[postgis-devel] Unknown SRID (Was: #1223: Unknown SRID to NULL)

Fri Sep 30 18:30:10 PDT 2011

Paul,
Me thinks you worry too much.

1) First of all, most of the tools our users are used to using are only
starting to support 3DM because our support for those wasn't anything to
write home about (so if you are going to make precedents now is the time
when we have a lot of 3D goodness to offer) and the other tools that did
either weren't GIS based or followed the ISO/MM more closely.  So the fact
your output looks more standard is a good thing as it opens up possibliites
of tools not currently supporting us having an easier time to.

2) My point about the SRID is if we are going to be SQL/MM compliant in
every other way, why make an exception here?  It's like traveling to the
moon and missing because you were too scared to put your foot on it. Older
versions of a lot of apps don't work for 2.0 for a myriad of reasons, QGIS
etc. have already made some effort for us, they are already in the mindset
of change. OSM got a bit burned because they hard coded the gist type in the
index builde etc.  One extra thing especially SRID is not going to increase
the pitch forks against us as Hobu so politely suggested, but it wil if we
waiver.  The apps are already undergoing surgery.  Better to make all the
necessary incisions before we close up the patient. My only regret is that
we didn't do it earlier in the development phase.

Thanks,
Regina

> -----Original Message-----
> From: postgis-devel-bounces at postgis.refractions.net 
> [mailto:postgis-devel-bounces at postgis.refractions.net] On 
> Behalf Of Paul Ramsey
> Sent: Friday, September 30, 2011 7:23 PM
> To: PostGIS Development Discussion
> Subject: Re: [postgis-devel] Unknown SRID (Was: #1223: 
> Unknown SRID to NULL)
> 
> Speaking of which, another major breaking change isn't turned on yet:
> return ISO WKT and WKB in ST_AsText and ST_AsBinary.
> 
> Will not affect anyone working with 2D data, but will slap 
> upside the head anyone working with Z's and M's.
> 
> I just worry about ruining our user base. worry!
> 
> P.
> 
> On Fri, Sep 30, 2011 at 10:51 AM, Paragon Corporation 
> <lr at pcorp.us> wrote:
> > What compatibility reasons do we have that are any worse than the 
> > others we have had?
> >
> > For example PostGIS 2.0 --  We got rid of all those functions, we 
> > changed our index types.  Those all broke a lot of tools some fixed 
> > some not.  We survived.  Everyone survived without too much 
> grumbling.
> >
> > 1) Dealling with SRID -1 is even less minor because few people that 
> > care to do transforms use them and when you are not doing 
> transforms, 
> > the srid is rarely used for anything but display anyway.
> > Most apps that use it do it out of ignorance or because they are 
> > tracking non-GIS things.
> >
> > 2) Not allowing negative SRIDs in our typmod gives us more space to 
> > store other things and if we don't store it that way on 
> earth would we 
> > output it that way.
> >
> > 3) OGC compliance
> > 4) If apps are really concerned about supporting both 2.0 and prior 
> > versions
> > -- what's wrong with anything < 1 is unknown?  Works for 
> both 1.5 and 
> > 2.0
> >
> > Strk -- your null idea is very unappealing.  If I were expecting a 
> > number and I got  a NULL back that I think will break a lot more 
> > things than returning -1 e.g you can't glue it If you tried to do 
> > ST_SRID() || ST_ASText  --- you get NULL back.  Very scary.
> >
> > .  Besides I am not aware
> > Of any spatial spec that dictates NULL so its going from one 
> > iconoclasm to another.  Remember we are trying to conform.  Yah it 
> > means missing or whatever -- but no spatial database does that Or 
> > respects that notion of SRID.What happens when I export the data to 
> > another spatial database like SQL Server that expects 0.
> >
> >
> >  5) Guys we've been planning this for how many years?  I 
> knew it was 
> > going to happen two years ago - many other people were waiting and 
> > asking why it hadn't happened in 2.0 already.  Some people I have 
> > talked to were happy about the conformance.  I didn't here any 
> > complains except from Hobu.  Why are we breaking our 
> promise of changing.
> > Why this sudden want to backout and after we already did 
> half the work 
> > and changed our tests?
> >
> >
> > Thanks,
> > Regina
> >
> >
> > -----Original Message-----
> > From: postgis-devel-bounces at postgis.refractions.net
> > [mailto:postgis-devel-bounces at postgis.refractions.net] On Behalf Of 
> > Paul Ramsey
> > Sent: Friday, September 30, 2011 11:56 AM
> > To: PostGIS Development Discussion
> > Subject: Re: [postgis-devel] Unknown SRID (Was: #1223: 
> Unknown SRID to 
> > NULL)
> >
> > Yeah, a little context would be good. I want zero too, but only for 
> > aesthetic reasons that are easily overwhelmed by backwards 
> > compatibility stuff.
> > P.
> >
> > On Fri, Sep 30, 2011 at 8:54 AM, Sandro Santilli 
> <strk at keybit.net> wrote:
> >> On Fri, Sep 30, 2011 at 03:30:50PM -0000, rob wrote:
> >>
> >>>  I want 0.
> >>
> >> Why is that ?
> >>
> >> --strk;
> >>
> >>  ()   Free GIS & Flash consultant/developer
> >>  /\   http://strk.keybit.net/services.html
> >> _______________________________________________
> >> postgis-devel mailing list
> >> postgis-devel at postgis.refractions.net
> >> http://postgis.refractions.net/mailman/listinfo/postgis-devel
> >>
> > _______________________________________________
> > postgis-devel mailing list
> > postgis-devel at postgis.refractions.net
> > http://postgis.refractions.net/mailman/listinfo/postgis-devel
> >
> >
> > _______________________________________________
> > postgis-devel mailing list
> > postgis-devel at postgis.refractions.net
> > http://postgis.refractions.net/mailman/listinfo/postgis-devel
> >
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-devel
> 