[comp.sources.d] v16i034: Larry Wall's Configure

kebsch@nixpbe (10/27/88)

Jay Libove wrote:

> Well, Larry (for the first time I've ever seen) made a number of very
> very bad assumptions in his configure distribution.
>
> 1) that you have Perl. Perl doesn't run on 80286 machines. Not at all.
			 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
That's wrong! I yacc'ed perl.y on a 32 bit machine and compiled
Perl-2.0, Patchlevel-14 on a 80286 machine. It runs pretty good and
I have no problems! My Operatingsystem is UNIX Sytem V Rel-2 (Microport).
So don't flame on Larry Wall, flame on SCO Xenix SysV/286 v2.2.1.

> 2) that your system can deal with a simple makefile unser /bin/csh
>    without SHELL=/bin/sh in the makefile - SCO Xenix SysV/286 v2.2.1
>    breaks here. You have to have the SHELL=/bin/sh.

That's your problem! So don't flame on Larry Wall, flame on
SCO Xenix SysV/286 v2.2.1.

> 3) the Configure that came with the configure package didn't ask
>    for other CFLAGS and LDFLAGS. Well, you need that on segmented
>    machines in order to give the model switches, else you have to
>   hand locate and edit the makefiles to add them.

Oh my dear...! Fix it (it's easy) and don't flame on Larry Wall!

> What happenned Larry? Your utilities used to be the best, and most
> portable programs on the net...

That's really true!! So, please, never flame on Larry Wall.
Don't forget: Software from Larry Wall is free and takes only
a little bit work for you.

That's all. Sorry for my bad English, but I think you will understand
what I am meaning.

Waldemar Kebsch

PS: Some month ago I installed SCO Xenix SysV/286 v2.2.1 on my private
    80286 machine. All what I remember are a lot of messages like
    this:"Compiler internal error". So I reformated the disk and
    installed Microport System V/AT 2.2.2. I like it.
--------------------------------------------------------------------------
UUCP: you are --> in the USA:   uunet!linus!nixbur!kebsch
UUCP: you are not in the USA:   mcvax!unido!nixpbe!kebsch
UUCP:              (however):   kebsch@nixpbe.UUCP
NERV ......:   kebsch.pad
Papermail .:   Waldemar Kebsch, Nixdorf Computer AG, Entwicklungstechnik,
	       Pontanusstr. 55, D-4790 Paderborn, W-Germany
Phone .....:   [nation]-5251-14-6155

erik@mpx2.UUCP (Erik Murrey) (10/31/88)

In article <3100001@nixpbe.UUCP>, kebsch@nixpbe writes:
> Jay Libove wrote:
> > 1) that you have Perl. Perl doesn't run on 80286 machines. Not at all.
> 			 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> That's wrong! I yacc'ed perl.y on a 32 bit machine and compiled
> Perl-2.0, Patchlevel-14 on a 80286 machine. It runs pretty good and
> I have no problems! My Operatingsystem is UNIX Sytem V Rel-2 (Microport).
> So don't flame on Larry Wall, flame on SCO Xenix SysV/286 v2.2.1.
> 


Oh, no.  Not again...  I would note that your microport yacc wouldn't
compile perl.y either, unless you just wanted to yacc it on a 32
bit box for no reason.  This is an 80286 problem, not SCO's, not Microport's.

Let's not start another SCO/Microport war...


-- 
Erik Murrey
MPX Data Systems, Inc.
erik@mpx2.UUCP
...!{bpa,spl1,cbmvax,vu-vlsi}!mpx1!erik

jtc@tessera.UUCP (J.T. Conklin) (10/31/88)

In article <3100001@nixpbe.UUCP> kebsch@nixpbe writes:
>Jay Libove wrote:
>
>> Well, Larry (for the first time I've ever seen) made a number of very
>> very bad assumptions in his configure distribution.
>>
>> 1) that you have Perl. Perl doesn't run on 80286 machines. Not at all.
>			 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>That's wrong! I yacc'ed perl.y on a 32 bit machine and compiled
>Perl-2.0, Patchlevel-14 on a 80286 machine. It runs pretty good and
>I have no problems! My Operatingsystem is UNIX Sytem V Rel-2 (Microport).
>So don't flame on Larry Wall, flame on SCO Xenix SysV/286 v2.2.1.

