[GRASS-dev] [release planning] GRASS GIS 8.4.0

Hi devs,

I'd like to discuss where we stand with the 8.4.0 release originally
planned for this month.

The remainder for the 8.4.0 milestone are (currently) 104 open pull
requests (PRs) which still need testing or review or to be postponed:
https://github.com/OSGeo/grass/pulls?q=is%3Aopen+is%3Apr+milestone%3A8.4.0

This is the full 8.4.0 milestone (open issues and pull requests):
https://github.com/OSGeo/grass/milestone/19

Please let's rank what still should go into 8.4.0 and what to postpone
to 8.4.1 or later.
Opinions welcome.

Best,
Markus

--
Markus Neteler, PhD
https://www.mundialis.de - company
https://grass.osgeo.org - FOSS
https://neteler.org - freelancing & blog

On Fri, 26 Apr 2024 at 05:36, Markus Neteler via grass-dev <grass-dev@lists.osgeo.org> wrote:

The remainder for the 8.4.0 milestone are (currently) 104 open pull
requests (PRs) which still need testing or review or to be postponed:
https://github.com/OSGeo/grass/pulls?q=is%3Aopen+is%3Apr+milestone%3A8.4.0

This is now down to 27. Kudos to Anna, myself and others.

I created a Future milestone discussed earlier to help with the management.

Vaclav Petras <wenzeslaus@gmail.com> schrieb am Di., 30. Apr. 2024, 16:29:

On Fri, 26 Apr 2024 at 05:36, Markus Neteler via grass-dev <grass-dev@lists.osgeo.org> wrote:

The remainder for the 8.4.0 milestone are (currently) 104 open pull
requests (PRs) which still need testing or review or to be postponed:
https://github.com/OSGeo/grass/pulls?q=is%3Aopen+is%3Apr+milestone%3A8.4.0

This is now down to 27. Kudos to Anna, myself and others.

We are down to 14, as far as I see.
Excellent.

I created a Future milestone discussed earlier to help with the management.

Thanks
Markus

Dear all,

It's been some weeks that we have released G 8.4.0RC1.

The milestones "8.4.0" still shows many open PRs/Issues:

https://github.com/OSGeo/grass/milestone/19

with 2 marked as "blockers":
https://github.com/OSGeo/grass/issues?q=is%3Aopen+is%3Aissue+label%3Ablocker

Really blockers or could they be downgraded in their importance (and
moved to the next 8.4.1 milestone)?

Thanks
Markus

--
Markus Neteler, PhD
https://www.mundialis.de - company
https://grass.osgeo.org - FOSS
https://neteler.org - freelancing & blog

On Thu, Jul 11, 2024 at 7:09 PM Markus Neteler <neteler@osgeo.org> wrote:

Dear all,

It's been some weeks that we have released G 8.4.0RC1.

The milestones "8.4.0" still shows many open PRs/Issues:

https://github.com/OSGeo/grass/milestone/19

--> 48 issues/PR still open.

So, I'll bump them to 8.4.1 shortly in order to move on with the 8.4.0 release.

with 2 marked as "blockers":
https://github.com/OSGeo/grass/issues?q=is%3Aopen+is%3Aissue+label%3Ablocker

Really blockers or could they be downgraded in their importance (and
moved to the next 8.4.1 milestone)?

... the blockers have been bumped to 8.4.1.

Still open is also the Weblate mess.

Markus

On Fri, Jul 19, 2024 at 4:30 PM Markus Neteler <neteler@osgeo.org> wrote:

On Thu, Jul 11, 2024 at 7:09 PM Markus Neteler <neteler@osgeo.org> wrote:
>
> Dear all,
>
> It's been some weeks that we have released G 8.4.0RC1.
>
> The milestones "8.4.0" still shows many open PRs/Issues:
>
> https://github.com/OSGeo/grass/milestone/19

--> 48 issues/PR still open.

So, I'll bump them to 8.4.1 shortly in order to move on with the 8.4.0 release.

So, only talking to myself here...

So I'll bump them now.

Still open is also the Weblate mess.

Markus

Hi,

I have started the G8.4.0 release procedure.
Will report once completed (takes some hours).

Markus

--
Markus Neteler, PhD
https://www.mundialis.de - company
https://grass.osgeo.org - FOSS
https://neteler.org - freelancing & blog

Hi Markus,

Great! :tada:

···

Dra. Verónica Andreo
Investigadora Adjunta de CONICET

Instituto Gulich (CONAE - UNC)

Centro Espacial Teófilo Tabanera (CETT)

Falda del Cañete - Córdoba, Argentina

+54 3547 400000 int. 1153
https://veroandreo.gitlab.io/

Hi all,

The release is done:
https://github.com/OSGeo/grass/releases/tag/8.4.0

Cronjob update PR:
https://github.com/OSGeo/grass-addons/pull/1147

Web news PR:
https://github.com/OSGeo/grass-website/pull/444

Please check them... and screenshot and binaries needed.

