From: UnixOS2 Archive
To: "UnixOS2 Archive" <UnixOS2_Archive at unixos2.com>
Date: Sat, 23 Feb 2002 04:16:24 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600
Subject: [UnixOS2_Archive] No. 144

**************************************************
Friday 22 February 2002
 Number  144
**************************************************

Subjects for today
 
1   Re: Emacs customisation : Masaru Nomiya <nomiya at ttmy.ne.jp>
2  Re:  make 3.79.1 : John Poltorak <jp at eyup.org>
3  Re:  EMACS build error : John Poltorak <jp at eyup.org>
4  Re:  Re: Emacs customisation : John Poltorak <jp at eyup.org>
5  Re:  Re: Emacs customisation : Masaru Nomiya <nomiya at ttmy.ne.jp>
6  Re:  Identifying sh.exe : Jun Sawataishi <jsawa at attglobal.net>
7  Re:  Re: Emacs customisation : Jun Sawataishi <jsawa at attglobal.net>
8  Re:  GNAT 3.14p : DWParsons at t-online.de (Dave Parsons)
9  Re:  UnixOS/2 Build System - proposal : Neil Waldhauer <zonker at well.com>
10   Migrating from gcc 2.8.1 to 3.0.3 : John Poltorak <jp at eyup.org>
11  Re:  Migrating from gcc 2.8.1 to 3.0.3 : Henry Sobotka <sobotka at axess.com>
12  Re:  cpp0.exe error msg : Henry Sobotka <sobotka at axess.com>
13  Re:  UnixOS/2 distro : John Poltorak <jp at eyup.org>
14  Re:  (un)protoize : Thomas Dickey <dickey at herndon4.his.com>
15  Re:  make 3.79.1 : Andreas Buening <andreas.buening at nexgo.de>
16  Re:  make 3.79.1 : John Poltorak <jp at eyup.org>
17   cpp0.exe error msg : John Poltorak <jp at eyup.org>
18  Re:  OSFree... : Stefan Neis <neis at cdc.informatik.tu-darmstadt.de>
19  Re:  Migrating from gcc 2.8.1 to 3.0.3 : Henry Sobotka <sobotka at axess.com>
20  Re:  OSFree... : Sebastian Wittmeier (ShadoW)" <Sebastian.Wittmeier at ginko.de>
21   UnixOS/2 Build System - proposal : John Poltorak <jp at eyup.org>
22   gcc 3.0.3. _optlink questions for Andy : Henry Sobotka <sobotka at axess.com>
23  Re:  UnixOS/2 Build System - proposal : John Poltorak <jp at eyup.org>
24  Re:  (un)protoize : Stefan Neis <neis at cdc.informatik.tu-darmstadt.de>
25  Re:  Migrating from gcc 2.8.1 to 3.0.3 : Michel SUCH" <msuch at free.fr>
26  Re:  gcc 3.0.3. _optlink questions for Andy : Henry Sobotka <sobotka at axess.com>
27  Re:  cpp0.exe error msg : Henry Sobotka <sobotka at axess.com>
28  Re:  Migrating from gcc 2.8.1 to 3.0.3 : Henry Sobotka <sobotka at axess.com>

**= Email   1 ==========================**

Date:  Sat, 23 Feb 2002 00:20:56 +0900
From:  Masaru Nomiya <nomiya at ttmy.ne.jp>
Subject:   Re: Emacs customisation

Hello,

In the Message; 

  Subject    : Re:  Re: Emacs customisation
  Message-ID : <20020222131850.D86 at eyup.org>
  Date & Time: Fri, 22 Feb 2002 13:18:50 +0000

[John] == John Poltorak <jp at eyup.org> has written:

John> I'm only trying to make it easy for other people to enjoy the benefits ofJohn> some hard work done by a few porters.

Yes, I know.
I do appreciate your efforts.

John> Would I'd like to have if possible is OS/2 based versions of some emacs 
John> packages which appear in Slackware here:-

I would like, but ...
Mr. Sasaki, Japanese great porter, wrote detailed Readme, and it is
very important to use Emacs 20.7 for OS/2.
Mr. Nakagawa, succesor of Mr. Sasaki's great work, do hope to
distribute Emacs 20.7 for OS/2 with Mr. Sasaki's Readme.
At the moment, translation is my work.....
 
Me> I'm very glad, and I must tell to Mr. Sasaki, Mr.Nakagawa.

John> It would be good to have the porters on the list.
 
Thaks.

John> Can be run full screen, ie not PM. 

By double clicking the title bar, it will become full screen.
How about this?
 
Me> Could you show me your .emacs and .XDefaults?

John> It's not very extensive, I'm just starting to collect a few macros. 
John> I have not used emacs under XFree so far.
[...]

