapObj(""); > > Is this creating a new file named or does the file already exist > and it simple allows the use of this file in the script? > > PGA From steve.lime at dnr.state.mn.us Mon Dec 20 07:08:11 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Mon, 20 Dec 1999 09:08:11 -0600 Subject: Mapscript (Perl) question Message-ID: Correct. At present a map object has to be initialized using a map file. However, I'll eventually allow for the creation of an "empty" map object that you could build upon. You'll see the use of constructors throughout mapscript. I deceided to go with a object oriented interface for a number of reasons- it's real clean and scripts written in other languages will look very similar. Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 >>> "Brent Fraser" 12/20/99 08:58AM >>> Paul, It creates a "map object" with properties set according to the values in your already-existing "filename.map". These properties are then available to the rest of your Perl script. Brent ----- Original Message ----- From: "Paul G. Allen" To: "MapServer Users List" Sent: Monday, December 20, 1999 7:27 AM Subject: Mapscript (Perl) question > I'm still a bit new to Perl, so please bear with me. > > I'm a little confused about 'new'. When a line such as: > > $map = new mapObj(""); > > Is this creating a new file named or does the file already exist > and it simple allows the use of this file in the script? > > PGA From steve.lime at dnr.state.mn.us Mon Dec 20 09:41:49 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Mon, 20 Dec 1999 11:41:49 -0600 Subject: NT Binary of 3.3 available. Message-ID: Brent Fraser has graciously provided his latest NT build of MapServer 3.3 for download. In his words, "it makes a map GIF" but is otherwise untested. Your millage may vary. Anyway, nab it at: http://mapserver.gis.umn.edu/3.3/download.html Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 From steve.lime at dnr.state.mn.us Mon Dec 20 11:03:10 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Mon, 20 Dec 1999 13:03:10 -0600 Subject: Line symbol changes... Message-ID: Greetings: Just an FYI. The line symbol format has changed from 3.2 to 3.3. I added more flexability to STYLED line symbols so that more complex symbols like railroad tracks or utility lines are now possible with some caveats. This change is documented in the 3.3 online docs and I sent new symbol sets with the 3.3 distribution. You will need to swap the line sets as you move from 3.2 to 3.3. Shade and marker sets are unchanged except that you can now name symbols and refer to them by name from a map file. Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 From pgallen at randomlogic.com Mon Dec 20 11:20:28 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Mon, 20 Dec 1999 11:20:28 -0800 Subject: Mapscript question Message-ID: <385E817B.5965A8F7@randomlogic.com> I didn't see my other post show up on the list, so here it is again. I'm a little confused about the use of 'new'. What does the following line do: $map = new mapObj(".map"); Does it create a map file with the name .map or does it create an object in Mapscript using an existing file whose name is .map, so that it can be referenced/used in Mapscript? PGA From pgallen at randomlogic.com Mon Dec 20 17:31:37 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Mon, 20 Dec 1999 17:31:37 -0800 Subject: Mapscript (Perl) question References: Message-ID: <385ED879.FB0391CA@randomlogic.com> I finally received the replies to my question. Wonder what took them so long? Oh well, thanks. PGA From steve.lime at dnr.state.mn.us Wed Dec 22 09:35:52 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Wed, 22 Dec 1999 11:35:52 -0600 Subject: 3.3.001 is available Message-ID: Just letting folks know that 3.3.001 is available via the main MapServer website. It contains fixes for a few bugs reported by users. (Note: Brent it does not contain the hacks for the parser or