Greg Hudson wrote:
> No. I don't think we should be targeting the Linux kernel's current
> developer model. It is an albatross among open source projects and it
> does not produce good results. I've written a document about this
I have to agree 100% with this. As an open source developer myself, none
of my projects follow the Linux model. Having 20 different kernel trees
and picking the one that includes the fixes you _may_ need is poor
development, open or closed source.
Linux is really a victim of its own success. There is quite a bit of pride
associated with submitting patches for the kernel; and that means that
lots of people want to do it. In that light, the "Linus Patch Filter"
makes a bit more sense, but it still leads to very disorganized
development.
Most projects simply don't work this way. Usually there is a small
group of core developers who actually do commits and take care of the
tree. Take for example the BSD's. People submit patches but they do so by
checking out and e-mailing in diff's. The maintainers then do the commits.
That scheme seems to work very well in practice (I say this mostly from
personal experience).
I think having Subversion target those projects, in particular, one of the
BSD's is very important. The model of CVS is fine, but the implementation
of CVS isn't. For that, Subversion is the logical replacement.
L8r,
Mark G.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sun Mar 16 21:38:08 2003