Re: A Question from a Hopeful Future Contributor

From: Junio C Hamano <gitster@pobox.com>
Date: Jul 28, 2025, 15:02 UTC
Message-ID: <xmqq5xfcz7d6.fsf@gitster.g>
In-Reply-To: <CALnO6CC+Tn5xFYcHwH-M7kLNt2+gut7Okx7fCuWRdOsQHoDWFw@mail.gmail.com>

"D. Ben Knoble" <ben.knoble@gmail.com> writes:

Show 14 quoted lines

>> >> * "git status" on intent-to-add index entries (say "I" in the first
>> >>   column instead of "A" for short status, add "(needs 'git add')" at the
>> >>   end of "new file: $path " in long status).
>> >
>> > I am interpreting this todo message as meaning that the following behavior
>> > should be implemented:
>>
>> I think that was done long time ago.  The entry may have been a wish-item
>> in April 2011, but not anymore.
>
> At least in 2.48.1 and in git version 2.50.0.rc0.48.g74dbe4346c, this
> behavior isn't present. So I wouldn't call it "done"—maybe there was a
> discussion that it shouldn't be done, though? I didn't search the
> list.

I didn't either, so this response may not be adding much value, but I think the intent of the above entry is so that the output allows users to tell between already added entries and about to be added entries. Perhaps in an ancient version of "git status" did not give enough information to allow us to tell the status of the index entry for the path and the status of the working tree file for the path separately, or something?

In any case, we do not need to add anything to achieve that today.

    $ >P
    $ cp P Q
    $ git add -N P
    $ git add Q

The user expresses an intent to add P and actually adds Q.

    $ git status -suno
     A P
    A  Q

The first column of the output is about how the path in the index compares with the HEAD, and the second column is about how working tree file compares with the index. A file that hasn't been added (e.g., untracked cruft) is not even known to the index so the latter is _usually_ only for tracked files. The fact that P appears as "_A" is a sign enough that its contents haven't been added to the index, but only that the index has been made aware of the presence of the path, which is exactly what intent-to-add is about.

If you further add P, this time for real, the status would become "A_" (i.e. to the index, new contents added relative to HEAD; to the working tree there is no change relative to the index), but the intent-to-add bit is gone once the index has (a version of) contents for the path P.

Normal output without "-s" is even easier to reason about.

    $ git status -uno
    On branch foo
    Changes to be committed:
      (use "git restore --staged <file>..." to unstage)
            new file:   Q

    Changes not staged for commit:
      (use "git add <file>..." to update what will be committed)
      (use "git restore <file>..." to discard changes in working directory)
            new file:   P

P is _known_ to the system but its contents is not added at all, so it is "not staged for commit", and known as a "new" file.