Re: [PATCH] git am/mailinfo: Don't look at in-body headers when rebasing

[Philip Hofstetter <phofstetter@sensational.ch>]

Hi,

On Wed, Nov 18, 2009 at 11:45 PM, Lukas Sandström <luksan@gmail.com> wrote:

> The actual change is that mailinfo doesn't look for in-body headers
> at all if --no-inbody-headers is passed. git-am now passes this option
> to mailinfo when rebasing.

after all the earlier discussion and a lot of thinking, I have to say,
that IMHO, this is the best option as it doesn't rely on heuristics
and now that you chose a descriptive command line switch, even the
small problem of "why exactly is this switch here?" seems to go away.

As I have no experience in git's codebase at all, I'll leave the
commenting on the patch itself to the people with clue, but
conceptionally, this feels much better than the method 1

> This won't handle the case when a "bad" patch is passed to git-am from
> somewhere else than git rebase.

of course, that leaves the question what "somewhere else" can contain.
If it's just manual calls to git-am, this is a non-issue as it's
easily fixed by the caller. If it's being called from other
higher-level operations though, you might run into the same issue
again.

Here too, I can't really provide any meaningful input though as I just
don't know well enough what really makes git tick.

Just my two cents :-)

Philip