.emacs for Emacs 20.7 for OS/2 must include

;; For OS/2
(setq os2-process-start-session t)

(setq os2-process-connection-type nil)

;; For ls
(setq dired-re-perm-bits "\\([^ ]\\)[-r][-w]\\([^ ]\\)\\([-a]\\)\\([-h]\\)[-s]")

and these are specific part for Japanese

(set-language-environment "japanese")
(prefer-coding-system 'japanese-shift-jis)
(setq default-buffer-coding-system 'sjis)
(setq file-coding-system 'sjis)
(set-buffer-file-coding-system 'sjis)
(set-keyboard-coding-system 'sjis)
(set-terminal-coding-system 'sjis)
(setq file-name-coding-system 'sjis)
(auto-compression-mode t)
(setq truncate-partial-width-windows nil) 

Moreover, to resolve your font problem in the window, you must write
.XDefaults like this;

*verticalScrollBars: right
Emacs.Font: fontset-MIN16
*Fontset-0: -*-fixed-medium-r-normal-*-14-*-*-*-*-*-fontset-MIN14, \
	ascii:-os2-MINCHO-medium-r-*--14-*-iso8859-1, \
	katakana-jisx0201:-os2-MINCHO-medium-r-*--14-*-jisx0201.1976-*, \
	japanese-jisx0208:-os2-MINCHO-medium-r-*--14-*-jisx0208.1983-*
*Fontset-1: -*-fixed-medium-r-normal-*-16-*-*-*-*-*-fontset-MIN18, \
	ascii:-os2-MINCHO-medium-r-normal-*-16-*-*-*-*-*-iso8859-1, \
	katakana-jisx0201:-os2-MINCHO-medium-r-normal-*-16-*-jisx0201.1976-*, \
	japanese-jisx0208:-os2-MINCHO-medium-r-normal-*-16-*-jisx0208.1983-*
*Fontset-2: -*-fixed-medium-r-normal-*-24-*-*-*-*-*-fontset-MIN24, \
	ascii:-os2-MINCHO-medium-r-*--24-*-iso8859-1, \
	katakana-jisx0201:-os2-MINCHO-medium-r-*-*-24-*-jisx0201.1976-*, \
	japanese-jisx0208:-os2-MINCHO-medium-r-*-*-24-*-jisx0208.1983-*

Yes, this is for Japanese edition for OS/2 except the first line.
I think you must write the first line as well.

Me> Then you'll get multiligual world!

John> This sounds useful. I'm not very familiar with font formats. 

John> Is there any info on BDF2MFN ? 

I must translate....

John> Where does it need as an input file?

Basically, Emacs is a multiligal editor, as you might know.
That is, even on the Uk version of OS/2, I can write Japanese sentence.

BTW. With mfn fonts, you can write .XDefaults as follows (when you use
16 bit font). I think;

*FontSetList: 16, 14, 24
*FontSet-14: W
	-ETL-Fixed-Medium-R-Normal--14-ISO8859-1
*FontSet-16: W
	-ETL-Fixed-Medium-R-Normal--16-ISO8859-1
*Fontset-18: W
        -ETL-Fixed-Medium-R-Normal--24-ISO8859-1

Please try this.
To make bdf fonts

    g:\>bdf2mfn jiskan16.bdf jiskan16.mfn

then see the FONTSET.OS2.

You can get bdf fonts from 

ftp://ftp.gnu.org/pub/gnu/intlfonts/intlfonts-1.2.tar.gz

# Could you join os2-emacs-ja ML? More powerful users help you.
# To join os2-emacs-ja ML send your mail to
#
#    os2-emacs-ja-ctl at m17n.org
#
#   with the messeage "subscribe".

Regards,

---
  Masaru Nomiya       mail-to: nomiya at ttmy.ne.jp

          "Bill! You married with Computers.
           Not with Me!"
          "No..h, with Money."

**= Email   2 ==========================**

Date:  Sat, 23 Feb 2002 00:26:45 +0000
From:  John Poltorak <jp at eyup.org>
Subject:  Re:  make 3.79.1

On Fri, Feb 22, 2002 at 10:53:37PM +0100, Andreas Buening wrote:
> John Poltorak wrote:
> > 
> > There was no %SHELL% or %MAKESHELL% variable set in the environment or in
> > the Makefile.
> > 
> > I set both on in turn, both with BASH and PDKSH as /bin/sh, and it made
> > no difference. There was no '$SHELL changed' msg.
> 
> Did you use "MAKESHELL=/bin/sh"? "export MAKESHELL=/bin/sh"
> might be required.

Originally I only tried running make from a CMD prompt, but have just 
tried BASH and PDKSH, but the results are the same.  

