Date: Tue, 25 Jan 2005 00:04:20 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600
Subject: [UnixOS2_Archive] No. 512

**************************************************
Monday 24 January 2005
 Number  512
**************************************************

Subjects for today
 
1  Re:  wildcard expansion, was: coreutils with gcc 3.3.5 b4 : John Poltorak <jp at warpix.org>
2  Re:  wildcard expansion, was: coreutils with gcc 3.3.5 b4 : lamikr <lamikr at cc.jyu.fi>
3   Perl help : John Poltorak <jp at warpix.org>
4  Re:  Perl help : Lyn St George" <lyn at zolotek.net>
5  Re:  Perl help : Stefan.Neis at t-online.de
6  Re:  Perl help : lsunley at mb.sympatico.ca
7  Re:  Perl help : John Poltorak <jp at warpix.org>
8  Re:  Perl help : lsunley at mb.sympatico.ca
9  Re:  Perl help : John Poltorak <jp at warpix.org>
10  Re:  Perl help : Lyn St George" <lyn at zolotek.net>
11  Re:  coreutils with gcc 3.3.5 b4 : Andy Willis <abwillis at comcast.net>
12  Re:  Perl help : John Poltorak <jp at warpix.org>
13  Re:  Perl help : Lyn St George" <lyn at zolotek.net>
14  Re:  Perl help : Stefan.Neis at t-online.de
15  Re:  coreutils with gcc 3.3.5 b4 : Andy Willis <abwillis at comcast.net>
16  Re:  Perl help : John Poltorak <jp at warpix.org>
17  Re:  Perl help : Steven Levine" <steve53 at earthlink.net>
18  Re:  Perl help : Lyn St George" <lyn at zolotek.net>
19   Locales and EURO qualifier : lsunley at mb.sympatico.ca
20  Re:  Perl help : Dave Yeo" <dave_yeo at paralynx.com>
21   Re: Xfree86 with libc : Dave Yeo" <dave_yeo at paralynx.com>
22  Re:  zlib : Dave Yeo" <dave_yeo at paralynx.com>
23  Re:  Locales and EURO qualifier : Stefan.Neis at t-online.de
24  Re:  Perl help : Stefan.Neis at t-online.de
25  Re:  Re: Xfree86 with libc : Henry Sobotka <sobotka at axess.com>
26  Re:  Locales and EURO qualifier : Knut St. Osmundsen" <bird at anduin.net>
27  Re:  wildcard expansion, was: coreutils with gcc 3.3.5 b4 : Andrew MacIntyre <andymac at bullseye.apana.org.au>
28   search order for libs with Innotek_libc : Dave Yeo" <dave_yeo at paralynx.com>
29  Re:  Re: Xfree86 with libc : Dave Yeo" <dave_yeo at paralynx.com>
30  Re:  search order for libs with Innotek_libc : Henry Sobotka <sobotka at axess.com>
31  Re:  search order for libs with Innotek_libc : Yuri Dario" <warp at os2power.com>
32  Re:  Re: Xfree86 with libc : Holger Veit <holger.veit at ais.fraunhofer.de>
33   please REMOVE me from this list. : Johannes Fahrenkrug <jfa at gebeco.de>
34  Re:  Perl help : Dave Saville" <dave at deezee.org>
35  Re:  Perl help : Dave Saville" <dave at deezee.org>
36  Re:  Perl help : John Poltorak <jp at warpix.org>
37  Re:  Perl help : Dave Saville" <dave at deezee.org>
38  Re:  Perl help : Dave Saville" <dave at deezee.org>


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

Date:  Sun, 23 Jan 2005 14:47:28 +0000
From:  John Poltorak <jp at warpix.org>
Subject:  Re:  wildcard expansion, was: coreutils with gcc 3.3.5 b4

