On 4/1/07, Erik Huelsmann <ehuels@gmail.com> wrote:
> In all of the discussion about having a patch format which allows for
> exchange of tree modifications to better fit Subversion, people come
> up with a lot of reasons why the proposal on the table (there have
> been more than the one now under discussion). One of those reasons is
> that we should work with other projects to invent a cross-VC-system
> format.
>
> When I see comments like that, I have a number of questions. Let me
> start out that I think the goal in itself is admireable though.
>
> 1) Do any of the other VC systems have a patch format which is able to
> represent tree changes?
> 2) Do any of the other VC systems have a patch format which is suited
> for code-review as well as changeset-exchange (as we're now demanding
> from 'our' new format)
> 3) Do any of the other systems even try to be exchangeable with other ones?
>
> If the answer to any of the above is 'yes', then shouldn't we
> investigate how they do it?
>
> And last but definitely not least, I'd like to add that solving the
> problem for Subversion alone seems hard enough, as things stand. Also,
> being able to exchange patches between VC systems has only marginally
> more value in the short term than being able to exchange correct tree
> mods at all. I guess this paragraph is a plea to start out small, then
> when we succeeded we can take the next step: integrate the world!
>
+1, I almost agree with you. I think for step we should try to use xml
representation of commit operation, that Subversion was using on the
beginning. And make patch command work as merge internally.
And only after that we can move on next step, to integrate the world.
--
Ivan Zhakov
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sun Apr 1 13:46:31 2007