Re: [gnome-flashback] Releasing gnome-panel 3.8.1, 3.10 for trusty ?, the state of master, RFH for code review



Hi,

as you may know, gnome-panel master has been a huge grab-bag of stuff
and fluff since shortly after the 3.8.0 release, containing some good
(and important) bug fixes, dependency bumps, ABI breaks, code
modernization patches, and translation updates.

For this reason it's not possible to just cut a 3.8.1 release from
master as some have suggested - we need to stay compatible to the Gnome
3.8 release, the ABI we shipped with 3.8.0, and the dependency
requirements we published at that time.

Are you serious? Do you know how many distributions are updated to GNOME Panel 3.8? I found that GNOME Panel 3.8 is available in Ubuntu 14.04, Debian Jessie and Debian Sid. For example all of them have GTK+ 3.10 or newer. There is no distribution that use GNOME Panel 3.8 with pure GNOME 3.8.

As GNOME Panel 3.8 is used with GTK+ 3.10 that means that we could make 3.8.1 release even with multi-screen code removal and it would change nothing. Nothing!
 
I have now finally finished reviewing all the commits that went into the
master branch more or less unsorted - only a tiny fraction has ever
received proper code review.

I created a CSV table, attached, which lists which commit should go into
each branch. Feedback on this is very welcome.

Credit where credit is due, the majority of this work has been created
by Alberts Muskupavels (at least as far as the patch authorship
information is correct).

The next step is to cut a 3.8.1 release (from the gnome-3-8 branch
obviously).
I have pushed the relevant code commits to that branch. Translation
commits will follow.

I see... You pushed many commits that you previously called as mess in master.

I have resurrected a saucy system with mixed Gnome 3.8 / Gnome 3.10
components, and the branch builds and works there, even loading applets
from the old gnome-applets 3.6.5 package.
(TODO: test jhbuild of 3.8 also)

Any reason to test with ubuntu version that will be obsolete in less than 10 days? Saucy is now end of life since July 17.
 
Please test building and running the gnome-3-8 branch, especially on
Gnome 3.8 and 3.10 installations and in jhbuild.

You want so much that 3.8 branch. I guess you are only one who needs it. Why someone would do it for you? Master has been tested and reported that works. Why someone would want to test it again?

GNOME Panel 3.8 is not used with pure GNOME 3.8. That branch should be deleted! Make 3.9.0 release if you think that changes in master are way to big for minor release.

@lanoxx: I have marked your .ui file modernization patches as being OK
to go into 3.8 - does that seem sensible to you ?

---

QUESTION to Ubuntu downstreams: You were the only ones that seemed
interested in getting a gnome-panel 3.10 release done. Is anybody still
interested in creating a backport for this  in trusty, and seeing it
through to completions ?

If not, I'm inclined to skip 3.10 completely. Please answer until
Saturday at the latest.
---

Ubuntu is waiting for bug fix release, but you can not do even that. How are you planing to make 3.12 or 3.14 release?
You unnecessarily spend time to re-commit existing commits to 3.8 branch. It is 18 days now since that and again progress has returned to 0.
 
I'm also looking for help to conduct a proper code review round on all
the work in master. Please check the CSV file, if the split of patches
into feature branches makes sense. Then I will publish proper RFCs for
each branch on this ML¹, and am hoping the other devs / downstreams will
help me to review this work.

¹ (As I had asked the author to do, who couldn't be bothered.)

I guess that this is one of your biggest problems. No one has helped you to review something (you have asked it in past too). Don't you think that it sign that this time will be no different?

Or you will simply use this as excuse why you have not still made a new release?

And as always, Gnome Flashback needs some love, resurrecting the missing
gnome-settings-daemon parts. (I know ubuntu downstream has created a
workaround, in the form of a g-s-d fork. Upstream we are still in bad
shape on this front, and AFAICT so is GF in Debian.)

Any progress? We would like to test it. Oh, wait, you are maintainer - you don't need to do anything, someone will do it for you. I am pretty sure that you have nothing to show. Am I wrong? I hope so.

This time I am not going to apologize for anything you find offensive. This has to stop. We need real progress! It is 10 months since 3.8 release.

No one wants to see this bullshit here. You have proven that you are not able to maintain GNOME Panel / GNOME Flashback. I think that no progress in 10 months is fairly good reason to say this.

Start to maintain this project. If you can not do that then I am ready to take over it.

--
Alberts Muktupāvels


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]