On Mon, Feb 8, 2010 at 7:39 PM, Jeff King <peff@peff.net> wrote:
quoted text
> On Mon, Feb 08, 2010 at 02:12:41PM -0500, Chris Packham wrote:
>
>> > $ git add -u nonexistent-file
>> > $ echo $?
>> > 0
>> [...]
>> It looks like in the case you've highlighted git is ignoring the extra
>> non-option parameters on the command line. I'll let other people argue
>> whether this is by design or omission.
>
> It's not ignoring the extra parameters. They limit the scope of the
> operation. So:
>
>  $ git init
>  $ touch file && mkdir subdir && touch subdir/file
>  $ git add . && git commit -m one
>  $ echo changes >file && echo changes >subdir/file
>  $ git add -u subdir
>  $ git status
>  # On branch master
>  # Changes to be committed:
>  #       modified:   subdir/file
>  #
>  # Changed but not updated:
>  #       modified:   file
>  #

Yep my bad. I tried the non-existent case but not the "normal" case.
Re reading the man page it makes sense it just happens that the
<filepattern> part scrolls off the top when I get to the -u part.

quoted text
> That being said, you noticed that the regular add case notes unused
> pathspecs on the command line:
>
>  $ git add bogus
>  fatal: pathspec 'bogus' did not match any files
>
> We could probably do the same here.

I think so. By not having this error it led me to think "OK it just
throws away the pathspec". Gabor rightly thought that it does use it
so shouldn't it be giving an error.

If I get brave enough I could attempt a patch but I wouldn't let that
dissuade anyone that actually knows what they're doing from jumping in
with a patch.
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html