Re: Why I'm working on a NetBSD conversion

[Jens Rehsack <rehsack%gmail.com@localhost>]

Am 25.10.2014 um 17:31 schrieb Eric S. Raymond <esr%thyrsus.com@localhost>:

> S.P.Zeidler <spz%NetBSD.org@localhost>:
> [...]
> I understand the problem.  I can't solve the politics, but I have some
> technical advice that may be helpful.
> 
> The effective demise of bzr simplifies matters. There are now only two
> even remotely realistic choices going forward: git, and hg.

What's with fossil? I experienced some minor performance issues because
of missing local optimizations in fossil - nothing which can't be solved
when no as much over-busy'ed ...

> Subversion 
> is, practically speaking, out of the running - you'll find out why
> if you benchmark checkouts on a history this size.
> 
> But that raises an interesting possibility.  Since late last year it has
> been possible to present an hg view of a git repository. It would not
> be very difficult, now, to move to a setup that presents to developers
> as either git or hg depending on what client configuration they choose.
> 
> Do you think you could sell your gitophobic developers on a system that
> does not lock them into using git, but allows hg? Obviously the anti-hg
> people would be immediately satisfied.  Unless you have a large contingency
> that gets hives from *both* systems, this might be a solution.

As far as I understood Joerg, that would satisfy fossil people as well.
But a very important point in tooling was, that the utility to checkout
is distributed with the base system. Neither hg nor git can satisfy here.

Cheers
-- 
Jens Rehsack
rehsack%gmail.com@localhost