Re: Improvements to ‘debian/watch’ for fetching from VCS
Ben Finney wrote:
> Raphael Geissert <firstname.lastname@example.org> writes:
>> How does bzr recognise whether it should use http or not?
> I don't understand the question. I'll try to answer what might be the
> question; if this doesn't answer you, please re-phrase.
> Bazaar can use HTTP as a transport for file-level operations; that's
> what a ‘http://’; URL specifies. The server doesn't need to be anything
> except a static HTTP server.
> Bazaar can alternatively use HTTP as a transport for Bazaar “smart
> server” commands, requiring a Bazaar server to be answering at the
> other end. That's what a ‘bzr+http://’; URL specifies.
Er, right. I think I was too sleepy when I asked that.
Probably the best way to handle VCS' without making it too vcs-oriented
would be to introduce 'handlers' so that:
would do the right thing.
Actually, am thinking about the following handlers for svn:
svn-ls (a.k.a just 'svn'): the equiv of svn ls on the url
svn-info: svn info url
the former causing uscan to try to match the regex on every line, and the
latter only on the revision (if that's of any use of course).
Or what else is needed?