On Tue, Mar 4, 2008 at 11:32 AM, David Glasser <glasser_at_davidglasser.net> wrote:
> On Tue, Mar 4, 2008 at 8:29 AM, Erik Huelsmann <ehuels_at_gmail.com> wrote:
> > On 3/4/08, Paul Burba <ptburba_at_gmail.com> wrote:
> > > Oops, forgot to include dev on my reply this morning...
> > >
> > > On Mon, Mar 3, 2008 at 9:46 PM, Mark Phippard <markphip_at_gmail.com> wrote:
> > > > On Mon, Mar 3, 2008 at 8:42 PM, Paul Burba <ptburba_at_gmail.com> wrote:
> > > >
> > > > > The problem is that the whole tree rooted at 'A_COPY' gets locked at
> > > > > the start of the merge. Then the svn_wc_diff_callbacks2_t callback
> > > > > merge_dir_added() doesn't find 'A_COPY/d' in the admin access (since
> > > > > that *is* case sensitive) and not finding it attempts to open a new
> > > > > access, but in doing so runs into the existing 'A_COPY/D'. I haven't
> > > > > thought much on how to solve this; I want to solve the simpler file
> > > > > case first.
> > > >
> > > > I think it is relevant to know if this behavior is any different from 1.4.
> > >
> > > This behavior exists in 1.4.6 and has AFAICT has been with us for a
> > > very long time. I didn't mean to imply it was a 1.5 problem.
> > It's exactly the same problem as the 'Case-only rename fails' problem.
> > We don't have a good way of detecting case-sensitivity of a filesystem
> > (yet)... Having that would at least enable us to resolve this problem.
> I thought we were mostly able to avoid this by making the reporter's
> editor driver report all deletes before adds. Why doesn't that help
> us here?
Paul touched on that. The issue is that merge is like an edit of the
working copy. On Windows and OSX, you cannot do a case only-rename of
a file. You can receive an update because it is altering the working
copy as it goes.
To unsubscribe, e-mail: dev-unsubscribe_at_subversion.tigris.org
For additional commands, e-mail: dev-help_at_subversion.tigris.org
Received on 2008-03-04 18:06:31 CET