 
> > I tried using the debug flag, but didn't gain any additional info.
> 
> I think I know what happened. If there is no SHELL statement
> within the Makefile then cmd is used as default shell
> and you get no "$SHELL changed" message. And cmd cannot handle
> those backslash escaped quotes like
> -DVERSION="\"20.7\"" etags.c getopt.o getopt1.o regex.o -lc -o etags
> 
> make 3.76 and earlier has some magic built in but therefore it
> fails on more complicated commands.

I'm sure it's not easy trying to sort this out... but this new Make seems 
to be running OK otherwise, and it would be great to finally have a Make 
that everyone can use rather than having a choice of six for different 
situations.
 
 
> bye,
> Andreas
> 
> -- 
> One OS to rule them all, One OS to find them,
> One OS to bring them all and in the darkness bind them
> In the Land of Redmond where the Shadows lie.


-- 
John


**= Email   3 ==========================**

Date:  Sat, 23 Feb 2002 00:31:38 +0000
From:  John Poltorak <jp at eyup.org>
Subject:  Re:  EMACS build error

On Fri, Feb 22, 2002 at 08:58:04PM +0100, Andreas Buening wrote:

> You mean EM won't put a simple "#define strcasecmp stricmp"
> into string.h?

Maybe EM won't but we can, if it's right for UnixOS/2... 

 
> 
> bye,
> Andreas
> 
> -- 
> One OS to rule them all, One OS to find them,
> One OS to bring them all and in the darkness bind them
> In the Land of Redmond where the Shadows lie.


-- 
John


**= Email   4 ==========================**

Date:  Sat, 23 Feb 2002 00:47:51 +0000
From:  John Poltorak <jp at eyup.org>
Subject:  Re:  Re: Emacs customisation

On Sat, Feb 23, 2002 at 01:42:21AM +0900, Jun Sawataishi wrote:
> Nomyya>>> I'm sorry, but I can't understand the meaning of "text mode".
> JP>Can be run full screen, ie not PM. 
> Run emacs with "-nw" option
>    $ emacs -nw


Thanks. That's just what I was looking for. 

It's good to see that it still works as normal.

Is there a list of available options anywhere?

 
> #   OS/2 is not a question, it's a solution.
> #        SAWATAISHI Jun <jsawa at attglobal.net>
> #   http://www2s.biglobe.ne.jp/~vtgf3mpr/
> 


-- 
John


**= Email   5 ==========================**

Date:  Sat, 23 Feb 2002 01:37:04 +0900
From:  Masaru Nomiya <nomiya at ttmy.ne.jp>
Subject:  Re:  Re: Emacs customisation

Hello,

Sorry, I have mistaken.

In the Message; 

  Subject    :  Re: Emacs customisation
  Message-ID : <664pfspz.wl%nomiya at ttmy.ne.jp>
  Date & Time: Sat, 23 Feb 2002 00:20:56 +0900

[Me] == Masaru Nomiya <nomiya at ttmy.ne.jp> has written:

Me> *FontSetList: 16, 14, 24
Me> *FontSet-14: W
Me> 	-ETL-Fixed-Medium-R-Normal--14-ISO8859-1
Me> *FontSet-16: W
Me> 	-ETL-Fixed-Medium-R-Normal--16-ISO8859-1
Me> *Fontset-18: W
Me>         -ETL-Fixed-Medium-R-Normal--24-ISO8859-1

Maybe, like this;

Emacs.font: fontset-BDF16_etl

*FontSet-1 -*-fixed-medium-r-normal-*-16-*-*-*-*-*-fontset-BDF16_etl,\
	ascii:-etl-fixed-medium-r-normal-*-16-*-iso8859-1

Regards,

---
  Masaru Nomiya        mail-to: nomiya at ttmy.ne.jp

   "No WIndows, no gains!" ..... "Why, I am wrong?"
        
                                        -- Bill --

**= Email   6 ==========================**

Date:  Sat, 23 Feb 2002 01:39:03 +0900
From:  Jun Sawataishi <jsawa at attglobal.net>
Subject:  Re:  Identifying sh.exe

>
>Is there any way to identify which version of shell is being used from 
>within a shell script?
>
>-- 
>John
How about dowing like this

if [ -n "$SH_VERSION" ] ; then
  echo "shell: $SH_VERSION"
fi
if [ -n "$KSH_VERSION" ] ; then
  echo "shell: $KSH_VERSION"
fi
if [ -n "$BASH_VERSION" ] ; then
  echo "shell: $BASH_VERSION"
fi
#   OS/2 is not a question, it's a solution.
#        SAWATAISHI Jun <jsawa at attglobal.net>
#   http://www2s.biglobe.ne.jp/~vtgf3mpr/


