Re: [PATCH] Fix a regression from 1.4 in trunk and branch

On Jan 30, 2008 2:03 PM, Augie Fackler <durin42_at_gmail.com> wrote:
>
> On Jan 30, 2008, at 3:17 PM, Augie Fackler wrote:
>
> >
> > On Jan 30, 2008, at 3:12 PM, C. Michael Pilato wrote:
> >
> >> Augie Fackler wrote:
> >>> Ah, that information comes from the RA layer (I was just trying to
> >>> figure this out) - I'll go inspect them all and make sure they
> >>> DTRT for this case. Should I keep the original part of the patch
> >>> and include it with the work on the RA layer?
> >>
> >> We shouldn't need the original patch, but I won't complain if we
> >> include that code. Should make us more resilient in the case of
> >> some other unseen programmer error.
> >
> > Acutally, in looking more - it's a bug in libsvn_ra_serf anyway. If
> > I switch my RA layer to neon and re-run the checkout without my
> > patch, it works. I'm working on a fix for serf now.
>
> This is the simplest fix I can come up with for serf:
> Index: subversion/libsvn_ra_serf/serf.c
> ===================================================================
> --- subversion/libsvn_ra_serf/serf.c (revision 29079)
> +++ subversion/libsvn_ra_serf/serf.c (working copy)
> @@ -992,6 +992,7 @@
> }
>
> *url = session->repos_root_str;
> + *url = svn_path_canonicalize(*url, pool);
> return SVN_NO_ERROR;
> }
>
> Anything else will require a little fiddling around inside serf to
> make some of the code in update.c work correctly.

So I guess the alternative here is to make session->repos_root_str be
set correctly in svn_ra_serf__discover_root, but this might cause
weirdness in the other code that looks at that field in update.c?

Lieven or Justin, any opinions?

--dave

-- 
David Glasser | glasser@davidglasser.net | http://www.davidglasser.net/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: dev-help_at_subversion.tigris.org