Re: [MinGW-dvlpr] Problems with cross compiler w32api/lib/Makefile.in

[Christopher F. <cgf...@cy...>]

On Wed, Sep 13, 2006 at 12:01:19AM +0100, Keith Marshall wrote:
>On Monday 11 September 2006 5:07 pm, Christopher Faylor wrote:
>> >>I'd appreciate some help here, given that autotools are really not my
>> >>strong suit.
>> >
>> >It used to work. ?What changed?
>>
>> Corinna was trying to get things working so that a cross-mingw
>> environment could be built for Linux. ?Looking at the configure code,
>> it's obvious that the with_cross_host test is too broad of a test for
>> what she was trying to do.
>
>Perhaps Corinna should talk to us.  Both Michael Gerdau and I have been 
>successfully building, and running cross-mingw on Linux for quite some 
>time, and I know we are not alone.

She discussed what she was trying to do when she submitted the patch to
mingw-patches.

>> Since she isn't going to be around for a while, I'm going to back out
>> the problematic part of her code. ?I suspect that just changing the
>>
>> ifneq (,$with_cross_host)
>
>Well, that is syntactically invalid.  I've already pointed out the typo, 
>and Chris Sutcliffe has corrected it.

The email that you are responding to was sent at 2006-09-11 16:07 GMT.
My CVS checking was a couple of minutes earlier.  Your email noticing of
the typo was on something like 2006-09-11 22:33 GMT.  Chris Suttcliffe's
email was from today.

Perhaps the confusion comes from the fact that although you only approved
my email to mingw-dvlpr today, it was actually sent yesterday.

However, I just erroneously backed out Corinna's change from the mingw
Makefile rather than the w32api/lib Makefile and friends.  That was my
mistake.

Regardless, I didn't think that the Makefile.in change was right anyway
and since it was apparently only to benefit Corinna's particular usage,
I thought backing it out and working with her when she returns made some
sense.  I just backed out the change in the wrong file and I apologize
for putting Chris to extra work for that.

>>to something which specifically tests for mingw is probably the right
>>solution, but I'll work with Corinna on that when she gets back.
>
>I won't claim to be an autoconf expert, but I have been using it
>extensively for between two and three years now, so I'd like to think I
>have attained a fair level of competence.

If nothing else, you certainly seem to have achieved the je ne sais quoi
attitude that goes with working on free software.

>I've also been looking at the autoconfigury for both w32api and
>mingw-runtime.  It is a mess,

Yes, I got this point.  You have mentioned it repeatedly.  I can't
speak for anyone else, but I don't think you need to mention it again.

>and I've already started to clean it up.  I do not understand some of
>the things Corinna did, and would welcome an opportunity to discuss
>them with her.  Please ask her to contact me on her return.

Mentioning that I would work with her was merely meant to convey that I
wasn't expecting anyone else's help.  It wasn't an indication that she
worked for me or vice versa.

If you want to talk to her, then go right ahead and send her email.

>BTW, I've moderated this posting through, on this occasion; please use 
>your subscribed address for future postings, or advise if you would like 
>me to enter a subscription for this one.

The cygwin.com pseudo-address is what I use when sending email to cygwin
lists.  It will only show up when messages are cross-posted.

If you have some way to allow that email address in the From: without
actually trying to send email to it, then, sure, go ahead and set that
up.

cgf