Johnson, Steven E wrote: > Rob, > You've zeroed in on a point I tried to make, and not very clearly. Software > developers must be concerned with their products conforming to specs. Mere > mortals (and I consider myself one) need only concern themselves with how > they might apply and conform to the standard. The Hollow Earth framework is > an excellent example of how to *gently* bring end users to applying the > standard. I rather like the idea of a 'service profile' to help end users > *implement*. > > There seems to be two issues at play here: > 1) Educating the broader user community on how to apply standards, and > 2) the point raised by Bryce about the cost of 'open' standards > To be sure, the two are closely related and if that's the case how do we > reduce the cost, in terms of cost and culture, of adopting standards? > There is a lot to be said for making the tools compliant and easy to use...
I've got a whole raft of ideas in the wings. Once geoserver is able to cope (internally) with compliant data standards there is a lot we can do to turn configuration into a hand-holding process. Currently we half-support a "try to describe my existing database in a meaningful way", which I believe is actually so complex, in the context of any open system with multiple potential users, its impossible for experts, let alone infrequent service deployers. The only successful solutions are ones which put the cost in the appropriate place. We should avoid end-users having to pay the cost (of non-compliance) through a mess of services and data ill or inconsistently defined. The second priority is to let deployers configure in a no-brainer fashion. This means some up-front costs need to be borne by developers, but good libraries encapsualting the standards compliance issues, like Bryce has been working on can go a long way. Cheers Rob Atkinson > Cheers, > SEJ > ___________________________________ > Steven E. Johnson > Booz Allen Hamilton > http://www.boozallen.com > É 703.902.5149 voice > Ê 703.902.3613 fax > - [EMAIL PROTECTED] > þ 38 55 19.89 N / 077 13 55.45 W > ë USNG Coordinate: 18SUJ0650010511 > > -----Original Message----- > From: Rob Atkinson [mailto:[EMAIL PROTECTED] > Sent: Thursday, March 08, 2007 5:13 PM > To: Bryce L Nordgren > Cc: Andrea Aime; Johnson, Steven E; McQueen, Joshua; Cole, Deborah; Fox, > Thomas P; Geotools-Devel list; Chris Holmes; Andersen, Norman C.; Brill, > Gregory A.; Justin Deoliveira; geotools-devel@lists.sourceforge.net; [EMAIL > PROTECTED] > Subject: Re: [Geotools-devel] ISO Modular Standards & Cost Efficiency > > > G'day Bryce. > > welcome to my day job.... > > for many years I've been collaborating with Simon Cox (GML co-editor, > Observations & Measurements) and more recently with Clemens Portele (chair > INSPIRE Data Standards Implementation Guideline Drafting Team) - both > involved in "realising" the ISO standards stack. > > I dont think we should expect any but the standards architects, and a few key > software module developers to read the ISO specs. After that mortals should > work with APIs and simplified profiles. > > Even standards architects should break down into meta-modellers and domain > modellers, and domain modellers should be following well thought out patterns > (meta-models) which are in effect implementation profiles of the ISO stack.. > > Of particular interest are: > > 1) The "HollowWorld" modelling framework. This is available at > > https://www.seegrid.csiro.au/twiki/bin/view/AppSchemas/HollowWorld > > The HollowWorld framework makes the ISO conceptual framework available for > formal compliance within domain modelling activities. Much of the key > definitions etc are. > > ps. -we've been adding stuff on governance and version control it would be > neat to have reviewed, and to see if we can bridge across different UML tools > at some stage. > > 2) Use of service profiles > > Service profiles mean that implementers simply dont need to read the ISO > specs... > > https://www.seegrid.csiro.au/twiki/bin/view/AppSchemas/ServiceProfiles > > > Bryce L Nordgren wrote: > >> Andrea & GeoTools developers, >> >> I'm including a number of people from the US Technical Advisory Group >> to ISO TC/211 in this email, particularly the ones interested in "outreach". >> Believe it or not, that's a big thing right now, at least for the US >> contingent. Bearing in mind that no one here has the power to get ISO >> to stop charging for standards, and that the 19100 family is now and >> will forever be highly modular (i.e. interrelated), we'd be interested >> in hearing any