On Sun, Jan 23, 2005 at 01:17:28PM +0100, Knut St. Osmundsen wrote:
> 
> > I don't know what the scope of programmes handling EAs might be, but
> > it would be great if we could automatically have it in tar, bzip2,
> > gzip, rsync, wget among others besides cp and mv.
> 
> Linux got EA support a while back. If stuff like coreutils starts using 
> it would make sense to implement that interface in LIBC too. The linux 
> guys got it from SGI it seems, so it's not a standarized api yet.
> 
> Anyway, look for these functions: [lf]*setxattr, [lf]*getxattr, 
> [lf]*listxattr and [lf]*removexattr.
> And these commands: attr, getfattr and setattr.
> 
> The BSD guys seems to be making a extattr_* api family. We'll have to 
> see which ends up being used.

Does that mean there is some remote possibility of something rsync 
having code for handling EA's in the original source?

I guess the best of handling this would be by creating some configure test 
which would would end up with config.h having something like

WANT_EA_SUPPORT=1

Just a thought...

> Kind Regards,
>   knut


-- 
John



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

Date:  Sun, 23 Jan 2005 16:51:08 +0200
From:  lamikr <lamikr at cc.jyu.fi>
Subject:  Re:  wildcard expansion, was: coreutils with gcc 3.3.5 b4

So if EA usage for functions like open would be optional and not used in 
the utils like
tar, bzip2, gzip, cp or mv then they would not understanding either 
wildcard extensions
or symlinks? And if that option is turned on then both symlinks and 
wildcards could be understood.

Mika

Stefan.Neis at t-online.de wrote:

>        Hi,
>
>  
>
>>What about Symlink extension that Martin made about one year ago.
>>Is that using EAs for storing symlink information.
>>    
>>
>
>Probably.
>
>  
>
>>Is symlink supported btw. in the Innoteks lib?
>>    
>>
>
>Yes. And it's using EAs for this purpose. AFAIK, it's even compatible
>with the way Linux creates symlinks on HPFS in the latest release.
>
>        Regards,
>	        Stefan	
>
>
>
>
>  
>



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

Date:  Sun, 23 Jan 2005 16:00:11 +0000
From:  John Poltorak <jp at warpix.org>
Subject:   Perl help




Can anyone tell me what this tests for:-  ?



if [ "$?" != "0" ]; then



Here is the context:-


# Create temp files directory
$perl "$srcdir/maketemp.pl"
if [ "$?" != "0" ]; then
	echo "ERROR: Failed to create or check temp files directory /tmp/.webmin"
	echo ""
	exit 2
fi




maketemp.pl:-

# maketemp.pl
# Create the /tmp/.webmin directory if needed

$tmp_dir = $ENV{'tempdir'} || "/tmp/.webmin";

while($tries++ < 10) {
	local  at st = lstat($tmp_dir);
	exit(0) if ($st[4] == $< && (-d _) && ($st[2] & 0777) == 0755);
	if ( at st) {
		unlink($tmp_dir) || rmdir($tmp_dir) ||
			system("/bin/rm -rf ".quotemeta($tmp_dir));
		}
	mkdir($tmp_dir, 0755) || next;
	chown($<, $(, $tmp_dir);
	chmod(0755, $tmp_dir);
	}
exit(1);




The directory /tmp/.webmin actually gets created so how do I change this 
code so that it shows that creating the directory has been successful?



-- 
John



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

Date:  Sun, 23 Jan 2005 16:21:53 +0000 (GMT)
From:  "Lyn St George" <lyn at zolotek.net>
Subject:  Re:  Perl help

On Sun, 23 Jan 2005 16:00:11 +0000, John Poltorak wrote:

>
>
>Can anyone tell me what this tests for:-  ?
>
>
>
>if [ "$?" != "0" ]; then
>

Tests for the result of a child process. 

if [ -d "tempdir" ]; then

does the equivalent but works better on my installations of perl (which
may well be faulty ..)

