[OSGeo-Discuss] OSGEO & OGC spec development

Michael P. Gerlek mpg at lizardtech.com
Thu Jul 26 19:34:53 PDT 2007 

A few more things:

I had a good chat w/ Sam B. from OGC tonight here at GeoWeb: they are very sincere in their desire to find some means by which OGC and OSGeo can work together, and are moving in the right direction for this.  Maybe not as fast as all of us would like or in exactly the right direction, but the effort is being made, they do want to engage with us.  Raj will be key here, I think, as would a meeting of the minds in Victoria.

We (Sam and I) also seem to agree that OSGeo makes a good environment for prototype / reference implementation work (or perhaps interop, as Allan discusses below).  Chris Holmes and TOPP are leading the way on this one.

Another issue that I've been thinking about is "the right to fork".  Let's say we have a model whereby a "specification" developed by OSGeo-type folks goes into the OGC process to become a "formal" spec (whatever that means).  Let's further hypothesize that the OGC members later on make a change to that spec in a way that the original OSGeo authors disagree with, or, equivalently, that the original OSGeo authors wish to make a change to the spec that OGC is unwilling to ratify.  Obviously we'd want to work this out amicably if at all possible -- but in the worst case OGC would need to agree to have a non-exclusive license to the spec work (whatever that means), such that it could be "forked" by any disaffected parties at any time...  This is, of course, the normal open source development practice and philosophy.  Sam (unofficially) seemed to feel this was not unreasonable.

I don't know what our policy is, but to prevent boring anyone on this topic I'll just move for the creation of "standards at osgeo.org".  Rough charter: "This is a discussion list for topics having to do with creation of standards, specifications, and related beasties within the OSGeo ecosystem.  This includes, but is not limited to, discussions of how to best work with OGC and other such bodies.  Note that discussions about issues specific to a given standard or spec, e.g. should it be lat/long or long/lat, should be held on some other mailing list about that standard or spec; this mailing list is more for meta level issues."

Do I hear a second?

-mpg

 