PROJ.4 with win32) One significant change is in how polygons/polylines are labeled. I added "floating" labels for polygons that behave similarly to line labeling. Now, the visable portion of a polygon is labeled. You can tune this in 2 ways. You can use the label parameter MINFEATURESIZE to help avoid labeling slivers or small inclusions. The value of MINFEATURESIZE coresponds to the minimum width or height in pixels of the bounding box for a particular feature. You can also use the label parameter PARTIALS to set whether partial labels are allowed (ie. can labels run off the edge of an image). The default is true. The floating polygon labels seem to work quite well. let me know what you all think. Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 From mdh20 at axe.humboldt.edu Wed Dec 22 13:46:10 1999 From: mdh20 at axe.humboldt.edu (Michael Hass) Date: Wed, 22 Dec 1999 13:46:10 -0800 (PST) Subject: mapimage java question In-Reply-To: Message-ID: This may be a long shot... I am trying to get example 5 up and running (java/javascript version) and the applet won't start. Debugging with the appletviewer, I get the following error message: java.lang.UnsatisfiedLinkError:InitClass at netscape.javascript.JSObject(JSObject.java:61) at mapimage.init(mapimage.java:29) at sun.applet.AppletPanel.run(AppletPanel.java:343) at java.lang.Thread.run(Thread.java:479) I am using linux and installed java1.2.2 I set the path and classpath to netscape.javascript (java40.jar). Does anyone know if the error message is stating that classpath for javascript is in error? Thanks for any help. I'm anxious to get on to 3.3, but obviously need to figure this out first. Mike Hass From fsvpss at eng.chula.ac.th Thu Dec 23 02:03:23 1999 From: fsvpss at eng.chula.ac.th (Phisan Santitamnont) Date: Thu, 23 Dec 1999 17:03:23 +0700 Subject: [Fwd: good news from good guy] Message-ID: <3861F36B.226B31BE@eng.chula.ac.th> Rob Atkinson wrote: > Good to hear from you. > > The latest news is that the OpenGIS consortium (www.opengis.org) is now > developing formal standards for web mapping. Social Change Online is > involved in this process and will be releasing an application server > product in 2000 to use these. > > We are starting to develop online deomstrations etc at > webmap.socialchange.net.au > > I recommend following this Open Standards based approach. > > I believe Mapserver will be modified to become compliant with these > standards. > > If you have specific challenges you would like advice on, or you wish to > use our expertise or products in a project please feel free to contact > us. > > Regards > > Rob Atkinson > > Phisan Santitamnont wrote: > > > Dear sir, > > > > i am very glad to get to know another > > Fornet Mapserver's user. > > > > Where can i get the bunch of > > the "Common Spatial Gateway Inteface" developed by Social Change > > Online. > -- > _______________________________________________________________ > Rob Atkinson Email: rob at socialchange.net > Director > Social Change Online Mob: 0419 202 973 > 6A Nelson Street Tel: +61 (0) 2 9557 6500 > Annandale NSW 2038 Fax: +61 (0) 2 9519 8940 > Sydney, Australia WWW: http://online.socialchange.net.au -- Phisan Santitamnont, Dr.-Ing. Survey Engineering Department Chulalongkorn University Pathumwan Rd, Bangkok 10330 Tel. +66 (2) 218 66 61 Fax. +66 (2) 218 66 53 mailto:fsvpss at eng.chula.ac.th http://www.sv.eng.chula.ac.th/Staff/Phisan From fsvpss at eng.chula.ac.th Thu Dec 23 01:59:30 1999 From: fsvpss at eng.chula.ac.th (Phisan Santitamnont) Date: Thu, 23 Dec 1999 16:59:30 +0700 Subject: mapimage java question Message-ID: <3861F282.865074C8@eng.chula.ac.th> Hi Mike, >This may be a long shot... >I am trying to get example 5 up and running (java/javascript version) and >the applet won't start. Debugging with the appletviewer, I get the >following error message: > java.lang.UnsatisfiedLinkError:InitClass > at netscape.javascript.JSObject(JSObject.java:61) > at mapimage.init(mapimage.java:29) > at sun.applet.AppletPanel.run(AppletPanel.java:343) > at java.lang.Thread.run(Thread.java:479) i insist that you got problem with classpath , try this javac -classpath /usr/lib/netscape/java/classes/java40.jar mapimage.java Rgds, -- Phisan Santitamnont, Dr.