Markus

--
Markus Neteler, PhD
https://www.mundialis.de - company
https://grass.osgeo.org - FOSS
https://neteler.org - freelancing & blog

Thanks Markus!!!

On 27 Jul 2024, at 18:09, Markus Neteler via grass-dev <grass-dev@lists.osgeo.org> wrote:

Hi all,

The release is done:
https://github.com/OSGeo/grass/releases/tag/8.4.0

Cronjob update PR:
https://github.com/OSGeo/grass-addons/pull/1147

Web news PR:
https://github.com/OSGeo/grass-website/pull/444

Please check them... and screenshot and binaries needed.

Markus

--
Markus Neteler, PhD
https://www.mundialis.de - company
https://grass.osgeo.org - FOSS
https://neteler.org - freelancing & blog
_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
https://lists.osgeo.org/mailman/listinfo/grass-dev

Hi devs,

I just discovered that the tarball no longer contains the version in
the tarball "grass" subdir (expected "grass-8.4.0"):

tar tvfz ~/grass-8.4.0.tar.gz | head
drwxr-xr-x runner/docker 0 2024-07-27 15:34 grass/
-rwxr-xr-x runner/docker 458838 2024-07-27 15:32 grass/configure
-rw-r--r-- runner/docker 778 2024-07-27 15:32 grass/pyproject.toml
-rw-r--r-- runner/docker 4067 2024-07-27 15:32 grass/CITATION.cff
-rw-r--r-- runner/docker 436 2024-07-27 15:32 grass/.dockerignore
-rw-r--r-- runner/docker 3272 2024-07-27 15:32 grass/Vagrantfile
-rw-r--r-- runner/docker 1488 2024-07-27 15:32 grass/TODO
drwxr-xr-x runner/docker 0 2024-07-27 15:32 grass/display/
drwxr-xr-x runner/docker 0 2024-07-27 15:32 grass/display/d.rhumbline/
-rw-r--r-- runner/docker 53 2024-07-27 15:32
grass/display/d.rhumbline/local_proto.h

Why is that? Is it a feature?
It breaks e.g. the (Fedora) RPM packaging procedure...

Markus

Markus,

On 28 Jul 2024, at 11:54, Markus Neteler via grass-dev grass-dev@lists.osgeo.org wrote:

Hi devs,

I just discovered that the tarball no longer contains the version in
the tarball “grass” subdir (expected “grass-8.4.0”):

tar tvfz ~/grass-8.4.0.tar.gz | head
drwxr-xr-x runner/docker 0 2024-07-27 15:34 grass/
-rwxr-xr-x runner/docker 458838 2024-07-27 15:32 grass/configure
-rw-r–r-- runner/docker 778 2024-07-27 15:32 grass/pyproject.toml
-rw-r–r-- runner/docker 4067 2024-07-27 15:32 grass/CITATION.cff
-rw-r–r-- runner/docker 436 2024-07-27 15:32 grass/.dockerignore
-rw-r–r-- runner/docker 3272 2024-07-27 15:32 grass/Vagrantfile
-rw-r–r-- runner/docker 1488 2024-07-27 15:32 grass/TODO
drwxr-xr-x runner/docker 0 2024-07-27 15:32 grass/display/
drwxr-xr-x runner/docker 0 2024-07-27 15:32 grass/display/d.rhumbline/
-rw-r–r-- runner/docker 53 2024-07-27 15:32
grass/display/d.rhumbline/local_proto.h

Why is that? Is it a feature?
It breaks e.g. the (Fedora) RPM packaging procedure…

It is unfortunately not a feature, but a regression by PR 3437 [1]. Too bad that
one slipped. Breaks current procedure for MacPorts too, but is patchable.

Is a fix urgent or is a quick 8.4.1 release necessary?

Nicklas

[1] https://github.com/OSGeo/grass/pull/3437

On 7/28/24 11:54 AM, Markus Neteler via grass-dev wrote:

I just discovered that the tarball no longer contains the version in
the tarball "grass" subdir (expected "grass-8.4.0"):

tar tvfz ~/grass-8.4.0.tar.gz | head
drwxr-xr-x runner/docker 0 2024-07-27 15:34 grass/
[...]

Why is that? Is it a feature?
It breaks e.g. the (Fedora) RPM packaging procedure...

