Re: fast-forward only policy

[Felipe Contreras <felipe contreras gmail com>]

On Wed, May 6, 2009 at 11:34 PM, Germán Póo-Caamaño <gpoo gnome org> wrote:
> On Wed, 2009-05-06 at 23:26 +0300, Felipe Contreras wrote:
>> On Wed, May 6, 2009 at 2:04 PM, Ross Burton <ross burtonini com> wrote:
>> > On Wed, 2009-05-06 at 12:27 +0200, Vincent Untz wrote:
>> >> Le mercredi 06 mai 2009, à 02:21 +0300, Felipe Contreras a écrit :
>> >> > Debian patches are debian patches, they control them, and they make
>> >> > debian releases. If GNOME decides to remove those commits the
>> >> > distributions will not loose their patches.
>> >>
>> >> I think this summarize well the whole thing: we do not want to remove
>> >> commits.
>> >
>> > Agreed.  All the way through this thread I've been wondering what
>> > possible reason there would be for throwing away a commit on a
>> > historical branch.
>>
>> It's not about throwing away commits, it's about throwing away unused branches.
>>
>> I've already explained two ways in which the branches can be thrown
>> away without loosing the commits although personally I would just
>> throw the commits away.
>>
>> My feeling is that if GNOME were using git at the time of those legacy
>> commits where made, the people developing them would have kept the
>> changes locally, and by this time, the commits would have been thrown
>> away anyway. In practice there's no difference between throwing away
>> local commits and throwing away public commits that nobody will use.
>
> They are used by software archeologist's, for mining purposes.  It is
> part of the project's history, and you should never regret of your
> history.

Am I denying my inheritance when I do undo? When I do 'git commit
--amend', or 'git rebase'? Nah.

Did the linux project made a fatal mistake when they decided to drop
*all* history (from bitkeeper) and start from scratch (in git)? Nah.
You can find all the history of the linux project divided into 3
repos, even from the very first release, for archaeologists or
whatever reason.

Would you fight to keep alive the branch Linus just found too crappy
and just killed it? If a commit never made it to a release and
probably never would, is it really that important?

I guess this is like the abortion debate. When is a commit really
alive? Does commits feel pain when they are killed before being
pushed?

-- 
Felipe Contreras