-
Lyn



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

Date:  Sun, 23 Jan 2005 17:21:35 +0100
From:  Stefan.Neis at t-online.de
Subject:  Re:  Perl help

        Hi,

> Can anyone tell me what this tests for:-  ?
>   
> if [ "$?" != "0" ]; then

IIRC, "$?" in shell scripts is the exit code of the
previous command, i.e. that thing tests if the previous
command was successfull (rc == 0) or not (rc!=0).

        Regards,
	        Stefan	





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

Date:  Sun, 23 Jan 2005 10:21:08 -0500
From:  lsunley at mb.sympatico.ca
Subject:  Re:  Perl help

In <20050123160011.M18360 at warpix.org>, on 01/23/05 
   at 04:00 PM, John Poltorak <jp at warpix.org> said:




>Can anyone tell me what this tests for:-  ?



>if [ "$?" != "0" ]; then

It tests the return code from the perl script.

>Here is the context:-


># Create temp files directory
>$perl "$srcdir/maketemp.pl"
>if [ "$?" != "0" ]; then
>	echo "ERROR: Failed to create or check temp files directory
>/tmp/.webmin"
>	echo ""
>	exit 2
>fi




>maketemp.pl:-

># maketemp.pl
># Create the /tmp/.webmin directory if needed

>$tmp_dir = $ENV{'tempdir'} || "/tmp/.webmin";

>while($tries++ < 10) {
>	local  at st = lstat($tmp_dir);
>	exit(0) if ($st[4] == $< && (-d _) && ($st[2] & 0777) == 0755);
>	if ( at st) {
>		unlink($tmp_dir) || rmdir($tmp_dir) ||
>			system("/bin/rm -rf ".quotemeta($tmp_dir));
>		}
>	mkdir($tmp_dir, 0755) || next;
>	chown($<, $(, $tmp_dir);
>	chmod(0755, $tmp_dir);
>	}
>exit(1);




>The directory /tmp/.webmin actually gets created so how do I change this 
>code so that it shows that creating the directory has been successful?





-- 
-----------------------------------------------------------
lsunley at mb.sympatico.ca
-----------------------------------------------------------



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

Date:  Sun, 23 Jan 2005 16:49:59 +0000
From:  John Poltorak <jp at warpix.org>
Subject:  Re:  Perl help

On Sun, Jan 23, 2005 at 10:21:08AM -0500, lsunley at mb.sympatico.ca wrote:
> In <20050123160011.M18360 at warpix.org>, on 01/23/05 
>    at 04:00 PM, John Poltorak <jp at warpix.org> said:
> 
> 
> 
> 
> >Can anyone tell me what this tests for:-  ?
> 
> 
> 
> >if [ "$?" != "0" ]; then
> 
> It tests the return code from the perl script.

In that case how do I change this code to work properly on OS/2:- ?

> ># maketemp.pl
> ># Create the /tmp/.webmin directory if needed
> 
> >$tmp_dir = $ENV{'tempdir'} || "/tmp/.webmin";
> 
> >while($tries++ < 10) {
> >	local  at st = lstat($tmp_dir);
> >	exit(0) if ($st[4] == $< && (-d _) && ($st[2] & 0777) == 0755);

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

I want an exit(0) here but I'm not sure what all the conditions mean. I 
doubt whether they apply under OS/2...

Can someone explain the checks being made?

> 
> -- 
> -----------------------------------------------------------
> lsunley at mb.sympatico.ca
> -----------------------------------------------------------
> 


-- 
John




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

Date:  Sun, 23 Jan 2005 11:09:22 -0500
From:  lsunley at mb.sympatico.ca
Subject:  Re:  Perl help

In <20050123164959.O18360 at warpix.org>, on 01/23/05 
   at 04:49 PM, John Poltorak <jp at warpix.org> said:

>On Sun, Jan 23, 2005 at 10:21:08AM -0500, lsunley at mb.sympatico.ca wrote:
>> In <20050123160011.M18360 at warpix.org>, on 01/23/05 
>>    at 04:00 PM, John Poltorak <jp at warpix.org> said:
>> 
>> 
>> 
>> 
>> >Can anyone tell me what this tests for:-  ?
>> 
>> 
>> 
>> >if [ "$?" != "0" ]; then
>> 
>> It tests the return code from the perl script.

