[postgis-users] ST_StartPoint fails with MultiLineString

[Jose Carlos Martinez Llario]

I agree that imagination and creative thinking is a good thing as 
PostGIS did improving the standard (not following it) many times.
But we are not talking about that now.
ST_StartPoint, ST_EndPoint, ST_GeometryN, etc. dont have an homogenous 
behavior, and others functions either.

If you guys are changing a lot of things in PostGIS 2.0 (what is really 
nice),  why you dont try to use
a homogenous  approach to these kind of functions?  I dont mind if 
following the standard or not  but my opinion is that PostGIS team 
should chose one side. If not we users are going to need a memo with 
every function to know what is the expected behavior because it does not 
follow any pattern, it means PostGIS is less easy to use than it could be.

Anyways this is just an opinion that I wanted to share with you.
Jose

On 24/11/2011 23:17, Paul Ramsey wrote:
> On Thu, Nov 24, 2011 at 1:39 PM, Sandro Santilli<strk at keybit.net>  wrote:
>
>> I guess nowadays it's not that worth being silly to follow
>> the standard ...
> Right, non-brain-dead behavior should be our default. Silly that we
> have st_npoints that works and st_numpoints that doesn't.
> P.
> _______________________________________________
> postgis-users mailing list
> postgis-users at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-users