**= Email   7 ==========================**

Date:  Sat, 23 Feb 2002 01:42:21 +0900
From:  Jun Sawataishi <jsawa at attglobal.net>
Subject:  Re:  Re: Emacs customisation

Nomyya>>> I'm sorry, but I can't understand the meaning of "text mode".
JP>Can be run full screen, ie not PM. 
Run emacs with "-nw" option
   $ emacs -nw

#   OS/2 is not a question, it's a solution.
#        SAWATAISHI Jun <jsawa at attglobal.net>
#   http://www2s.biglobe.ne.jp/~vtgf3mpr/


**= Email   8 ==========================**

Date:  Sat, 23 Feb 2002 07:47:51 +0100 (CET)
From:  DWParsons at t-online.de (Dave Parsons)
Subject:  Re:  GNAT 3.14p

On Fri, 22 Feb 2002 16:12:28 +0000, John Poltorak wrote:
 
> Anyone looking at building GNAT 3.14p?

Yes

> Is this one based on gcc 3.0.3?

No

Dave


**= Email   9 ==========================**

Date:  Sat, 23 Feb 2002 08:39:34 -0800
From:  Neil Waldhauer <zonker at well.com>
Subject:  Re:  UnixOS/2 Build System - proposal

On Sat, 23 Feb 2002 15:58:52 +0000, John Poltorak <jp at eyup.org> wrote:

> The UnixOS/2 Build System

I'd like a sanity check as part of the build system to ensure that the
prerequisites are in place. The sanity check would check if an appropriate
shell, make, gcc and other parts were there before doing any building.

The checkinstall.cmd included with XFree86 finds environment problems with that
product. The UnixOS/2 Build System would benefit from something similar.

Neil

--
Neil Waldhauer, zonker at well.com

**= Email   10 ==========================**

Date:  Sat, 23 Feb 2002 10:20:19 +0000
From:  John Poltorak <jp at eyup.org>
Subject:   Migrating from gcc 2.8.1 to 3.0.3


Is it possible to move to gcc v3.0.3 without using any binaries from 
v2.8.1? Or is there a requirement to install over the top of 2.8.1?
I'm ignoring include and lib and any documentation, just the files which 
normally go in bin. I'm thinking about programs like NM, LISTOMF etc...
	

-- 
John 



**= Email   11 ==========================**

Date:  Sat, 23 Feb 2002 10:55:54 -0500
From:  Henry Sobotka <sobotka at axess.com>
Subject:  Re:  Migrating from gcc 2.8.1 to 3.0.3

John Poltorak wrote:
> 
> Is it possible to move to gcc v3.0.3 without using any binaries from
> v2.8.1? Or is there a requirement to install over the top of 2.8.1?
> I'm ignoring include and lib and any documentation, just the files which
> normally go in bin. I'm thinking about programs like NM, LISTOMF etc...

Yes. See emx/doc/install for details.

h~

**= Email   12 ==========================**

Date:  Sat, 23 Feb 2002 11:07:08 -0500
From:  Henry Sobotka <sobotka at axess.com>
Subject:  Re:  cpp0.exe error msg

John Poltorak wrote:
> 
> Can someone translate this error msg? :-

<snip>

There's nothing wrong with malloc.c nor cpp0.exe, but your gcc3
installation is somehow messed up. Doublecheck the instructions to make
sure you didn't miss anything.

h~

**= Email   13 ==========================**

Date:  Sat, 23 Feb 2002 12:00:23 +0000
From:  John Poltorak <jp at eyup.org>
Subject:  Re:  UnixOS/2 distro

On Fri, Feb 22, 2002 at 06:50:45AM -0500, Thomas Dickey wrote:
> On Fri, Feb 22, 2002 at 09:13:13AM +0000, John Poltorak wrote:
> > 
> > ftp://unixos2.com/pub/unixos2/unixos2-current/unixos2/d1/byacc.zip
> > 
> > it's just that I don't have a definitive location as the home of BYACC.
> 
> I'll make a small webpage for the copy of byacc that I use (am doing so now).
> 
> 	http://invisible-island.net/byacc/
> 	ftp://invisible-island.net/byacc/

Thanks, for that.

A couple of issues arise...

It comes with a Makefile rather than a configure script, which would be 
OK, but insists on using 'cc'. Could it be changed to use $(CC) instead?

Also it ends up creating a binary called YACC. Isn't it normally refered 
to as BYACC?


BTW if I run 'Make', I need to run EMXBIND subsequently to create an EXE.
Is there a way to specify an EXE on the Make command line? 
 
