Re: Hurd on L4 Project

[Niels Möller]

Ian Duggan <address@hidden> writes:

> How much danger is there that we'd end up with a fork of the Hurd
> codebase if we did something like this.

For the first steps, getting a basic L4 environent to work with (as
outlined by Farid), there's no danger at all, as the Hurd code isn't
involved yet. Sooner or later we'll get to the point where servers and
libraries that are part of the Hurd have to be modified so that they
can run both in the new environment and on Mach. Then some
coordination and careful planning is needed, but I think it's really
unlikely that a fork will be needed.

(And generally, those few forks that have happened in the free
software world aren't evil, they're the result of different and
incompatible interests and priorities among the developers. If the
l4-hurd eventually has to fork, so be it, and we'll deal with it then,
but I thinks it's very unlikely, and definitely not something to worry
about today).

> Additionally, I've been looking at BitKeeper lately for source control.
> Has anyone tried it? It looks like a nice combination of CVS and
> Clearcase. It works distributed and disconnected like CVS, but has a
> branching and merging scheme more like Clearcase.

As Neal pointed out, it's not quite free, which rules out using it for
GNU projects. Anyway, CVS may not be perfect, but it is good enough
and people are used to it. Most people are naturally quite
conservative about version control tools. Trying new VC tools,
interesting as it may be, ought to be out of scope for the l4-hurd
project.

/Niels