Re: self bootstrap on Interix 3.5

[Simon Gerraty <sjg%juniper.net@localhost>]

On Wed, 12 Apr 2006 09:32:08 +0900, OBATA Akio writes:
>Hmm, if it is a bug, how about this?

Yes this is part of the bug.  I posted a fix to tech-toolchain earlier
today.  I've had a positive review, but will wait a little while in case
anyone else can think of an objection.

>   953                  ln = Lst_Find(sysMkPath, (ClientData)NULL, ReadMakefil
>e);

FWIW I tested your patch for removing the default_syspath and found it
broke the default boot-strapping I use to test new releases.  Fixing the
above is my prefered solution.

>> > >2. ${MAKEFILE} issue(devel/bmake)
>> 
>> >> I think we need Todd's input here...
>> >> I don't have access to one of these systems, so have to go by the input
>> >> from those that do. 
>> >
>> >He tell us about it yesterday in tech-pkg:
>> >http://www.netbsd.org/cgi-bin/query-pr-single.pl?number=28392
>> 
>> That tells me that he thinks it is unwise for pkgsrc to re-use
>> ${MAKEFILE}, which seems somewhat orthogonal.
>
>If using NFS on Interix, sometime change ${MAKEFILE} from "Makefile"
>to "makefile" even if using "Makefile".

Oh I see... this is due to the fact that when no makefile is specified,
make looks for 'makefile' before 'Makefile', which btw isn't something I'd
support changing - since I for one rely on it.  Anyway, you end up with
MAKEFILE=makefile even though the file was actually Makefile.

This is presumably only "interesting" when mixing a case insensitive fs
with a case sensitive one.  Did you say that the NFS client can be
configured to do case sensitivity correctly?  If so, perhaps adding that
info to the PR you mentioned would be very helpful.

>Or simply, useing "-f Makefile".

Exactly.  No need to frob make's internals. Ie. your patch to undo the
previous hack should be ok, but depending on the answer to the above NFS
question that may not help pkgsrc.

--sjg