> -- 
> Thomas E. Dickey <dickey at invisible-island.net>
> http://invisible-island.net
> ftp://invisible-island.net


-- 
John


**= Email   14 ==========================**

Date:  Sat, 23 Feb 2002 12:48:04 -0500
From:  Thomas Dickey <dickey at herndon4.his.com>
Subject:  Re:  (un)protoize

On Sat, Feb 23, 2002 at 06:28:59PM +0100, Stefan Neis wrote:
> On Fri, 22 Feb 2002, John Poltorak wrote:
> 
> > > "The program protoize is an optional part of GCC. You can use it to add
> > > prototypes to a program, thus converting the program to ISO C in one
> > > respect. The companion program unprotoize does the reverse: it removes
> > > argument types from any prototypes that are found."
> > 
> > Can anyone build an OS/2 version of this program?
> 
> Why? I can see a use for those programs on certain old platforms where
> native compilers are unable to handle ANSI prototypes, but they seem
> completely useless on OS/2.

protoize's doubly useless because it doesn't preserve the parameter promotion
(in some cases it breaks the program).

-- 
Thomas E. Dickey <dickey at invisible-island.net>
http://invisible-island.net
ftp://invisible-island.net

**= Email   15 ==========================**

Date:  Sat, 23 Feb 2002 13:24:19 +0100
From:  Andreas Buening <andreas.buening at nexgo.de>
Subject:  Re:  make 3.79.1

John Poltorak wrote:
> 
> On Fri, Feb 22, 2002 at 10:53:37PM +0100, Andreas Buening wrote:
> > John Poltorak wrote:
> > >
> > > There was no %SHELL% or %MAKESHELL% variable set in the environment or in
> > > the Makefile.
> > >
> > > I set both on in turn, both with BASH and PDKSH as /bin/sh, and it made
> > > no difference. There was no '$SHELL changed' msg.
> >
> > Did you use "MAKESHELL=/bin/sh"? "export MAKESHELL=/bin/sh"
> > might be required.
> 
> Originally I only tried running make from a CMD prompt, but have just
> tried BASH and PDKSH, but the results are the same.

Okay, which package are you trying to compile (URL)?

[snip]

bye,
Andreas

-- 
One OS to rule them all, One OS to find them,
One OS to bring them all and in the darkness bind them
In the Land of Redmond where the Shadows lie.

**= Email   16 ==========================**

Date:  Sat, 23 Feb 2002 13:53:08 +0000
From:  John Poltorak <jp at eyup.org>
Subject:  Re:  make 3.79.1

On Sat, Feb 23, 2002 at 01:24:19PM +0100, Andreas Buening wrote:
 
> Okay, which package are you trying to compile (URL)?

EMACS - check the recent EMACS thread for build instructions.


This should do quite a bit of it, but there are a few manual edits:-


c:
md c:\tmp
md c:\tmp\archives
cd \tmp\archives

wget ftp://ftp.gnu.org/pub/gnu/emacs/emacs-20.7.tar.gz
wget ftp://ftp.gnu.org/pub/gnu/emacs/leim-20.7.tar.gz
wget http://www.asahi-net.or.jp/~ik3a-nsmr/emacs-20.7-KIT1.9.tgz

set SOMEWHERE=c:\tmp\archives
set INCLUDE=%C_INCLUDE_PATH%
set NAME=Richard Stallman

tar zxf %SOMEWHERE%\emacs-20.7.tar.gz
tar zxf %SOMEWHERE%\leim-20.7.tar.gz
cd emacs-20.7
chmod +rw -R *

rem 'tar zxf' does not work with patch kit

untgz %SOMEWHERE%\emacs-20.7-KIT1.9.tgz

rem 
rem     Edit configure.cmd 
rem 

pause

rem 
rem     Execution of configure.cmd. Enter name of PATCH program and press Enter key
rem 

pause

configure.cmd --with-bsd-socket --with-xf86sup-pty
make


 
> bye,
> Andreas
> 
> -- 
> One OS to rule them all, One OS to find them,
> One OS to bring them all and in the darkness bind them
> In the Land of Redmond where the Shadows lie.


-- 
John


**= Email   17 ==========================**

Date:  Sat, 23 Feb 2002 14:13:46 +0000
From:  John Poltorak <jp at eyup.org>
Subject:   cpp0.exe error msg


Can someone translate this error msg? :-


Finding dependencies for malloc.obj.
cpp0.exe: warning:  is shorter than expected
malloc.c:505:5: #endif without #if
malloc.c:524:2: #endif without #if

Exactly what 'is shorter than expeted'?

What do the numbers above signify? I guess 505 is the line no. Is 5 the 
position where the error occurs?

This is line 523-526 of malloc.c:-

static u_short buck_size[MAX_BUCKET_BY_TABLE + 1] = 
  { 
      0, 0, 0, 0, 4, 4, 8, 12, 16, 24, 32, 48, 64, 80,
  };


This msg occurs when using gcc 3.0.3 for building Perl. 

