Junio C Hamano <gitster@xxxxxxxxx> wrote: > This just makes me wonder if there is a practical reason why people > would not want this always enabled. I just looked at output from > > $ git log --grep='>>*From ' Missing '^' ? Auto-unescaping in mailsplit might throw off people on older versions of git if reserialized as mail, though. Maybe in a few years, we can consider it. Auto-escaping on the other hand, I think we start doing in --pretty=email by default soon/today. At least for lines matching the stricter is_from_line() check from mailsplit. > in the kernel repository, and I saw no cases where the committer > really wanted to preserve the leading one or more '>' on that line. > No, I didn't go through all of 150+ such commits, but I did check > the couple dozen of them from the recent history. > > Our history also have 5 instances of them, none of which should have > had the leading '>' if the committer were careful. Right, but I'm actually more worried about unescaped /^From / breaking further attempts to split. git log --grep='^From .*:.*:.* .*' ...finds 45 commits in Linux 4.6.1 which might cause incorrect splits. We have 5 of those ourselves. Technically it is backwards-incompatible, but I would rather leave an extra '>' in the commit than break a split. > > +>From the beginning, mbox should have been mboxrd > > Indeed ;-) Yes, I'm really wishing we did this 11 years ago :) -- 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