On Mon, Jun 28, 2010 at 08:24, Eric Wong <normalperson@xxxxxxxx> wrote: > Bert Wesarg <bert.wesarg@xxxxxxxxxxxxxx> wrote: >> On Tue, May 11, 2010 at 21:44, Paul Lindner <lindner@xxxxxxxxx> wrote: >> > This fixed it for my situation. Thanks! >> > On May 11, 2010, at 9:20 AM, Bert Wesarg wrote: >> >> Can you try this patch, it worked on my site: > > <snip> > >> Eric, have you picked this up, or should I re-send it as a propper patch? > > A proper patch + commit message would be easier, thanks Bert! I have invested some time to verify my patch before sending. And I think I will break something. The current patch changes the range argument for git rev-list from "r1^..r2" to "r1^!" & "r2", which should not be the same, because "r1^..r2" is equivalent to "^r1^" & "r2". This reads 'all commits reachable from r2 but not reachable from the parent of r1' where my conversion should read 'all reachable from r1 (excluding all of its parents) or reachable from r2'. But maybe the 'excluding all of its parents'-part quarantines that the semantic still holds? At least test 5 of t9151-svn-mergeinfo.sh does not pass, so my patch proved to be wrong. Bert > > -- > Eric Wong > -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html