Re: github.com/NetBSD/src 5 days old?

[Marc Balmer <mhbalmer%gmail.com@localhost>]

> Am 28.04.2020 um 10:50 schrieb maya%netbsd.org@localhost <maya%NetBSD.org@localhost>:
> 
> On Tue, Apr 28, 2020 at 08:30:43AM +0200, Marc Balmer wrote:
>> 
>> 
>>> Am 28.04.2020 um 08:29 schrieb Andreas Gustafsson <gson%gson.org@localhost>:
>>> 
>>> maya%NetBSD.org@localhost wrote:
>>>> Yes, I believe joerg and spz are changing the conversion from
>>>> cvs->??->git to hg->git, to match what will be done once we stop using
>>>> CVS.
>>> 
>>> Has there been a formal decision choosing hg over git?
>> 
>> I am also interested in this.
>> 
>> 
> 
> This feels like a protest. Since it's addressing me, I'd like to point
> out I'm just letting people know why the conversion is down, and don't
> get any more of a say over things than others.

No, that was not to be understood as a protest, and addressing you personally was by mistake - I hit reply-to-all and did not check the adresses.

Well, this time it is not by mistake, as I intended to reply to you ;)

> As a reminder, hg/git offer far better interoperability (than CVS).
> Much of my own NetBSD work is done on Git, and even if I don't stop
> doing this, I would be happier if the backend was Mercurial.
> 
> The CVS->??->git conversion loses information on the parents of branch
> merges, so we carry a growing graft file, and it has to be adjusted
> whenever there's a forced push.
> 
> Having Mercurial at the back would eliminate ~all forced pushes and have
> real merge commits. Exporting the commits would require a lot less
> threats and custom scripts on current-users, because pushing is
> distinct from committing.

Thanks for your explanations.