[Kde-scm-interest] Project release tags on Gitorious

[Eike Hein]

On 11/30/2009 12:23 AM, Chani wrote:
> the idea was to minimize the transition period: right now, if you move to git 
> then anyone who wants to contribute to it has to know git, and the techbase 
> documentation isn't where I'd like it to be at.

We've got a consensus amongst the active Konversation
developers, that was the prerequisite from our end as
well - "no developer left behind." As for outside con-
tributors, the Amarok guys tell me Gitorious has sig-
nificantly lowered the barrier of entry for contribu-
tors to Amarok and they're getting much more patch
traffic these days.

We're also being cheered on e.g. by our Debian packa-
ger MoDaX.


> main modules of KDE *will* stay in svn until we can provide a smooth 
> transition for them.

Yep, sure. The main modules have a much wider contri-
butor base, and in transitioning them it needs to be
ensured that nobody's work gets interrupted too much.

This isn't an issue in Konvi's case, due to the afore-
mentioned consensus we have in the team.


> projects that are on the fringe... well, in the end we can't force you to 
> stay, but please *please* consider all the consequences of a premature move to 
> git.

FWIW: We don't want to force our move, either. The
community is much more important to us than the SCM
we use.


> important technical consequences:
> -you won't have a log of who pushed what until whenever that TODO item gets 
> done. sure, this isn't *likely* to be important... but you're not likely to 
> need your backups, either, are you? :)

I've read the accountability thread, and it's defini-
tely an important item for the KDE-wide transition.

I think Konvi can do without for now, however. We're
not so big and have a pretty tight handle on what co-
de goes in. And it helps that IRC client devs tend to
naturally hang out on IRC, so there's constant commu-
nication of who's doing what.


> -you won't be part of commit-filter, you won't be able to use BUG: or CCMAIL: 
> or any of those hooks until they're done.

This is a bummer, but we knew that going in, yes.

BTW, something I forgot to mention: Eli 'argonel'
MacKenzie whom you might know from #kde-git and who
has been present at the last few meetings about the
git transition is a core developer at Konversation,
so we've been trying to involve ourselves in the
transition work already.

I've also done things like work a tiny bit on the
CIA post-commit hook for SVN in the past, so I might
be able to make myself useful on these kinds of
things somehow.


> -if you've got a docbook folder, remember to separate it out and leave it 
> somewhere in svn for the translators.

Our handbook is in extragear/network/docs right now; I
had figured we'd leave it there, and whenever we announce
a string and documention freeze on kde-i18n-doc we mention
where the docs are. Sound good? Albert?


> there are also social consequences - I think it'll create more pressure to 
> move to git before it's ready, and there'll be less pressure to actually 
> resolve that todolist. it also feels less "KDE" right now - we don't have a 
> kde page on gitorious or anything... or am I putting the cart before the horse 
> on that one? :)

Yes, the social consequences are perhaps the most worry-
ing.

However, my take on this is this: Konversation is one of
the oldest and most stable KDE Extragear projects around.
Konversation has been around for 6 1/2 years now, and our
mission has always been to be a KDE app and a champion
for the KDE platform first and foremost. Having more
"good old KDE stuff" on Gitorious can perhaps help in ma-
king sure we successfully import our culture there. I
would hate to see this as cutting any ties with KDE, I
see it more as extending KDE to a new place. Our spiri-
tual home will always be KDE, not Gitorious ;-).


> so... if you think the move is worth it in spite of that, I won't try to stop 
> you. however, I really think it would be better for all of us if you put 
> pressure on (or volunteered to help with) fixing the technical issues *first*.


> all I ask is that you consider all the consequences before making a
decision.

Well, it's definitely not like "we need to be on Gito-
rious by Monday!" We have a strong interest in not drag-
ging things out until the end of time, but we also don't
want to jump in over-hurriedly. This why I opened this
thread on the tagging problem: We started to look into
the consequences of moving now, and that was one of the
problems we identified and that we wanted to talk about
first.


-- 
Best regards,
Eike Hein