The EWKB/EWKT specs should be part of the PostGIS manual

[Ben Harper-2]

Sorry to rant about an open source product that's free,

But the EWKB/EWKT specs should really be included in the PostGIS manual.
Only after a fair amount of mistrial did I discover this:

http://postgis.refractions.net/pipermail/postgis-devel/2004-December/000710.html

If that was part of the manual, it would have saved me a fair amount
of confusion.

Additionally, the PostGIS manual claims that "every valid WKB/WKT is a
valid EWKB/EWKT" but I find this to be untrue.
Specifically, if I construct a WKB PointZ geometry, as specified by
http://portal.opengeospatial.org/files/?artifact_id=18241
and I try to read it via GeomFromEWKB(), I get garbage (A COMPOUND
CURVE is what comes out).
The docs should be updated to reflect this fact, or the WKB parser
should support this part of the spec.

Thanks,
Ben
_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[Mark Cave-Ayland-3]

Ben Harper wrote:

> Sorry to rant about an open source product that's free,
>
> But the EWKB/EWKT specs should really be included in the PostGIS manual.
> Only after a fair amount of mistrial did I discover this:
>
> http://postgis.refractions.net/pipermail/postgis-devel/2004-December/000710.html
>
> If that was part of the manual, it would have saved me a fair amount
> of confusion.

Thanks for the feedback. I probably speak for the majority of the
PostGIS team in that we are so familiar with the code that perhaps
things that are obvious to us aren't necessarily obvious to the users.
Most of our focus is on preparing for the 1.4 release at the moment, so
if you could contribute a patch then it would be gratefully accepted.

> Additionally, the PostGIS manual claims that "every valid WKB/WKT is a
> valid EWKB/EWKT" but I find this to be untrue.
> Specifically, if I construct a WKB PointZ geometry, as specified by
> http://portal.opengeospatial.org/files/?artifact_id=18241
> and I try to read it via GeomFromEWKB(), I get garbage (A COMPOUND
> CURVE is what comes out).
> The docs should be updated to reflect this fact, or the WKB parser
> should support this part of the spec.

Hmmm this smells like a bug. Are you working with 1.4 or 1.3? Can you
post a specific test case for us to look at?


Many thanks,

Mark.

--
Mark Cave-Ayland - Senior Technical Architect
PostgreSQL - PostGIS
Sirius Corporation plc - control through freedom
http://www.siriusit.co.uk
t: [phone]
_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[Ben Harper-2]

Thanks for the feedback.

I'll take some time on the weekend to prepare patches, etc.

Ben

On Fri, Jul 3, 2009 at 11:07 AM, Mark
Cave-Ayland<mark.cave-ayland@...> wrote:

> Ben Harper wrote:
>
>> Sorry to rant about an open source product that's free,
>>
>> But the EWKB/EWKT specs should really be included in the PostGIS manual.
>> Only after a fair amount of mistrial did I discover this:
>>
>>
>> http://postgis.refractions.net/pipermail/postgis-devel/2004-December/000710.html
>>
>> If that was part of the manual, it would have saved me a fair amount
>> of confusion.
>
> Thanks for the feedback. I probably speak for the majority of the PostGIS
> team in that we are so familiar with the code that perhaps things that are
> obvious to us aren't necessarily obvious to the users. Most of our focus is
> on preparing for the 1.4 release at the moment, so if you could contribute a
> patch then it would be gratefully accepted.
>
>> Additionally, the PostGIS manual claims that "every valid WKB/WKT is a
>> valid EWKB/EWKT" but I find this to be untrue.
>> Specifically, if I construct a WKB PointZ geometry, as specified by
>> http://portal.opengeospatial.org/files/?artifact_id=18241
>> and I try to read it via GeomFromEWKB(), I get garbage (A COMPOUND
>> CURVE is what comes out).
>> The docs should be updated to reflect this fact, or the WKB parser
>> should support this part of the spec.
>
> Hmmm this smells like a bug. Are you working with 1.4 or 1.3? Can you post a
> specific test case for us to look at?
>
>
> Many thanks,
>
> Mark.
>
> --
> Mark Cave-Ayland - Senior Technical Architect
> PostgreSQL - PostGIS
> Sirius Corporation plc - control through freedom
> http://www.siriusit.co.uk
> t: [phone]
> _______________________________________________
> postgis-users mailing list
> postgis-users@...
> http://postgis.refractions.net/mailman/listinfo/postgis-users
>

_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users

[Ben Harper-2]

And I'm on 1.3, but I'll test with 1.4 before calling anything out again.

On Fri, Jul 3, 2009 at 12:41 PM, Ben Harper<rogojin@...> wrote:

> Thanks for the feedback.
>
> I'll take some time on the weekend to prepare patches, etc.
>
> Ben
>
> On Fri, Jul 3, 2009 at 11:07 AM, Mark
> Cave-Ayland<mark.cave-ayland@...> wrote:
>> Ben Harper wrote:
>>
>>> Sorry to rant about an open source product that's free,
>>>
>>> But the EWKB/EWKT specs should really be included in the PostGIS manual.
>>> Only after a fair amount of mistrial did I discover this:
>>>
>>>
>>> http://postgis.refractions.net/pipermail/postgis-devel/2004-December/000710.html
>>>
>>> If that was part of the manual, it would have saved me a fair amount
>>> of confusion.
>>
>> Thanks for the feedback. I probably speak for the majority of the PostGIS
>> team in that we are so familiar with the code that perhaps things that are
>> obvious to us aren't necessarily obvious to the users. Most of our focus is
>> on preparing for the 1.4 release at the moment, so if you could contribute a
>> patch then it would be gratefully accepted.
>>
>>> Additionally, the PostGIS manual claims that "every valid WKB/WKT is a
>>> valid EWKB/EWKT" but I find this to be untrue.
>>> Specifically, if I construct a WKB PointZ geometry, as specified by
>>> http://portal.opengeospatial.org/files/?artifact_id=18241
>>> and I try to read it via GeomFromEWKB(), I get garbage (A COMPOUND
>>> CURVE is what comes out).
>>> The docs should be updated to reflect this fact, or the WKB parser
>>> should support this part of the spec.
>>
>> Hmmm this smells like a bug. Are you working with 1.4 or 1.3? Can you post a
>> specific test case for us to look at?
>>
>>
>> Many thanks,
>>
>> Mark.
>>
>> --
>> Mark Cave-Ayland - Senior Technical Architect
>> PostgreSQL - PostGIS
>> Sirius Corporation plc - control through freedom
>> http://www.siriusit.co.uk
>> t: [phone]
>> _______________________________________________
>> postgis-users mailing list
>> postgis-users@...
>> http://postgis.refractions.net/mailman/listinfo/postgis-users
>>
>

_______________________________________________
postgis-users mailing list
postgis-users@...
http://postgis.refractions.net/mailman/listinfo/postgis-users