>In that case how do I change this code to work properly on OS/2:- ?

>> ># maketemp.pl
>> ># Create the /tmp/.webmin directory if needed
>> 
>> >$tmp_dir = $ENV{'tempdir'} || "/tmp/.webmin";
>> 
>> >while($tries++ < 10) {
>> >	local  at st = lstat($tmp_dir);
>> >	exit(0) if ($st[4] == $< && (-d _) && ($st[2] & 0777) == 0755);

I think the $st[2] & 0777 == 0755 is testing the file ownership from the
results of a "stat" equivalent, you could probably ignore that. I think
the first one is testing to see if the directory exists as a directory.

Based on the code following the test

>	if ( at st) {
>		unlink($tmp_dir) || rmdir($tmp_dir) ||
>			system("/bin/rm -rf ".quotemeta($tmp_dir));
>		}
>	mkdir($tmp_dir, 0755) || next;
>	chown($<, $(, $tmp_dir);
>	chmod(0755, $tmp_dir);
>	}
>exit(1);

if the directory does not exist as a directory the code tries to remove
anything with the given name of the directory then create it...

But then I know very little about scripting or perl

>^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

>I want an exit(0) here but I'm not sure what all the conditions mean. I 
>doubt whether they apply under OS/2...

An exit(0) 
>Can someone explain the checks being made?

>> 
>> -- 
>> -----------------------------------------------------------
>> lsunley at mb.sympatico.ca
>> -----------------------------------------------------------
>> 




-- 
-----------------------------------------------------------
lsunley at mb.sympatico.ca
-----------------------------------------------------------



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

Date:  Sun, 23 Jan 2005 17:39:06 +0000
From:  John Poltorak <jp at warpix.org>
Subject:  Re:  Perl help

On Sun, Jan 23, 2005 at 04:21:53PM +0000, Lyn St George wrote:
> On Sun, 23 Jan 2005 16:00:11 +0000, John Poltorak wrote:
> 
> >
> >
> >Can anyone tell me what this tests for:-  ?
> >
> >
> >
> >if [ "$?" != "0" ]; then
> >
> 
> Tests for the result of a child process. 
> 
> if [ -d "tempdir" ]; then
> 
> does the equivalent but works better on my installations of perl (which
> may well be faulty ..)

I'm slowly working through some of your patches for Webmin and trying to 
figure them out. Ideally, there would be some way of making all this 
platform independent so it could 'just work'...

I find Perl particularly impenetrable an am amzed that you actually got 
Webmin working with so few changes.
 
> -
> Lyn
> 


-- 
John




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

Date:  Sun, 23 Jan 2005 18:14:11 +0000 (GMT)
From:  "Lyn St George" <lyn at zolotek.net>
Subject:  Re:  Perl help

On Sun, 23 Jan 2005 17:39:06 +0000, John Poltorak wrote:

>On Sun, Jan 23, 2005 at 04:21:53PM +0000, Lyn St George wrote:
>> On Sun, 23 Jan 2005 16:00:11 +0000, John Poltorak wrote:
>> 
>> >
>> >
>> >Can anyone tell me what this tests for:-  ?
>> >
>> >
>> >
>> >if [ "$?" != "0" ]; then
>> >
>> 
>> Tests for the result of a child process. 
>> 
>> if [ -d "tempdir" ]; then
>> 
>> does the equivalent but works better on my installations of perl (which
>> may well be faulty ..)
>
>I'm slowly working through some of your patches for Webmin and trying to 
>figure them out. Ideally, there would be some way of making all this 
>platform independent so it could 'just work'...
>
>I find Perl particularly impenetrable an am amzed that you actually got 
>Webmin working with so few changes.

