Re: svn commit: rev 278 - trunk/subversion/include trunk/subversion/libsvn_ra_local trunk/subversion/libsvn_client trunk/subversion/libsvn_ra_dav

On Mon, Oct 22, 2001 at 10:56:54AM -0500, kfogel@collab.net wrote:
> kevin@tigris.org writes:
> > Log:
> > Make nonrecursive checkouts work. (I know this was low priority, but it was
> > easy once I got nonrecursive updates working).
>
> Not just a "low priority" -- I thought we'd decided not to do it,
> because we don't know how to handle a nonrecursive (shallow) working
> copy.
>
> How does the working copy behave after one of these checkouts, when
> you run update?

Well of course it fails if there are updates to a deeper part of the tree.
with can't open directory './some/deep/dir.'

I can of course disable this functionality, but it is interesting interesting
to think about for now. If we forbid addition of directories to a wc checked
out non-recursively, it becomes easy to make updates work, simply by storing
a flag in the .svn dir, and passing recurse = FALSE to all future updates.

I don't know how useful that behaviour is, however. And I haven't thought/don't
think there is a way to handle it in the general case. The server will always
think that revision N contained some subdirectories, and pass them down to the
client. Unless either the client or server contained some logic that said
Only update this subdir if it was created after revision N, I don't see how it
could work.

Open to other thoughts on this though...

-- 
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Kevin Pilch-Bisson                    http://www.pilch-bisson.net
     "Historically speaking, the presences of wheels in Unix
     has never precluded their reinvention." - Larry Wall
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~