That tarball used for the Debian package from https://grass.osgeo.org/grass84/source/ includes the version in the toplevel directory:

  $ tar tavf ../grass-8.4.0.tar.gz | head -1
  drwxrwxr-x root/root 0 2024-07-27 13:27 grass-8.4.0/

Kind Regards,

Bas

--
  GPG Key ID: 4096R/6750F10AE88D4AF1
Fingerprint: 8182 DE41 7056 408D 6146 50D1 6750 F10A E88D 4AF1

On 28 Jul 2024, at 13:30, Sebastiaan Couwenberg via grass-dev <grass-dev@lists.osgeo.org> wrote:

On 7/28/24 11:54 AM, Markus Neteler via grass-dev wrote:

I just discovered that the tarball no longer contains the version in
the tarball "grass" subdir (expected "grass-8.4.0"):
tar tvfz ~/grass-8.4.0.tar.gz | head
drwxr-xr-x runner/docker 0 2024-07-27 15:34 grass/
[...]
Why is that? Is it a feature?
It breaks e.g. the (Fedora) RPM packaging procedure...

That tarball used for the Debian package from https://grass.osgeo.org/grass84/source/ includes the version in the toplevel directory:

$ tar tavf ../grass-8.4.0.tar.gz | head -1
drwxrwxr-x root/root 0 2024-07-27 13:27 grass-8.4.0/

The tarball from https://grass.osgeo.org/grass84/source/ is not the same as the one on Github, it also does not contain the "include/VERSION_GIT” file.

Best,
Nicklas

Kind Regards,

Bas

--
GPG Key ID: 4096R/6750F10AE88D4AF1
Fingerprint: 8182 DE41 7056 408D 6146 50D1 6750 F10A E88D 4AF1

_______________________________________________
grass-dev mailing list
grass-dev@lists.osgeo.org
https://lists.osgeo.org/mailman/listinfo/grass-dev

On 7/28/24 2:51 PM, Nicklas Larsson via grass-dev wrote:

On 28 Jul 2024, at 13:30, Sebastiaan Couwenberg via grass-dev <grass-dev@lists.osgeo.org> wrote:

On 7/28/24 11:54 AM, Markus Neteler via grass-dev wrote:

I just discovered that the tarball no longer contains the version in
the tarball "grass" subdir (expected "grass-8.4.0"):
tar tvfz ~/grass-8.4.0.tar.gz | head
drwxr-xr-x runner/docker 0 2024-07-27 15:34 grass/
[...]
Why is that? Is it a feature?
It breaks e.g. the (Fedora) RPM packaging procedure...

That tarball used for the Debian package from https://grass.osgeo.org/grass84/source/ includes the version in the toplevel directory:

$ tar tavf ../grass-8.4.0.tar.gz | head -1
drwxrwxr-x root/root 0 2024-07-27 13:27 grass-8.4.0/

The tarball from https://grass.osgeo.org/grass84/source/ is not the same as the one on Github, it also does not contain the "include/VERSION_GIT” file.

That's obvious.

Why that breaks the RPM packaging procedure is not, as it uses the tarballs from grass.osgeo.org too:

  https://src.fedoraproject.org/rpms/grass/blob/rawhide/f/grass.spec#_29

Kind Regards,

Bas

--
  GPG Key ID: 4096R/6750F10AE88D4AF1
Fingerprint: 8182 DE41 7056 408D 6146 50D1 6750 F10A E88D 4AF1

On Sun, Jul 28, 2024 at 3:07 PM Sebastiaan Couwenberg via grass-dev
<grass-dev@lists.osgeo.org> wrote:

On 7/28/24 2:51 PM, Nicklas Larsson via grass-dev wrote:
>> On 28 Jul 2024, at 13:30, Sebastiaan Couwenberg via grass-dev <grass-dev@lists.osgeo.org> wrote:
>>
>> On 7/28/24 11:54 AM, Markus Neteler via grass-dev wrote:
>>> I just discovered that the tarball no longer contains the version in
>>> the tarball "grass" subdir (expected "grass-8.4.0"):
>>> tar tvfz ~/grass-8.4.0.tar.gz | head
>>> drwxr-xr-x runner/docker 0 2024-07-27 15:34 grass/
>>> [...]
>>> Why is that? Is it a feature?
>>> It breaks e.g. the (Fedora) RPM packaging procedure...

I need to add that I had tried to use (this time) the tarball from GitHub.

>> That tarball used for the Debian package from https://grass.osgeo.org/grass84/source/ includes the version in the toplevel directory:
>>
>> $ tar tavf ../grass-8.4.0.tar.gz | head -1
>> drwxrwxr-x root/root 0 2024-07-27 13:27 grass-8.4.0/
>
> The tarball from https://grass.osgeo.org/grass84/source/ is not the same as the one on Github, it also does not contain the "include/VERSION_GIT” file.

That's obvious.

Not to me :slight_smile:

Why that breaks the RPM packaging procedure is not, as it uses the
tarballs from grass.osgeo.org too:

  https://src.fedoraproject.org/rpms/grass/blob/rawhide/f/grass.spec#_29

Yeah, using that one all goes fine.

So, we may ignore that but may consider to fix
https://github.com/OSGeo/grass/pull/3437

Best,
Markus