I have tried unsetting the paths which gcc 2.8.1 normally needs, as has 
been suggested, but this didn't make any difference.


-- 
John



**= Email   18 ==========================**

Date:  Sat, 23 Feb 2002 14:25:57 +0100 (CET)
From:  Stefan Neis <neis at cdc.informatik.tu-darmstadt.de>
Subject:  Re:  OSFree...

On Thu, 21 Feb 2002, Jack Troughton wrote:

> Kris Steenhaut wrote:
> > 
> > Jack Troughton schreef:
> > 
> > > Kris Steenhaut wrote:
> > > >
> > > > Mentore Siesto schreef:
> > > >
> > > > > KS >
> > > > > KS >Could only be done in court.
> > > > >
> > > > > But - I'm afraid - it CAN definitely be done.
> > > > >
> > > >
> > > > You damn well know you can't go to court in Europe about patented software, as
> > > > software can't be patented here. bsa seems to be working on it, but that's quite
> > > > another matter.
> > >
> > > It wouldn't be a patent issue; it would be a copyright issue. AFAIK, the
> > > common market is still signatory to the Berne Convention.
> > >
> > 
> > In a nutshell: programs can be copyrighted, source codes can't (The usa and Japan
> > being the exception).

You're confusing patent issues (you have to pay money if you develop your
own software using the same algorithm - e.g. any RSA implementation, no
matter what programming language and what author of the source code) and
copyright issues (you'll have to "pay" for stealing a specific existing
piece of source code, recompiling it and making it available).
The problem is to prove that it really is stolen source code and not
source code written by somebody else that just is intended to be
compatible.

 
> If there's no copyright on source code, how is the GPL enforceable? 

That probably was a particularly bad example. I doubt that this licence 
would really be enforceable. Anyway, this would make a really interesting
case. But yes, in principle you're absolutely correct.

> In short, the distinction you draw between "programs" and "source codes"
> is an artificial one at best...

Well, binaries are harder to modify. Somebody rename all variables and
functions, changes the indentation pattern, replaces some while by for
loops or vice versa and similar tricks. Now, have the original author
try to prove that it is stolen source code and not some new source code
which just succeeds exceptionally well in being compatible with the
original.
Do do the some kind of thing to a binary, you must know extremely well
how to handle assembler/disassembler...

	Regards,
		Stefan

-- 
Micro$oft is not an answer. It is a question. The answer is 'no'.


**= Email   19 ==========================**

Date:  Sat, 23 Feb 2002 14:36:23 -0500
From:  Henry Sobotka <sobotka at axess.com>
Subject:  Re:  Migrating from gcc 2.8.1 to 3.0.3

Michel SUCH wrote:
> 
> For those using GCC 3.03, note that the -funroll_loops option is broken
> (just experienced it with lame).

Do you mean -funroll-loops? Is the underscore instead of hyphen just a
typo in your message?

h~

**= Email   20 ==========================**

Date:  Sat, 23 Feb 2002 14:49:10 +0100 (CET)
From:  "Sebastian Wittmeier (ShadoW)" <Sebastian.Wittmeier at ginko.de>
Subject:  Re:  OSFree...

On Sat, 23 Feb 2002 14:25:57 +0100 (CET), Stefan Neis wrote:

>Do do the some kind of thing to a binary, you must know extremely well
>how to handle assembler/disassembler...

Writing a program that uses virus stealth technology (available on the
net) to automatically change a binary shouldn't be that difficult.

Sebastian



**= Email   21 ==========================**

Date:  Sat, 23 Feb 2002 15:58:52 +0000
From:  John Poltorak <jp at eyup.org>
Subject:   UnixOS/2 Build System - proposal

The UnixOS/2 Build System

The UnixOS/2 Build System consists of a standard set of directories and 
build routines, designed with the aim of being able to build any UnixOS/2
package without manual intervention simply by issuing the command:-

\> BUILD PKG

eg.

\> BUILD M4

would automatically build M4.

Directories:-

* archives

* lib

* logs

* patches

* scripts

* workdir


archives contains a copy of the original Open Source archive for a particular
app.

lib contains all the standard build scripts and configuration files for
performing builds (see below).

logs contains a copy of all msgs from the build of a particular app.

patches contains the OS/2 patches which need to be applied to a particular
app before it is built.

scripts contains app specific scripts which should be used when the standard
build script is not suitable.

workdir is the location for extracting a particular app from its archive and
building it.