-Ing. Survey Engineering Department Chulalongkorn University Pathumwan Rd, Bangkok 10330 Tel. +66 (2) 218 66 61 Fax. +66 (2) 218 66 53 mailto:fsvpss at eng.chula.ac.th http://www.sv.eng.chula.ac.th/Staff/Phisan From pgallen at randomlogic.com Thu Dec 23 07:36:31 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Thu, 23 Dec 1999 07:36:31 -0800 Subject: Mapscript problems Message-ID: <3862417F.B68A8979@randomlogic.com> I'm getting an error when trying to instantiate a new mapObj. The code looks like this: $map = new mapObj("/home/httpd/html/SD/sandiego.map"); and the error is: Usage: new_mapObj(filename,config); at /mapscript.pm line 1479 I am still using the last alpha version of MapServer 3.3. (I have not updated yet because I'd have to mess with the code to get it to work on my Alpha, and I've been saving that for a good stable release.) The map file does exist in the path indicated and permissions are correct. PGA From bfraser at geoanalytic.com Thu Dec 23 07:23:17 1999 From: bfraser at geoanalytic.com (Brent Fraser) Date: Thu, 23 Dec 1999 08:23:17 -0700 Subject: Mapscript problems References: <3862417F.B68A8979@randomlogic.com> Message-ID: <000b01bf4d59$a53cb330$030102c0@cghh1.ab.wave.home.com> Paul, There was a change in the msLoadMap function from 3.3 alpha to 3.3 beta (I'm assuming the change is reflected in MapScript too). Try using two parameters for mapObj, the first one is a map file as you've done, and the second one is the path to your mapserv.ini file (excluding the filename). For example: $map = new mapObj("/home/httpd/html/SD/sandiego.map","/home/httpd/cgi-bin/"); but I'm not a MapScript guy... Brent ----- Original Message ----- From: "Paul G. Allen" To: "MapServer Users List" Sent: Thursday, December 23, 1999 8:36 AM Subject: Mapscript problems > I'm getting an error when trying to instantiate a new mapObj. The code looks > like this: > > $map = new mapObj("/home/httpd/html/SD/sandiego.map"); > > and the error is: > > Usage: new_mapObj(filename,config); at /mapscript.pm line > 1479 > > I am still using the last alpha version of MapServer 3.3. (I have not updated > yet because I'd have to mess with the code to get it to work on my Alpha, and > I've been saving that for a good stable release.) > > The map file does exist in the path indicated and permissions are correct. > > PGA From pgallen at randomlogic.com Thu Dec 23 08:36:37 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Thu, 23 Dec 1999 08:36:37 -0800 Subject: Mapscript problems References: <3862417F.B68A8979@randomlogic.com> Message-ID: <38624F95.2DD78181@randomlogic.com> After looking at the mapscript wrapper document, I realized the problem was that the mapObj() method changed. So I updated to 3.3.001. Now I am having another problem: the command is failing and not creating the $map object. The command is: die mapscript::ms_error->{message} unless $map = new mapObj("/home/httpd/html/SD/sandiego.map"); PGA "Paul G. Allen" wrote: > > I'm getting an error when trying to instantiate a new mapObj. The code looks > like this: > > $map = new mapObj("/home/httpd/html/SD/sandiego.map"); > > and the error is: > > Usage: new_mapObj(filename,config); at /mapscript.pm line > 1479 > > I am still using the last alpha version of MapServer 3.3. (I have not updated > yet because I'd have to mess with the code to get it to work on my Alpha, and > I've been saving that for a good stable release.) > > The map file does exist in the path indicated and permissions are correct. > > PGA From pgallen at randomlogic.com Thu Dec 23 10:40:00 1999 From: pgallen at randomlogic.com (Paul G. Allen) Date: Thu, 