[gnome-flashback] State of Gnome Flashback

[Philipp Kaluza <floss ghostroute eu>]

Hi Alberts, hi team,

this answer is a few days late, but I wanted to take the time to (again)
try to clearly communicate where the project is at in my opinion, and
open the floor to discussions.

Am 22.05.2014 01:28, schrieb Alberts Muktupāvels:
> Hi!
>
> Dear maintainers are you still interested in gnome flashback project
> or not? How about some new releases for metacity and gnome-panel?
>
> Phillip, half year ago you wanted to release new version. What
> happened? Have I offended you?

I must admit that at times I felt you (Alberts) were missing the point
on purpose. But I know that this would be an unfair ad-hominem
characterization, and that our problem is rooted more deeply in our
flawed communication culture. It does not help, of course, that you
(still) seem unfamiliar with the communication customs in the rest of
the Gnome project.

At this point I have to concede, that the "Gnome Flashback" initiative,
as originally envisioned, is well on its path to failure. We have
consistently failed attract manpower to work on the most burning
problems, which are in my only somewhat humble opinion the regression
caused by the disappearing functionality in gnome-settings-daemon. This
makes Gnome Flashback upstream be in a pretty terrible state, although
some distributors (like Ubuntu) have managed to work around this - much
appreciated BTW - using a mix-n-match strategy, at the cost of keeping
other Gnome components in a stale version.

I am not free of blame here: like I expressed before, the time I can
devote to Gnome Flashback is mostly proportional to how much work around
terminal server projects comes my way (which has not turned out like I
expected), and is otherwise pretty limited. I _am_ sorry about that, but
that's where I come from at the moment.

The only one that has been really active coding-wise is you, Alberts.
While you and I may disagree on a lot of the details, your efforts are
highly appreciated. But while a modernization of the gnome-panel code
base is certainly a good thing for this project's long-term health, and
yet another, more focused effort to port metacity to GTK+ 3 might well
help revive that window manager, the problem that our "umbrella-project"
Gnome Flashback is in bad shape remains, and we still don't really have
a good answer to that.

Also, Dmitry, your efforts are appreciated as well, especially around
triage.

Repeatedly approaching the Mate project about cooperating also has not
yielded any results.

> metacity.
Like Dmitry said, metacity never made it into the fold of the Gnome
Flashback project. It's awesome you're working on this. Please consider
taking over maintainership, and potentially just ask for it.

> gnome-panel.
> I think master can be released as new release. Some patches from there
> are already used in ubuntu, debian and maybe in other distros as well.
Yeah, but what version number would that have ? With what Gnome release
should we target and test against ? The relatively brief window where
Gnome 3.10 was a consensus in the .deb world is now over as well.

> In case you don't want to continue to maintain those packages maybe
> you are ready to give those rights to others?
I am only maintainer of one package in the Gnome world, and that is
Gnome panel (which, for reasons, builds and ships the Flashback support
files as well). If people feel that I am the limiting factor, I am OK
with stepping aside, but I'd like to hear a compelling plan and
commitment on what will happen not only to gnome-panel, but also Gnome
Flashback - the whole thing including dependencies.

I hereby open the floor for discussions, and am hoping for a slightly
better response than my February 21st mail. :)

Cheers everybody, keep fighting the good fight,
  Philipp

-- 
Philipp Kaluza
Ghostroute IT Consulting