The lib directory contains the following files:-


build.cmd
build.sh
config.site
build.table


BUILD.CMD - a cmd file used essentially as a wrapper to create the correct
build environment:-


 at echo off
mode co100,70
set path=c:\usr\bin;c:\emx\bin;c:\usr\local\bin;c:\os2;
set buildroot=c:/unixos2
set workdir=c:/unixos2/workdir
set home=c:/home/john
set buildlogs=%buildroot%/logs
set CONFIG_SITE=c:/unixos2/lib/config.site
set EMXROOT=c:
c:
cd \unixos2\lib
c:\bin\sh build.sh %1 2>&1 | tee %buildlogs%/build_%1.log 



BUILD.SH  - a shell script which determines sundry parameter for
a specific app using the BUILD.TABLE and then extracts the required archive
from the ARCHIVES directory into the WORKDIR directory, applies any patches
and then performs either an app-specific build, depending on whether a script
exists in the SCRIPTS directory, or the standard build:-


#! /bin/sh

PKG=$1
export PATH=`echo -E $PATH |tr '\\\\' '/'`
export BUILDROOT=`echo -E $BUILDROOT | tr '\\\\' '/'`

ARCHIVE=`grep ^$PKG build.table | awk -F"\t" '{print $2}'`
PARMS=`grep ^$PKG build.table | awk -F"\t" '{print $3}'`
SRC=`grep ^$PKG build.table | awk -F"\t" '{print $4}'`

cd $WORKDIR
tar zxvf $BUILDROOT/archives/$ARCHIVE.tar.gz
cd $ARCHIVE

if test -f $BUILDROOT/scripts/build_$1.cmd; then
  { 
  echo "using $1 specific Build script"
  export CONFIG_SITE=
  $BUILDROOT/scripts/build_$1.cmd
  exit 1
  }

fi

test -f $BUILDROOT/patches/$ARCHIVE.diff && patch -p 2 < $BUILDROOT/patches/$ARCHIVE.diff

if test $PKG != "autoconf"; then
  { 
   autoconf
  }

fi

./configure $PARMS

make

echo Build finished.





CONFIG.SITE  - contains all the standard environment variables used for a
build:-



#! /bin/sh
#
# config.site
#
UXRT=${UNIXROOT-c:}
SHELL="$UXRT/bin/sh"
CC=gcc
CXX=gcc
AWK=awk
LEX=flex
RANLIB=echo
#CXXFLAGS='-D__EMX__ -DOS2 -Zmtd -D__ST_MT_ERRNO__ -O2'

export ac_executable_extensions=".exe"

prefix=$UXRT/usr/local
EMX=${EMXROOT-$UXRT/emx}

C_INCLUDE_PATH=$EMX/include
CPLUS_INCLUDE_PATH=$EMX/include/cpp';'$EMX/include
LIBRARY_PATH=$EMX/lib



BUILD.TABLE - is simply a lookup table containing sundry parameters for
every app. This is something that is like to get more comprehensive over
time which will help to eliminate the need for app-specific build scripts:-

 
ncurses	ncurses-5.2		.
tin	tin-1.5.11		.
m4	m4-1.4	--without-included-regex --without-included-gt	src
sed	sed-3.02	--with-regex=-lregex	sed
grep	grep-2.4.2	 --without-included-regex --without-included-gettext	src






This is just a proposal and needs more refinement, but hopefully it provides
for an outline build system which could be used in all UnixOS/2 apps.
It would also allow anyone to build any app for themselves with a good chance 
of success if they downloaded the appropriate patches and scripts for an app
from a standard repository such as unixos2.com.


Comments, suggestions invited...

 


-- 
John


**= Email   22 ==========================**

Date:  Sat, 23 Feb 2002 16:56:56 -0500
From:  Henry Sobotka <sobotka at axess.com>
Subject:   gcc 3.0.3. _optlink questions for Andy

...which I'm posting here for the benefit of anyone else interested in
these issues.

Simple tests indicate that gcc's _optlink is compatible with VAC++'s
_Optlink both ways: I was able to create an _optlink DLL with gcc and
use it with an icc-built program, as well as run a gcc-built program
with a VAC++ version of glib.dll.

Is there any particular reason for the case difference between _optlink
and _Optlink?

How feasible would it be to add a -moptlink flag to make that calling
convention the default?

I noticed that _optlink only kicks in with optimization. Without -O or
higher, args are moved from the registers to the stack instead of used
in situ. Is this a bug or just an unfortunate necessity due to gcc
internals and implementation details or...?

h~

**= Email   23 ==========================**

Date:  Sat, 23 Feb 2002 17:17:50 +0000
From:  John Poltorak <jp at eyup.org>
Subject:  Re:  UnixOS/2 Build System - proposal

On Sat, Feb 23, 2002 at 08:39:34AM -0800, Neil Waldhauer wrote:
> On Sat, 23 Feb 2002 15:58:52 +0000, John Poltorak <jp at eyup.org> wrote:
> 
> > The UnixOS/2 Build System
> 
> I'd like a sanity check as part of the build system to ensure that the
> prerequisites are in place. The sanity check would check if an appropriate
> shell, make, gcc and other parts were there before doing any building.

