On Thu, 03 Aug 2006, Garrett Rooney wrote:
> On 8/3/06, Daniel Rall <dlr@collab.net> wrote:
> >On Thu, 03 Aug 2006, Malcolm Rowe wrote:
> >
> >> On Thu, Aug 03, 2006 at 10:25:55AM -0700, dlr@tigris.org wrote:
> >> > New Revision: 20951
> >> >
> >> > Modified:
> >> > branches/1.4.x/subversion/include/svn_wc.h
> >> > trunk/subversion/include/svn_wc.h
...
> So if I tried to merge the portion of the commit from trunk into
> 1.4.x, the merge tracking code would have no way to determine that it
> shouldn't be done, and a conflict would likely occur, right? So we
> should just tell people not to do stuff like that if they want merge
> tracking to be any good ;-)
Well, in this case you'd probably see 'G', and the hunk would apply
successfully. But in an uglier case, where branches had drifted
further apart, you could see a merge conflict, yes. This seems only
slightly worse than the typically case for merging when two branches
have drifted apart, no?
- application/pgp-signature attachment: stored
Received on Thu Aug 3 23:04:20 2006