Re: Win32 build broken

On 4/30/07, Rafael Garcia-Suarez <rgarciasuarezgmail.com> wrote:
> On 30/04/07, Steve Hay <steve.hayuk.radan.com> wrote:
> > The problem here (sadly not shown up in the
Compiler messages section)
> > is this:
> >
> > Error: 'Thread' not in typemap in Thread.xs, line
74
> > Error: 'Thread' not in typemap in Thread.xs, line
84
> > Error: 'Thread' not in typemap in Thread.xs, line
89
> > Error: 'Thread' not in typemap in Thread.xs, line
90
> > Error: 'Thread' not in typemap in Thread.xs, line
96
> > Error: 'Thread' not in typemap in Thread.xs, line
101
> > Error: 'Thread' not in typemap in Thread.xs, line
115
> >
> > which I guess is caused by change #31088:
> >
> > Change 31088 by rgsbenny on 2007/04/26
13:58:34
> >
> >         Remove the Thread typemap
> >
> > Why was that done?
>
> As a complement to :
>
> Change 31078 on 2007/04/25 by rgsstcosmo
>
>         Subject: [PATCH] Thread.xs doesn't need magic
sig anymore
>         From: "Jerry D. Hedden"
<jdheddencpan.org>
>         Date: Mon, 23 Apr 2007 13:25:20 -0400
>         Message-ID:
> <1ff86f510704231025y22d30ecfk1ebc418f2b3a0bf7mail.gmail.com>
>
> I can revert those two patches.
> However, I'm not sure why Thread is needed on Win32, or
whether
> ext/Thread should be completely removed. If someone as
some time to
> look into this... (Jerry, for example?)

Please could we revert these patches until then?

yves

-- 
perl -Mre=debug -e "/just|another|perl|hacker/"

On 4/30/07, demerphq <demerphqgmail.com> wrote:
> On 4/30/07, Rafael Garcia-Suarez <rgarciasuarezgmail.com> wrote:
> > On 30/04/07, Steve Hay <steve.hayuk.radan.com> wrote:
> > > The problem here (sadly not shown up in the
Compiler messages section)
> > > is this:
> > >
> > > Error: 'Thread' not in typemap in Thread.xs,
line 74
> > > Error: 'Thread' not in typemap in Thread.xs,
line 84
> > > Error: 'Thread' not in typemap in Thread.xs,
line 89
> > > Error: 'Thread' not in typemap in Thread.xs,
line 90
> > > Error: 'Thread' not in typemap in Thread.xs,
line 96
> > > Error: 'Thread' not in typemap in Thread.xs,
line 101
> > > Error: 'Thread' not in typemap in Thread.xs,
line 115
> > >
> > > which I guess is caused by change #31088:
> > >
> > > Change 31088 by rgsbenny on 2007/04/26
13:58:34
> > >
> > >         Remove the Thread typemap
> > >
> > > Why was that done?
> >
> > As a complement to :
> >
> > Change 31078 on 2007/04/25 by rgsstcosmo
> >
> >         Subject: [PATCH] Thread.xs doesn't need
magic sig anymore
> >         From: "Jerry D. Hedden"
<jdheddencpan.org>
> >         Date: Mon, 23 Apr 2007 13:25:20 -0400
> >         Message-ID:
> >
<1ff86f510704231025y22d30ecfk1ebc418f2b3a0bf7mail.gmail.com>
> >
> > I can revert those two patches.
> > However, I'm not sure why Thread is needed on
Win32, or whether
> > ext/Thread should be completely removed. If
someone as some time to
> > look into this... (Jerry, for example?)
>
> Please could we revert these patches until then?

Apparently Win32 is doing the same thing more or less that
VMS was,
namely building Thread when any thread-like configure option
was
chosen.  Unix (or anything that uses Configure) only builds
it when
-Dusethreads AND -Uuseithreads are both in effect, so the
problem you
are reporting is true everywhere, but the default settings
for an
unusual and possibly nonsensical set of configuration
options are
different with each configuration tool.  On VMS I dodged the
issue
(and mimicked Unix behavior) by not buiilding Thread when
ithreads
have been selected.  Thread is still broken, but fixing it,
to
paraphrase one of the local sages, is not my itch to
scratch.  And in
any case it looks like 5.9.x is about to undergo a
Thread-ectomy.