On Fri, 2002-12-13 at 09:53, mark benedetto king wrote:
> > The repository for a URL varies depending on how you access it; you
> > might use file:, http:, and svn: to access the same repository. I think
> > we should avoid storing any data in one repository about another
> > repository, with the exception of the guid.
> I think that would make repository synchronization difficult.
> The mapping between UUID and URL has to be stored somewhere,
> so that a process can, based on the repos UUID of a node in the database,
> go look for updates.
Granted; and if we had an agreed-on design document which involved
someone looking at a guid and going out and fetching updates, I'd say we
would need a mapping like that. (Maybe on the client, maybe on the
server; if on the server, maybe in properties, maybe in a BDB table.)
There are things you can do with a guid which don't involve that
operation. You can verify that an svnadmin switch --relocate isn't
crossing repositories, you can calculate which changesets to ignore in a
merge operation, and you can do repository synchronization when you've
been given the necessary URLs.
Perhaps I've gotten lost as to your plans. But I really don't want to
see people adding data structures to the repository before we, as a
project, have a good sense of what they're needed for.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Fri Dec 13 16:53:50 2002