Re: Confusing error message in rebase when commit becomes empty

From: Phil Hord <phil.hord@gmail.com>
Date: Jun 11, 2014, 19:04 UTC
Message-ID: <CABURp0q1c+75VzMfttDWVXhTFma3g=Tj938J_9i9xhG1yuMpdQ@mail.gmail.com>
In-Reply-To: <alpine.DEB.2.00.1406111854180.21431@ds9.cixit.se>

On Wed, Jun 11, 2014 at 1:57 PM, Peter Krefting <peter@softwolves.pp.se> wrote:

Show 14 quoted lines

> Phil Hord:
>
>
>> What does it mean when you say it worked as expected?  Did it leave
>> the empty commit, omit the empty commit, or leave some un-squashed
>> commit?
>
>
> Actually, it did not work as expected I noted afterward, it just dropped the
> reversion commit, and did not squash the next commit into it as I had asked,
> so from three commits, "change", "revert", "new-change" I had two, "change",
> "new-change" with the end result being the same (i.e., instead of squashing
> all three into one "new-change", I had "change" and "revert" +
> "new-change").

Did you have a series of three commits being squashed in your to-do list? I mean, did you have a list like this:

   pick ...  do foo
   squash ...  revert "do foo"
   squash ...  What I really meant to do.

I suppose the rebase stopped after the first squash failed due to the emptiness of the proposed result. Then rebase --continue proceeded, having decided that you were finished with the 'revert' commit. Then ... I would expect the next commit would actually be squashed, but I can only speculate at the reasons it might have decided not to after your continue.

This actually sounds like another case of a bug I reported a few weeks ago[1] and which Fabian Ruch was kindly investigating[2] and trying to fix. I don't think his fix would have helped in this case, but I do think it is worthy of consideration for that same patch series.

Show 6 quoted lines

>> It's not clear to me what --continue _should_ do in this case, but it does
>> seem like the two options here should be
>
> I sort of expect a squashed commit of "change" + "revert" to be an empty
> commit, and of "change" + "revert" + "new-change" to be a commit of
> "new-change".

Yes, but empty commits are discouraged on some projects. If you want your "change + revert = empty" commit to appear after the squash, I would expect you would want to use --keep-empty on your inital rebase command. But I'm not sure that will do what you expected either; it may only keep previously-empty commits during the rebase.

[1] http://article.gmane.org/gmane.comp.version-control.git/245688 [2] http://www.mail-archive.com/git%40vger.kernel.org/msg51703.html