Re: [Simulavr-devel] Simulavrxx fork started

[Petr Hluzín]

Hello

(Replying to 3 mails at once to not increase the heat by excessive traffic.)

On 19 March 2010 23:25, Joerg Wunsch <address@hidden> wrote:
> I'm an administrator
> here, yes, but that doesn't mean I would be patching a CVS tree that
> is not even used by the developers at all.

Well, someone has to. No one with commit access did commit the
existing patches (to the CVS repo).
Since all work is done against the CVS repo (me) or against the git
repository (Onno and some other guy) which already contains the CVS
repository then the effect is the CVS repository is used.

> Also, there's a knob on the Savannah web interface people who want to
> join a project could press

I already tried it and it did something which did not look like it is working.
I believe it persisted after logging in, maybe cache.
I assumed some other mechanism is used and I gave up.

On 19 March 2010 23:41, Weddington, Eric <address@hidden> wrote:
> There seems to be some of you more interested in using git, then in actual 
> development. Who really cares about the type of revision control system that 
> is being used? If git is required for this project then we would use it. As 
> it stands, this project is so small that either CVS or SVN is sufficient for 
> the project.

Regarding version control tool: I prefer Subversion and I know CVS well.
Distributed tools are more difficult (unless effort was made to hide
it) and have little benefit in our case.
And git has geeky reputation - which scares me.
I am willing (to try) to use any versioning tool, though.

On 19 March 2010 23:50, Joerg Wunsch <address@hidden> wrote:
> But I've constantly got the impression that those who are doing the
> development work were simply waiting for a mystical someone "above
> them" to eventually debate their decisions, and somehow integrating
> all that, or whatever.

Since I do not have a commit access I expect to see either discussion
over a patch or a mail like "the patch looks sane and no comments
raised in 7 days, commited in".
If I had a commit access I would want a guidelines when/what to commit.
(I do not handle rejections well - I am willing to read/polish things
more to avoid that.)

Summary: I just want patches to be merged to every repository possible
and I prefer to obtain changes from just one repository.
And the patches which affect users immediately (e.g.
non-compile-ability of the legacy simulavr) commited to be where users
expect to find them.
I do not care who applies them. Both git and CVS are acceptable to me.
(And Onno's repository seems alive while simulavr does not.)

-- 
Petr Hluzin