From: Daniel Barkalow <barkalow@iabervon.org>
Date: Mon, 16 Mar 2009 23:11:03 GMT
Subject: Re: What's cooking in git.git (Mar 2009, #04; Sat, 14)
Message-ID: <alpine.LNX.1.00.0903161834530.19665@iabervon.org>
In-Reply-To: <7vr60z8fkl.fsf@gitster.siamese.dyndns.org>

On Sat, 14 Mar 2009, Junio C Hamano wrote:

> * db/foreign-scm (Sun Jan 11 15:12:10 2009 -0500) 3 commits
>  - Support fetching from foreign VCSes
>  - Add specification of git-vcs helpers
>  - Add "vcs" config option in remotes
> 
> The "spec" did not seem quite well cooked yet, but in the longer term I
> think something like this to allow interoperating with other SCMs as if
> the other end is a native git repository is a very worthy goal.

I recently figured out a way to alter the transport API slightly to make 
it possible to attach foreign VCSes there. This amounts to allowing 
get_refs_list() to leave the sha1s null, and allowing fetch() to set them. 
This just requires a copy of a ref list in the clone path, and a check to 
make null (for not having a ref) not match null (for not knowing the hash 
of the ref) in the fetch logic.

I think I will end up changing the spec, mostly to allow a single 
non-trivial call to the helper to provide both the list of refs and the 
fast-import stream; this matches how the pack protocol works (except, of 
course, that the initial list doesn't have hashes, and the output goes to 
fast-import instead of unpack-objects). I'm going to rebase the series at 
some point, but I've been focused more on the more core transport and 
remote stuff now that I think I can go through that.

	-Daniel
*This .sig left intentionally blank*