Hi Matthieu!

Thanks for taking the time to provide a good plan for the next steps.

> Hi all,
>
> Although none of us have much time to spend on DVC, the project is
> evolving, and starts IMO to be useable for several back-ends. It is
> probably time to make a bit more noise about it to attract users and
> developers (what I want to avoid is that other independant projects to
> manage distributed VCS in Emacs become mature before people know about
> DVC - this would be a waste of time for everybody).
>
> I think we can plan a release (0.5 or something, to mean "we're not
> version 0.0001, but still far from 1.0) in a few weeks. Regarding my
> availability, this probably means end of may or june.

That is definitely a good idea!

I also think a 0.5 release number reflects the current status.

> This may increase our motivation a bit too ;-).
>
> The things I'd like to be done before that :
>
> * Get rid of the "Xtla" name. This means:
>
>   - Move the home page away from the GNU Arch wiki. Probably a static
>     HTML page can do since we received almost no contribution from the
>     wiki itself.
>
>   - create a new Gna! project called DVC, with a dvc-dev mailing list.

Would you create that project?
If you want I can do it also.

> * Migrate away from GNU Arch. We didn't take the final decision about
>   which VCS to choose. Although I'm a bit angry against Canonical
>   regarding the way they abandonned Bazaar and threw away my work (I'm
>   OK with the fact that they progressively move away from baz to bzr,
>   but they announced a release of 1.5 for september, I worked hard to
>   prepare it, and it's still not there), bzr seems to me to be
>   technically the best. If we chose bzr, I suggest we wait for the 0.8
>   release, or a pre-release, before the migration.

I still didn't take the time to test bzr with my setup.

The important things for me are:
* publish on a dumb server (via ftp)
* bandwith optimized (I still use a 56k modem at home)
* storage optimized on the server side

tla works very well for that setup.

However I have seen some very nice new features in bzr. So I think it
is a good idea to switch to bzr. Hopefully the points mentioned above
work well with bzr also.

>   Or we can have a game on this : release 0.5 with bzr, then 0.6 with
>   hg, then 0.7 with Darcs, then ... ;-).

That would have the benefit of providing good support for all the
mentioned systems to ease our work... ;-)

> * As a result of the above, get rid of tla|baz build dependancies of
>   DVC. DVC should install and run on a computer without tla and baz
>   installed, and as much as possible, should continue to work after a
>   "rm tla*.el baz*.el".
>
> * Work a bit more on the revlist mode, but it's already in good shape
>   after my commits of today. 'would be cool to have this revlist mode
>   for Mercurial too.

I will try to reserve a time slot to implement that feature for hg and
cogito in the next weeks. To have it before the release.

> * Would be cool to have a more-than-draft implementation of bookmarks.
>   Stephan, I had a look at your code, it's only the beginning but I
>   like it !

Thanks. The next function I'd like to add is killing and yanking of bookmarks
items to have basic editing capabilities. Then I will add saving/loading and try
to migrate my tla bookmarks to the new system.

> * A merge operator should be doable from the revision list quite
>   easily.
>
> Just let me know what you think about the above points, and I think
> the next steps are:
>
> * Implement the above
>
> * A "stabilization/test" period of a few weeks
>
> * Release

Sounds o.k.

We now have at least basic support (view log, view diff, commit) for tla/baz, 
bzr, hg, cg.
I'd like to add that basic support at least to darcs before our first release.
However darcs seems a bit different than the other systems we support. But I 
didn't
take a closer look up to now.


Stefan.

_______________________________________________
Xtla-el-dev mailing list
[email protected]
https://mail.gna.org/listinfo/xtla-el-dev

Reply via email to