> -----Original Message-----
> From: discuss-bounces at lists.osgeo.org 
> [mailto:discuss-bounces at lists.osgeo.org] On Behalf Of Allan Doyle
> Sent: Thursday, July 26, 2007 4:25 PM
> To: OSGeo Discussions
> Subject: Re: [OSGeo-Discuss] OSGEO & OGC spec development
> 
> 
> On Jul 26, 2007, at 18:08 , Michael P. Gerlek wrote:
> 
> > Sean has a good point: some of the OGC specs have been developed  
> > with relatively few members in the working group, which I 
> think can  
> > tend to lead to inclusion of some obscure feature just because  
> > there's not a wide enough group to object.
> 
> Some pretty odd things get by in the plenary, too. There seemed  
> always to be a general trend towards "if in doubt, keep it in" as  
> opposed to any real efforts to keep things simple.
> 
> I think the testbeds have also become a bit of a liability rather  
> than a benefit. There are too few people working on any given topic  
> and/or the people working on something are spread too thin. Thus the  
> work is often pretty superficial and in the end, hastily thrown  
> together into a demo that is far removed from a true interop style  
> trial of the interfaces. Then the results of the testbed are 
> packaged  
> into specs that tend to get approved pretty easily at the plenary.
> 
> In the end, you need to have a small number of people representing a  
> broad enough view who care passionately about the outcome and who  
> understand the technology well enough to build implementations and  
> evangelize the use of the spec. I don't see much of OGC or ISO work  
> fitting that profile. Interestingly enough, GML actually fits that  
> description.
> 
> I also think that it doesn't matter whether a specification gets  
> developed in a standards body or outside it. If it's any good, it  
> will get used. The notion that "governments like to use ISO 
> specs" is  
> really an excuse. I suspect there was a running web browser 
> on nearly  
> every government desktop computer before HTTP and HTML ever 
> even came  
> close to being IETF or W3C specs. Google Earth showed up in 
> the White  
> House [1] before KML was handed over to OGC and I bet there's not an  
> ISO 191xx spec in sight inside Google Earth.
> 
> 	Allan
> 
> [1] http://news.com.com/8301-10784_3-6128904-7.html
> 
> >
> > -mpg
> >
> >
> >
> >> -----Original Message-----
> >> From: discuss-bounces at lists.osgeo.org
> >> [mailto:discuss-bounces at lists.osgeo.org] On Behalf Of Sean Gillies
> >> Sent: Tuesday, July 24, 2007 10:24 AM
> >> To: OSGeo Discussions
> >> Subject: Re: [OSGeo-Discuss] OSGEO & OGC spec development
> >>
> >> Michael,
> >>
> >> Standards bodies are a good thing if they produce good standards.
> >>
> >> OGC standards tend toward fussy intricacy (compare WFS-T 
> to the Atom
> >> Publishing Protocol) and pointless abstraction (all the so-called
> >> distributed computing platforms that no one uses). I don't
> >> know why, but
> >> I suspect that it's cultural (no, I don't mean Canadian
> >> culture). Going
> >> public has the potential to reform the culture of the OGC.
> >>
> >> Regards,
> >> Sean
> >>
> >> Michael P. Gerlek wrote:
> >>> FYI, later this week at the GeoWeb conference in Vancouver
> >> we're having a discussion on this hot topic:
> >>>
> >>>> Ever wonder why we need standards bodies?  Can we just do
> >> it with a Wiki?
> >>>> We have open source, why not open source open standards?
> >> What about intellectual
> >>>> property protection?  Can I afford to belong to a
> >> standards body?  Can I afford
> >>>> not to? Do standards bodies impede or drive innovation?
> >> How should neo-geo and
> >>>> OGC work together?  What are the pitfalls of "going public"?
> >>>
> >>> I'll be on the panel, as will Carl Reed from OGC.
> >>>
> >>> (Feel free to send me any comments/positions you'd like me
> >> to put forward.)
> >>>
> >>> -mpg
> >>>
> >>>
> >>>
> >>>
> >>> ________________________________
> >>>
> >>> 	From: discuss-bounces at lists.osgeo.org
> >> [mailto:discuss-bounces at lists.osgeo.org] On Behalf Of 
> Jeroen Ticheler
> >>> 	Sent: Monday, July 16, 2007 2:12 PM
> >>> 	To: OSGeo Discussions
> >>> 	Subject: [OSGeo-Discuss] OSGEO & OGC spec development
> >>> 	
> >>> 	
> >>> 	Hi all,
> >>> 	Last week I attended the Open Geospatial Consortium
> >> Technical Committee (OGC-TC) meeting in Paris.
> >>>
> >>> 	For those not to familiar with this meeting, it
> >> consists of a series of Working Group (WG) meetings that
> >> mostly run around the development of specifications (or
> >> standards if you wish) dealing with geo-informatics. The most
> >> prominent specifications coming from OGC are Web Map Service
> >> (WMS), Web Feature Service (WFS) and Geographic Markup
> >> Language (GML). There's a whole list of other specs available
> >> or under development. OSGEO projects work with a substantial
> >> number of them. See http://www.opengeospatial org for more details.
> >>>
> >>> 	With this email I would like to touch upon two issues
> >> that I think are relevant to OSGEO. I hope bringing this up
> >> can trigger some discussion on how OSGEO would best benefit
> >> from the OGC spec development process:
> >>>
> >>> 	1- Discussions related to Google's KML and Web Map Context
> >>> 	2- Discussions related to a Tiled Web Map Service specifications
> >>>
> >>> 	There was discussion on the possibility that KML
> >> becomes an OGC specification and, more importantly, that it
> >> could be used to replace the wining Web Map Context (WMC)
> >> specification. A number of OSGEO projects use the Styled
> >> Layer Descriptors (SLD (symbology)) specification and the
> >> WMC. There's a great deal of overlap between these and KML.
> >> It is likely in the interest of these projects to share their
> >> experience with OGC and see some of that reflected in future
> >> OGC specs.
> >>>
> >>> 	There was also discussion about a new Tiled WMS
> >> specification. Such spec can have different forms, and could
> >> be conceived as a new spec or as an extension (or application
> >> profile) of a Web Map Service. Two approaches were presented
> >> and two other approaches were mentioned, among which the
> >> approach taken within the OSGEO community.
> >>>
> >>> 	Observing these discussions, my impression is that
> >> OSGEO has an important role to play in the further
> >> development of these OGC specs. We can obviously take the
> >> easy route and let OGC go its way. We could than come up with
> >> in-house, open specifications that will compete with OGC
> >> specs still under development. The development of the specs
> >> is likely to be quicker than going through OGC. However, I
> >> feel that with limited effort by the community we can have a
> >> very positive influence on the OGC spec development. We can
> >> make sure experiences in OSGEO are reflected in the OGC
> >> specs. The WMS-T is an obvious example of this. It was kind
> >> of frustrating to not see that experience properly
> >> represented at the WMS-WG.
> >>>
> >>> 	OSGEO is very young still, so frustration is not an
> >> expression of dissatisfaction in this case :-) rather, I
> >> think it might be time to establish a way to formally
> >> represent OSGEO in OGC. This could be through those OSGEO
> >> members that already hold a TC level membership to OGC (the
> >> logical first step I would think) and later possibly through
> >> a direct OSGEO TC Membership to OGC. Also, we could consider
> >> a focal point in OSGEO where specification development is
> >> discussed and coordinated. This may have the form of a
> >> Committee for instance. I'm hesitant to propose new
> >> Committees, but if there's enough interest to have a central
> >> coordination point dealing with standards and specs, it may
> >> make sense :-)
> >>>
> >>> 	Greetings from Rome,
> >>> 	Jeroen
> >>>
> >>> 		_______________________
> >>> 	Jeroen Ticheler
> >>> 	FAO-UN
> >>> 	Tel: +39 06 57056041
> >>> 	http://www.fao.org/geonetwork
> >>> 	42.07420°N 12.34343°E
> >>>
> >>> 	
> >>>
> >>> _______________________________________________
> >>> Discuss mailing list
> >>> Discuss at lists.osgeo.org
> >>> http://lists.osgeo.org/mailman/listinfo/discuss
> >>>
> >>
> >> _______________________________________________
> >> Discuss mailing list
> >> Discuss at lists.osgeo.org
> >> http://lists.osgeo.org/mailman/listinfo/discuss
> >>
> > _______________________________________________
> > Discuss mailing list
> > Discuss at lists.osgeo.org
> > http://lists.osgeo.org/mailman/listinfo/discuss
> 
> -- 
> Allan Doyle
> +1.781.433.2695
> adoyle at eogeo.org
> 
> 
> 
> _______________________________________________
> Discuss mailing list
> Discuss at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/discuss
>

More information about the Discuss mailing list