23 Dec 1999 10:40:00 -0800 Subject: Mapscript question Message-ID: <38626C80.1954312A@randomlogic.com> I'm a little at a loss. If I have a multi-selection list box on my web page for selecting layers to display, how would I draw these layers using mapscript? PGA From steve.lime at dnr.state.mn.us Thu Dec 23 10:52:58 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Thu, 23 Dec 1999 12:52:58 -0600 Subject: Mapscript question Message-ID: Remember MapScript is NOT mapserv. You can build mapservish apps with it, and can do things you simply couldn't do with mapserv. Here's one simple way to process a request for a set of layers that mimics the way mapserv does it: ... $cgi = CGI; $map = new mapObj('test.map'); @user_layers = $cgi->param('layer'); for($i=0; $i<$map->{numlayers}; $i++) { # for all layers $layer = $map->getLayer($i); # get the layer object if($layer->{status} != $mapscript::MS_DEFAULT && $layer->{status} != $mapscript::MS_QUERY) { # leave these layers alone $layer->{status} = $mapscript::MS_OFF; # by default foreach (@user_layers) { if($_ eq $layer->{name} || $_ eq $layer->{group}) { # user wants this layer $layer->{status} = $mapscript::MS_ON; break; } } } } # now modify extents and whatever else you might want to do $img = $map->draw(); # all layers defined so draw the map # now save the map some place or even send it back to the client using Mime: image/gif. ... This chunk of code simply turns all layers off, then sets those selected by the user to on. It ignores DEFAULT and QUERY layers. It may be that this is so common that we'll want to set up a separate module mapscript::mapserv to mimic a few mapserver code pieces. Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 >>> "Paul G. Allen" 12/23/99 12:40PM >>> I'm a little at a loss. If I have a multi-selection list box on my web page for selecting layers to display, how would I draw these layers using mapscript? PGA From steve.lime at dnr.state.mn.us Tue Dec 28 09:02:29 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Tue, 28 Dec 1999 11:02:29 -0600 Subject: looking for... Message-ID: Greetings: I've ported a quadtree spatial index originally written for the shapelib to work under MapServer. It seems to perform nicely, at least as fast as my original grid indexer and hopefully faster. It's easier to use than my old code. I'll release it later today with version 3.3.003 but I'd like to run some timing tests and need some really huge (eg. lots of shapes) shapefiles to play with. Anyone got some single files with 250K to 1 million shapes in them? Please let me know and I'll arrange a transfer somehow. Thanks - Steve Stephen Lime Internet Applications Analyst Minnesota DNR 500 Lafayette Road St. Paul, MN 55155 651-297-2937 From imap at chesapeake.net Thu Dec 30 12:02:18 1999 From: imap at chesapeake.net (imap at chesapeake.net) Date: Thu, 30 Dec 1999 15:02:18 -0500 Subject: 3.3 beta compile notes Message-ID: <386BBA4A.87BF4A3B@chesapeake.net> Steve/mapserver list, On the 3.3_001 beta, if you are compiling on NT, please note the following In mapstring.c, needed the strcasecmp and strncasecmp stuff added back in as a conditional compile. That does not exist in the MSVC6 world... and probably built-in to the gnu libraries. In the mapparser.c, there are references to /usr/lib/bison.simple. On MSVC6, you need to need to a copy of this in the compilation directory and the reference needs to change to "bison.simple" Also in mapparser.c, you need to add a directive at near the top: #ifdef _WIN32 #include #endif Everything else *appears* to compile okay as far as I can tell. -- Chris Stuber (mapsurfer) Silicon Mapping Solutions, Inc. From banister at banister.com Tue Dec 28 16:43:12 1999 From: banister at banister.com (Trevor Banister) Date: Tue, 28 Dec 1999 17:43:12 -0700 Subject: Modified shpindex that reports status Message-ID: <6D5E4B9090F1D011AC3A00A0C903430804CF8D@www.banister.com> I dug into the source to answer my own question about the grid settings. I added a status function to shpindex.c so I could see how long the indexing process would take on some of my large (>500,000) shape files. Turns out that is will take a few days for some of them :-) Hopefully the speedup in map rendering will be worth it. Add this to the next release if you want. Is there enough mods and stuff like this that people send in to start a "contrib" folder in the distribution? Trevor /* shpindex.c, from Mapserver 3.3 by Stephen Lime.