Agreed, I run perl 2.0 on my SCO Xenix/286 machine without a hitch,
but I did have to cross compile it on a '386.  It compiles all right on
the '286 but it fails some of the regression tests.

With a little work on the compiler, a working perl executable could
be generated solely on a '286.  Perhaps bug fixes/updates are being
held up in favor of ansification?  In the mean time, I do have a working
binary that I'm willing to share -- send me mail if you want it.

	--jtc

-- 
J.T. Conklin
    ...!{ubc-cs,uunet}!van-bc!tessera!jtc

mjy@sdti.UUCP (Michael J. Young) (11/01/88)

In article <265@mpx2.UUCP> erik@mpx2.UUCP (Erik Murrey) writes:
>In article <3100001@nixpbe.UUCP>, kebsch@nixpbe writes:
>> Jay Libove wrote:
>> > 1) that you have Perl. Perl doesn't run on 80286 machines. Not at all.
>> 			 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>> That's wrong! I yacc'ed perl.y on a 32 bit machine and compiled
>> Perl-2.0, Patchlevel-14 on a 80286 machine. It runs pretty good and
>> I have no problems! My Operatingsystem is UNIX Sytem V Rel-2 (Microport).
>
>Oh, no.  Not again...  I would note that your microport yacc wouldn't
>compile perl.y either, unless you just wanted to yacc it on a 32
>bit box for no reason.  This is an 80286 problem, not SCO's, not Microport's.

No, it's a yacc problem.  Using the new version of yacc from Microport's
bbs that was compiled with larger tables, I was able to yacc perl.y.  I
have a working perl, although I'm only running at patchlevel 10 (just
lazy, I guess).

But let's face it.  It didn't come easy.  There are a number of constructs
in the perl source that cause the Microport compiler to bite the big one.
I'm still looking forward to Larry's patches to make my hacked up copy more
"official".  Hopefully, the next update will include the yacc output as
well as the .y file for those who can't get access to a large-table yacc.
-- 
Mike Young
Software Development Technologies, Inc., Sudbury MA       Tel: +1 508 443 5779
Internet: mjy@sdti.sdti.com                 UUCP: {harvard,mit-eddie}!sdti!mjy

PLS@cup.portal.com (Paul L Schauble) (11/02/88)

All this discussion about the configure generator has been interesting. Now,
would someone please tell us what this program does?

   ++PLS

greim@sbsvax.UUCP (Michael Greim) (11/11/88)

In article <3100001@nixpbe>, kebsch@nixpbe.UUCP writes:
< 
< Jay Libove wrote:
< 
< > Well, Larry (for the first time I've ever seen) made a number of very
< > very bad assumptions in his configure distribution.
< >
< > 1) that you have Perl. Perl doesn't run on 80286 machines. Not at all.
< 			 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
< That's wrong! I yacc'ed perl.y on a 32 bit machine and compiled
< Perl-2.0, Patchlevel-14 on a 80286 machine. It runs pretty good and
< I have no problems! My Operatingsystem is UNIX Sytem V Rel-2 (Microport).
< So don't flame on Larry Wall, flame on SCO Xenix SysV/286 v2.2.1.
< 
I had a problem to yacc perl on my system, too. Like Jay I yacc'ed perl on
a 32 bit machine.
Recently I ported bison and tested it, among other things, on perl.y.
It worked ok.  And if you are coming up against some built-in limits in bison
you can always change them.

	-mg
-- 
email : greim@sbsvax.informatik.uni-saarland.dbp.de
  (some mailers might not like this. Then use greim@sbsvax.uucp)
  or  : ...!uunet!unido!sbsvax!greim
# include <disclaimers/std.h>

mjy@sdti.UUCP (Michael J. Young) (11/15/88)

In article <642@sbsvax.UUCP> greim@sbsvax.UUCP (Michael Greim) writes:
>I had a problem to yacc perl on my system, too. Like Jay I yacc'ed perl on
>a 32 bit machine.

For those of you who are using Microport System V/AT, there is a new
version of yacc on Microport's bbs that was compiled with larger tables.
I used it when I installed perl on my system and had no problems.

I don't know if the larger yacc was included in the 2.4 distribution.
-- 
Mike Young
Software Development Technologies, Inc., Sudbury MA       Tel: +1 508 443 5779
Internet: mjy@sdti.sdti.com                 UUCP: {harvard,mit-eddie}!sdti!mjy