constructive suggestions about how the concepts and >> content of the standards could be disseminated at an appropriate level >> in a more cost effective manner. Please feel free to respond even if >> you're not from the US. :) >> >> [EMAIL PROTECTED] wrote on 03/07/2007 >> 06:25:49 >> AM: >> >> >> >>> Warning, personal opinion here. If I'll ever have to work with ISO >>> something, and be paid to do so, I'll buy the relevant ISO standard. >>> >>> Yet, I'm not sure an organisation that makes "for buy" standards >>> deserves open source implementations of them. >>> >>> If we were to start copying with ISO seriously, we would end up using >>> so many inter-related standards that each one of us would have to pay >>> 1000$, not 30. That's totally unacceptable to me. >>> I'm not asking a dime to people downloading the stuff I did develop >>> in my own spare time. >>> >>> >> Oh boy do I understand that! ($800 and counting) However, there's a >> couple of factors which force us in this direction whether we like it or not: >> >> 1] GML3 (free/OGC) encodes time using the data model in 19108 (not >> free/ISO). >> 2] GML3 (free) provides only an encoding and does not provide much in >> the way of "explanatory text" or legal values (but is still 600 pages long). >> >> Also, observe that C++ is an ISO standard. Is any open source >> software written in C++? (yes) Are there any open-source >> implementations of the c++ language? (yes) Is it common practice to >> learn about the C++ standard template library by buying the ISO >> standard? (no) >> >> As to the problem of users not knowing how to use an ISO-based library >> without buying ISO standards: There is a difference in the level of >> knowledge required by an implementor and the level required by a user >> of a library. By and large, users never need to see standards or know >> algorithms. They just need a decent grasp of the rules which >> ultimately derive from the standard. What we need is an O'Reilly "ISO >> GIS in a Nutshell" book. I don't believe that ISO (or ANSI or OGC) >> would publish a "Nutshell" book, but there are at least two things we >> may be able to look into in order to ease the pain of newcomers, and >> I'd like to hear your opinions about them: >> >> 1] ANSI sometimes sells standards as sets for a discounted rate. What >> if the entire 19100 series was sold for one fixed price? Or if we >> made sets for Basic Geospatial Modeling using Features, Geospatial >> Services, Geospatial XML (not just GML), and other "topic areas"? >> What groupings would you like to see? What would help people get >> started with the least pain? >> >> 2] It is my understanding that ISO also produces documents which are >> reports, not standards. What if ISO produced one or more reports in >> various topic areas? These reports would be designed as summaries of >> the suite of standards targeted towards "users", not "implementors". >> These could serve as an introduction to the 45+ standards in the >> family, they could focus on how the standards build on each other >> (i.e. what tasks require what pieces from what standards) and have a much >> more "explanatory" >> style than the detail-oriented approach required by an actual standard. >> >> 3] Do we (GeoTools) want to start documenting various non-free >> standards in the spirit of the various primers I've written? >> Implementors still need to buy the standard (and will always have to do so). >> Users won't. >> >> Right now these are just thoughts, as I don't know exactly what we can >> and can't do or even what we are willing to do. Practically speaking, >> OGC and ISO TC/211 standards are married now and there's not a divorce in >> sight. >> So--we want to encourage adoption of these standards, you want to be >> able to interoperate without having ISO capitalize on your gratis >> implementation; how can we help each other out? >> >> Bryce >> >> >> ---------------------------------------------------------------------- >> --- Take Surveys. Earn Cash. Influence the Future of IT Join >> SourceForge.net's Techsay panel and you'll get the chance to share >> your opinions on IT & business topics through brief surveys-and earn >> cash >> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEV >> DEV _______________________________________________ >> Geotools-devel mailing list >> Geotools-devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/geotools-devel >> >> ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Geotools-devel mailing list Geotools-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/geotools-devel
