[postgis-devel] Aggregate STYPE

[Paul Ramsey]

OK, I think I've done all the due-diligence in
https://trac.osgeo.org/postgis/ticket/4035
P.

On Mon, Mar 5, 2018 at 9:16 AM, Regina Obe <lr at pcorp.us> wrote:
> One more thing, make sure to note it in the NEWS that it's a breaking change.
>
> It's a breaking change because it requires drop and recreate of all the aggregates so anyone that has views tied to the aggs, which seems a surprising number of people do
> Will have to drop their views before upgrade and recreate after upgrading.
>
> -----Original Message-----
> From: Regina Obe [mailto:lr at pcorp.us]
> Sent: Monday, March 05, 2018 12:01 PM
> To: 'PostGIS Development Discussion' <postgis-devel at lists.osgeo.org>
> Subject: RE: [postgis-devel] Aggregate STYPE
>
> Paul,
> Can you please put in a change notice on all the aggs and functions you changed.
>
> NOTE:  that all aggs would need to be dropped and recreated and our updater relies on those comments to know to drop the aggregate since we no longer always drop aggs.
> Functions we always replace so not quite as big of a deal to have change notice on that.
>
> Strk,
> Correct me if I misspoke.
>
> Thanks,
> Regina
>
> -----Original Message-----
> From: postgis-devel [mailto:postgis-devel-bounces at lists.osgeo.org] On Behalf Of Paul Ramsey
> Sent: Monday, March 05, 2018 9:15 AM
> To: PostGIS Development Discussion <postgis-devel at lists.osgeo.org>
> Subject: Re: [postgis-devel] Aggregate STYPE
>
> I've removed pgis_abs in https://trac.osgeo.org/postgis/ticket/4035
>
> On Sat, Mar 3, 2018 at 4:17 AM, Sandro Santilli <strk at kbt.io> wrote:
>> On Fri, Mar 02, 2018 at 02:52:03PM -0800, Paul Ramsey wrote:
>>> Memory test, ancient developers:
>>>
>>> The CREATE AGGREGATE for ST_Union has
>>>
>>>    stype = pgis_abs
>>>
>>> The CREATE AGGREGATE for ST_AsMVT has
>>>
>>>    stype = internal
>>>
>>> Now, pgis_abs is just a stub type, it's not real at all, the
>>> input/output functions don't do anything, etc. So, is the retention
>>> of that type an echo of an earlier time when it wasn't possible to
>>> aggregate with an internal stype, or is there an *extra* reason why
>>> we don't just use stype=internal for ST_Union()?
>>
>> My guess is we want "internal" nowadays, and "pgis_abs" is a leftover
>> from the far past. A quick look reveals that use of "internal" for
>> state types was introduced in PostGIS-2.4. PostgreSQL documentation
>> talk about an "internal" type even in version 7.3.
>>
>> --strk;
>> _______________________________________________
>> postgis-devel mailing list
>> postgis-devel at lists.osgeo.org
>> https://lists.osgeo.org/mailman/listinfo/postgis-devel
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/postgis-devel
>
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/postgis-devel