Yes, that's a good idea. The trouble is knowing which utils to check...
There could be so many.

 
> The checkinstall.cmd included with XFree86 finds environment problems with that
> product. The UnixOS/2 Build System would benefit from something similar.

I'll check it out for a few ideas...

 
> Neil
> 
> --
> Neil Waldhauer, zonker at well.com


-- 
John



**= Email   24 ==========================**

Date:  Sat, 23 Feb 2002 18:28:59 +0100 (CET)
From:  Stefan Neis <neis at cdc.informatik.tu-darmstadt.de>
Subject:  Re:  (un)protoize

On Fri, 22 Feb 2002, John Poltorak wrote:

> > "The program protoize is an optional part of GCC. You can use it to add
> > prototypes to a program, thus converting the program to ISO C in one
> > respect. The companion program unprotoize does the reverse: it removes
> > argument types from any prototypes that are found."
> 
> Can anyone build an OS/2 version of this program?

Why? I can see a use for those programs on certain old platforms where
native compilers are unable to handle ANSI prototypes, but they seem
completely useless on OS/2.
 
	Regards,
		Stefan
-- 
Micro$oft is not an answer. It is a question. The answer is 'no'.


**= Email   25 ==========================**

Date:  Sat, 23 Feb 2002 19:28:04 +0100 (CET)
From:  "Michel SUCH" <msuch at free.fr>
Subject:  Re:  Migrating from gcc 2.8.1 to 3.0.3

Hi all,

For those using GCC 3.03, note that the -funroll_loops option is broken
(just experienced it with lame).

On Sat, 23 Feb 2002 10:55:54 -0500, Henry Sobotka wrote:

>John Poltorak wrote:
>> 
>> Is it possible to move to gcc v3.0.3 without using any binaries from
>> v2.8.1? Or is there a requirement to install over the top of 2.8.1?
>> I'm ignoring include and lib and any documentation, just the files which
>> normally go in bin. I'm thinking about programs like NM, LISTOMF etc...
>
>Yes. See emx/doc/install for details.
>
>h~
>
>

----------------------------
Michel SUCH TEAM OS/2 FRANCE	
ICQ # 51654489


**= Email   26 ==========================**

Date:  Sat, 23 Feb 2002 22:39:19 -0500
From:  Henry Sobotka <sobotka at axess.com>
Subject:  Re:  gcc 3.0.3. _optlink questions for Andy

Holger Veit wrote:
> 
> This sounds good. EM had reported long ago that the old 2.8.1 already had
> some sort of Optlink calling convention, but mentioned it would work
> for scalar types only. So, does the new optlink also work for the
> not so "simple cases" like float/double arguments (mixed with integer
> args), structure passing and returns, and pointers to obscure data types
> (like function pointers)?

Alas, no for mixed args. It just moves everything to the stack in the
prolog to a foo(int, float, double, int, double, float, long, int) call,
instead of pushing the last arg and loading the general and FP registers
as expected with the rest. (Why? - another implicit question for Andy.)
For structs I have look up how _Optlink handles them. Pointers may be
okay because IIRC they're all just treated as integral types that can go
into %eax, %edx or %ecx; nonetheless a must-test. It would be great if
we could get this working right.

h~

**= Email   27 ==========================**

Date:  Sat, 23 Feb 2002 22:51:19 -0500
From:  Henry Sobotka <sobotka at axess.com>
Subject:  Re:  cpp0.exe error msg

John Poltorak wrote:
> 
> Is there any way to pinpoint the error?

Can't think of a method offhand. But you might make sure that Andy's
gettext is being used instead of some other variant (I have no idea
if/how cpp0 uses it, just a possible source of strangeness that comes to
mind). Also, I gather you've run make in emx/lib and
emx/lib/gcc-lib/.../3.0.3 after unpacking?

h~

**= Email   28 ==========================**

Date:  Sat, 23 Feb 2002 22:54:06 -0500
From:  Henry Sobotka <sobotka at axess.com>
Subject:  Re:  Migrating from gcc 2.8.1 to 3.0.3

Michel SUCH wrote:
> 
> On Sat, 23 Feb 2002 14:36:23 -0500, Henry Sobotka wrote:
> 
> >Michel SUCH wrote:
> >>
> >> For those using GCC 3.03, note that the -funroll_loops option is broken
> >> (just experienced it with lame).
> >
> >Do you mean -funroll-loops? Is the underscore instead of hyphen just a
> >typo in your message?
> >
> Sorry, yes it is.

YAQFA (yet another question for Andy).

h~