*/ /* Modified by Trevor Banister */ /* Added verbose mode, prints status every 100 shapes if argv[4] isn't NULL*/ /* i.e. shpindex myfile 10 10 verbose */ #include "map.h" #include "mapindex.h" int main(int argc, char *argv[]) { SHPHandle hSHP; int nShapeType; int nEntities; rectObj shpBounds; double tmpBounds[4]; int i; int verbose; char *filename; indexObj *idx; if(argc<4) { fprintf(stdout,"Syntax: shpindex [shpfile] [sx] [sy] [verbose]\n" ); exit(0); } if (argv[4] != NULL) { verbose=1; } else { verbose=0; } /* ** Open shapefile */ if((hSHP = SHPOpen(argv[1], "r" )) == NULL) { fprintf(stderr, "Error opening shapefile %s.\n", argv[1]); exit(0); } /* ** Load some information about this shapefile, don't need much */ SHPGetInfo(hSHP, &nEntities, &nShapeType); SHPReadBounds(hSHP, -1, tmpBounds); shpBounds.minx = tmpBounds[0]; shpBounds.miny = tmpBounds[1]; shpBounds.maxx = tmpBounds[2]; shpBounds.maxy = tmpBounds[3]; /* ** Initialize the index */ if((idx = msNewIndex(shpBounds, atoi(argv[2]), atoi(argv[3]), nEntities)) == NULL) { msWriteError(stderr); exit(0); } /* ** For each shape */ for(i=0;i Message-ID: Is there a limit to the size of a tiff that can be effectively displayed? I've tried to display a tiff that is 26meg's and then a smaller one that was 4meg's and I am either getting a pixel here a pixel there (like it blew up) or its greying out half way (this is running the java/javascript version). The map file is written in the same manner as the demo, the tif has a wld file that works, etc. I'm wondering if I need to degrade the image to below a meg to get effective results. Thanks Mike Hass HSU From steve.lime at dnr.state.mn.us Thu Dec 30 20:36:56 1999 From: steve.lime at dnr.state.mn.us (Stephen Lime) Date: Thu, 30 Dec 1999 22:36:56 -0600 Subject: tiff size Message-ID: MapServer can easily display TIFs 15K on a side. The image is sampled using nearest neighbor which is all that makes sense for colormapped data. The effects of this sampling vary greatly from image to image. For example, DRGs pretty much suck at all but a couple of sampling intervals. This is because they are more like line art and lines (eg. roads and contours) can easily disapear and reappear. Conversely, continuous data like doqs, landsat images or wall-to-wall classifications look fine. I posted a url a week or so ago for a DRG viewer and the only way I could get it to work was to scrap infinite zooms in favor of specific displays at scales that I know the image will look good. With 8-bit output this is just the way things work no matter what package your using. 24 -bit output allows for cubic-convolution or bilinear interpolation resampling which will greatly enhance certain types of image viewing- at a cost though- speed. I personally have used image files larger than 300 Mb with no problems. Send a sample if this doesn't answer your question. Steve <<< Michael Hass 12/30 3:38p >>> Is there a limit to the size of a tiff that can be effectively displayed? I've tried to display a tiff that is 26meg's and then a smaller one that was 4meg's and I am either getting a pixel here a pixel there (like it blew up) or its greying out half way (this is running the java/javascript version). The map file is written in the same manner as the demo, the tif has a wld file that works, etc. I'm wondering if I need to degrade the image to below a meg to get effective results. Thanks Mike Hass HSU