Re: Linux-2.0.1

Kevin Buettner (kev@primenet.com)
Thu, 4 Jul 1996 11:02:27 -0700


On Jul 4, 10:50am, Andrew E. Mileski wrote:

> > > Developers: I will NOT accept large patches any more. If you find bugs,
> > > go for the one-liner obvious bugfix rather than a rewrite. Due to me
> > > being abroad I couldn't work as closely with people on 2.0.1 patches as I
> > > need to, but I'm back, and this kind of mega-patch will NOT happen again..
> >
> > Does this include drivers as we will need to add a fairly large driver
> > chunk because the Metricom firmware has changed. That should all be a
> > single clean driver not affecting other files.
>
> Personally, I don't have a problem with 1 big patch = a new version.
> Does anybody else? Does Mr. Torvalds?

Large patches increase the risk that something will break (as opposed
to getting fixed) which is a very bad thing for a non-development
(i.e. stable) series of kernels. This is particularly true when some
bit of functionality is rewritten (to be cleaner, more efficient,
whatever) when a simpler but less elegant bug fix would have sufficed.
I think this is reason for Linus' objections to large patches.

When the 2.1 development series starts up, I suspect that Linus
will be willing to accept large _focused_ patches.

> I say this, because PnP support will likely be 1 patch, because
> it touches just about everything.

Umm, yeah. I'm not knocking PnP or your work, but I suspect that when
it's integrated, there'll be much breakage for several kernel
versions. I don't personally have a problem with this just so long
as it happens in 2.1 instead of 2.0.

Kevin