----- Original Message ----- > From: "Junio C Hamano" <gits...@pobox.com> > Sent: Sunday, September 9, 2012 12:40:37 AM > Subject: [PATCH 0/2] Teaching -Xours/-Xtheirs to binary ll-merge driver > > The part that grants Stephen's wish is unchanged from the earlier > "perhaps like this" patch, but this time with a bit of documentation > and test. > > A more important change between the two actually is [PATCH 2/2]. > When a "binary" synthetic attribute is given to a path, we used to > (1) disable textual diff and (2) disable CR/LF conversion, but it is > also sane to disable the textual merge for a path marked as > "binary", and setting the "-merge" attribute to summon the "binary" > ll-merge driver is the way to do so. > > Junio C Hamano (2): > merge: teach -Xours/-Xtheirs to binary ll-merge driver > attr: "binary" attribute should choose built-in "binary" merge > driver > > Documentation/gitattributes.txt | 2 +- > Documentation/merge-strategies.txt | 3 ++- > attr.c | 2 +- > ll-merge.c | 25 ++++++++++++++++++++----- > t/t6037-merge-ours-theirs.sh | 14 +++++++++++++- > 5 files changed, 37 insertions(+), 9 deletions(-)
Thanks for this Junio. After figuring out how to make the corporate email server NOT munge the patches, I was able to apply these with no problem. Then $ git merge -Xtheirs maint mostly does what I want, but conflicting files deleted on the incoming branch are still marked unmerged/"deleted by them" (binary or not). I'm not sure what the best (most common?) resolution is for that. In my case I would be happy for the -Xtheirs to also delete the files, but that might not always be the right answer (but then again -Xtheirs is a pretty heavy hammer to begin with). Thanks, Stephen -- To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
