[postgis-devel] PostgIS 3.0 beta 1 ready?

[Regina Obe]

Are we ready for a PostGIS 3.0 beta 1release?

 

Any more features we plan to put in 3.0 before we nail the box?

 

I'd like to package a PostGIS 3.0 beta 1 to go with the PostgreSQL 12 beta 1
as I see PostgreSQL 12 now shipping with PostGIS 2.5.2 

which is discouraging given the effort we've put in to take advantage of
PostgreSQL 12 features with 3.0.

 

relevant thread on IRC:

 

[11:42] <Myon> trying postgis on PG12beta1:
https://pgdgbuild.dus.dg-i.net/job/postgis-binaries-beta/architecture=amd64,
distribution=sid/35/console

[11:42] <sigq> Title: postgis-binaries-beta > amd64,sid #35 Console
[Jenkins] (at pgdgbuild.dus.dg-i.net)

[11:42] <Myon> is that a PostGIS problem, or one in PostgreSQL? (timestamp
17:39:21)

[11:43] <robe2> Myon you should be using PostGIS 3.0 for 12

[11:43] <Myon> well that doesn't exist yet

[11:43] <robe2> Myon if we shipped a beta would that help?

[11:43] <robe2> We are planning to release PostGIS 3.0 around same time as
12

[11:44] <robe2> because it has a lot of stuff just for 12 :)

[11:44] <Myon> a beta would help sortish, because I wouldn't want to ship to
users before it's stable, while still trying to support PG12

[11:45] <Myon> it's just these extra "WARNING:  stats for "t.g" do not
exist" lines, the rest of the tests works

[11:45] <Myon> does it make sense that these are duplicated now, or is that
something that should be fixed in PostgreSQL itself? (I haven't yet read any
postgis code)

[11:45] <robe2> Okay let me bring it up for discussion on postgis dev

<robe2> we aren't testing PostgreSQL 12 on 2.5 -- so it may not work
completely until release time as we won't be in a rush to fix issues on 12
in 2.5 until release

[11:46] <robe2> and the most we'd promise is it compiles -- not necessarily
it regresses

[11:48] <Algunenano> About the duplicated messages, I think that the issue
had to do with CTEs and robe2 fixed it in 3.0 and refushed (:D) to fixed it
in 2.5

[11:50] <robe2> :D) yap I refused we didn't put in all this effort in 3.0
for 12 just to have people shipping 2.5 with 12

 

 

If we aren't ready maybe we should just wait till PG 13 comes out and forgo
the 3.0 release in September. As there is no point in shipping a 3.0 mid
cycle after PostgreSQL 12 is released and if packagers are going to ship a
2.5.  Everyone will be settled in on a 2.5 and won't waste their time
upgrading to a 3.0 even if packagers were willing to ship a 3.0 with 12,
which most WILL NOT.

 

Thanks,

Regina

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/postgis-devel/attachments/20190521/b34de84b/attachment.html>