Re: CVS commit: src

[Havard Eidnes <he%NetBSD.org@localhost>]

> > My guess is that the portmaster at this point needs to step up to
> > the plate and take the necessary steps to put the install kernel
> > and the ramdisk image on a diet.
>
> It's better to check why ramdisk overflows and also to consider
> how we can shrink it, before blindly bumping ramdisk size only
> for buildable state.

It would be better, yes.  But it would also require more time and
effort than I had available, and ivolves some of the "why does
the kernel in general grow", which is a general problem I'm not
up for solving singlehandedly.

Intentionally leaving a port non-buildable will probably cause
other subsequent build failures to go undetected in the mean time
(OK, granted, this is near the end of the build, but still...),
so I think that is also not a good strategy.

Regards,

- Håvard