| Re: [Sawfish] How to send patches |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/sawfish Archives
]
Just now, Christopher Roy Bratusek wrote:
> Generally the way you send patches in does not matter (unless it's
> not in unified format). Git shows me as author as I applied the
> patch.
>
> Even if 'git apply' would have worked it would have shown me, unless
> I'm wrong.
No -- git has two people for every commit: the author (who did the
code editing), and the commit (the person who committed it to whatever
repo). But looking at some random commit messages, it looks like
you've edited at least those messages (eg, I see "fix fsf address" and
"fixing the FSF address in rep-gtk", and in the git repo there's "fix
address of FSF").
For cases like touching up some minor detail in a commit, like
changing its message, you should apply the commit, do the touchup, and
then use "--ammend" after the touchup, which will replace the commit
by a new one, and should preserve the commit author. Another
alternative is to mention the original author in the commit message,
but in cases of heavier tweaking it's best to just add a second
commit.
> Either way I seem to remember one could change the author of a
> commit afterwards?
That's possible too (with an --ammend), but the above should work fine
if the goal is to preserve the author...
An hour and a half ago, Michal Maruska wrote:
> On Fri, Jul 6, 2012 at 4:37 PM, Christopher Roy Bratusek
> > Even if 'git apply'
>
> I guess you should use "git am" then. (which works with patches
> prepared with "git format-patch").
No, "git am" basically extracts the patches and passes them to "git
apply"... (It's kind of implicit in the man page -- it talks about
some flags that are passed on to "git apply".)
--
((lambda (x) (x x)) (lambda (x) (x x))) Eli Barzilay:
http://barzilay.org/ Maze is Life!
--
Sawfish ML