Re: irt: Re: Core statement on version control systems

["Constantine A. Murenin" <mureninc%gmail.com@localhost>]

On Mon, 8 Dec 2025 at 00:43, David Holland <dholland-tech%netbsd.org@localhost> wrote:
>
> On Mon, Dec 08, 2025 at 06:19:39AM +0000, David Holland wrote:
>  > Now try to rebase on main:
>  >
>  >    % git rebase main
>
> And I meant to mention: if you do these merges and rebases from hg,
> they work correctly without conflicts, even if the rename was
> committed on the git side.

It's certainly nice for things to work, as opposed to not work, correctly.

But how come none of the other large projects have the renaming
implementation as a deal-breaking issue?

Git was designed for the Linux kernel; which is no small project, and
they must certainly have their share of renames, too.

Is it basically our licence header that steers it off-track?

But doesn't Mozilla also features a rather large licence header in
each file with the MPL?  (Actually, just checked, they don't anymore;
I think this was actually a deliberate re-licensing cleanup that they
did back many years ago now, as per
https://github.com/mozilla-firefox/firefox/commit/cb6a072c2a187508fa46cf982e3bed75df972b30
that's dated May 2012, some 13 years ago; IIRC, it was done in many
stages, so, there's probably other commits, too.)  Why did they move
from Hg to Git still, just a few months ago, in May 2025, after first
announcing the definitive plan to migrate over a year prior, in
November 2023?

Mozilla's migration was clearly deliberate, as it was executed about
18mo past a very public announcement of plans.  Did a single
Mozilla/Firefox contributor object during said 18 months?  Did a
single Mozilla Contributor quit contributing over the migration?

These recent emails about renames certainly do help us better
understand your insistence on Mercurial, but I think it is still
unclear why it's a deal-breaker for NetBSD when noone else is facing
the same issues, because right now, it's far less than clear.  If
anything, Mozilla's experience shows that perhaps we should…  Yeah, I
guess I won't table such a suggestion, but I still think we should
have a better explanation on why renaming is such a huge issue for us,
and isn't for anyone else.

Best regards,
Constantine.