Re: Repo cpnversion progress report

[Karl Fogel <kfogel@red-bean.com>]

Juanma Barranquero <lekktu@gmail.com> writes:
>On Fri, Jan 31, 2014 at 6:31 PM, Karl Fogel <kfogel@red-bean.com> wrote:
>
>> I don't agree with Eli's comments in a later mail about how he wasn't
>> sure we should entrust this job to ESR.  It's a complex, time-consuming,
>> and apparently thankless task; IMHO ESR is doing a good job considering
>> all the inputs involved (both human and machine).
>
>Well, do you at least agree with him when he says
>
>> I'm ignored already.  You never accepted any argument of mine through
>> that long and tiresome "discussion".  You are always right and all
>> those who disagree with you are always wrong.
>
>and
>
>> I never had any influence over that, because you refuse to listen
>> to any dissenting arguments, and prefer to stubbornly do what you
>> decided was right ahead of any discussion.
>
>?

My impression -- I've followed the discussions, but have not read every
single message or subthread -- is that the above is an exaggeration.

>Because that's also my feeling about every single switching-to-git
>discussion we've had so far. The only thing Eric has backtracked about
>is the imminence of switching, and that because Stefan specifically
>said that it wouldn't happen during the freeze.

The most important fact about all our prior switching-to-git discussions
was that always RMS vetoed the idea, explicitly saying that the number
of developers who wanted to switch was not relevant to the decision.  He
even continued saying this after it was obvious that Bazaar was not in
any meaningful sense a GNU project.

Given that history, I find it ironic that ESR is now being accused of
not listening enough :-).

In another followup, ESR wrote:
>I hear the argument about forensics, but the Bazaar revision numbers
>are no more helpful for that than the action stamps. If anything,
>spelunking is an argument for appending those numbers to the change
>comment of each revision - leaving them in as references would a 
>bass-ackwards approach to the problem.
>
>But there's a better way.  We're going to have a complete
>revision-to-action-stamp map as part of the record.  It would be
>pretty easy to write Emacs code that uses that map to find revisions
>by Bazaar reference number.  That's the right solution, IMO.

Okay.  I'm not able to pay as close attention to the problem as others
here are, but if you feel you've got a plan that handles the forensic
case well enough, that's fine by me.

There are a million ways to convert a bzr history to git.  We just need
to pick one of those ways that will work for Emacs development.  (insert
the usual comment about diminishing returns, lather, rinse, repeat?)

Best,
-Karl