On 10 August 2014 19:26, Eli Zaretskii <eliz@gnu.org> wrote:

I'd prefer that all the changes for such a build be committed at once
together, not bit by bit, and only if you eventually succeed to
produce a working binary using this method.

Sorry, I was trying to separate my patch into parts with different implications (e.g. no effect on the current DOS build system; some effect on that system but not more generally; more general changes); but instead I'll concentrate on producing a single patch that, as you say, results in a working binary.

> I think you're also saying that I should test both HAVE_SELECT &&
> HAVE_SYS_SELECT_H therefore?

Yes, at the very least.

Noted.

But let's delay this until you have gone all
the way to building Emacs using this method. We won't know all of the
implications until you do.

Sure, OK.

I am closing this bug; since you diagnosed my other patch submission as a more general bug in configuration, I'll look into a fix for it at that level.

--
http://rrt.sc3d.org