I'm far from a top-gun at Perl, but I find it _very_ useful and hence
have to learn enough to achieve some results. 

I plan to put aside some time tonight so as to put out a new patch
that will do most stuff automagically. It should run a script over 
selected module.info files to add Warp 4, WSeB and eCS as
allowable platforms for Apache and similar. It should also return
most stuff and carry on with the setup without human interaction 
(I hope ..). Bits of this were done before for v1.121 (as I've discovered
by looking through my archives), but it all needs updating for v1.170.

Question: does anyone know what 'uname -sv' returns on eCS?
Or why it returns a version of 0.00 on WSeB (or does this only 
happen on the Aurora beta, which I'm running because I've lost
my Warp 4 CD)? 

>> -
>> Lyn
>> 
>
>
>-- 
>John
>
>
>
>


-
Lyn



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

Date:  Sun, 23 Jan 2005 11:27:59 -0700
From:  Andy Willis <abwillis at comcast.net>
Subject:  Re:  coreutils with gcc 3.3.5 b4

Stefan.Neis at t-online.de wrote:

>Knut St. Osmundsen schrieb:
>
>  
>
>>>Easier way to add -lsocket is to set LIBS=-lsocket
>>>before running  configure. Seems that really
>>>hostname() should be in libc not
>>>      
>>>
>> > libsocket since its in unistd.h.
>>    
>>
>
>I assume, that function should be gethostname?
>
>  
>
In the case of coreutils it was gethostname.
Andy


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

Date:  Sun, 23 Jan 2005 18:35:48 +0000
From:  John Poltorak <jp at warpix.org>
Subject:  Re:  Perl help

On Sun, Jan 23, 2005 at 06:14:11PM +0000, Lyn St George wrote:
> On Sun, 23 Jan 2005 17:39:06 +0000, John Poltorak wrote:
> 
> >On Sun, Jan 23, 2005 at 04:21:53PM +0000, Lyn St George wrote:
> >> On Sun, 23 Jan 2005 16:00:11 +0000, John Poltorak wrote:
> >> 
> >> >
> >> >
> >> >Can anyone tell me what this tests for:-  ?
> >> >
> >> >
> >> >
> >> >if [ "$?" != "0" ]; then
> >> >
> >> 
> >> Tests for the result of a child process. 
> >> 
> >> if [ -d "tempdir" ]; then
> >> 
> >> does the equivalent but works better on my installations of perl (which
> >> may well be faulty ..)
> >
> >I'm slowly working through some of your patches for Webmin and trying to 
> >figure them out. Ideally, there would be some way of making all this 
> >platform independent so it could 'just work'...
> >
> >I find Perl particularly impenetrable an am amzed that you actually got 
> >Webmin working with so few changes.
> 
> I'm far from a top-gun at Perl, but I find it _very_ useful and hence
> have to learn enough to achieve some results. 
> 
> I plan to put aside some time tonight so as to put out a new patch
> that will do most stuff automagically. It should run a script over 
> selected module.info files to add Warp 4, WSeB and eCS as
> allowable platforms for Apache and similar. It should also return
> most stuff and carry on with the setup without human interaction 
> (I hope ..). Bits of this were done before for v1.121 (as I've discovered
> by looking through my archives), but it all needs updating for v1.170.

Great. I'll look forward to it.

 
> Question: does anyone know what 'uname -sv' returns on eCS?
> Or why it returns a version of 0.00 on WSeB (or does this only 
> happen on the Aurora beta, which I'm running because I've lost
> my Warp 4 CD)? 

Why not just run 'uname' by itself? That returns 'OS/2'. (I don't know 
what eCS returns).

> -
> Lyn


-- 
John
 



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

Date:  Sun, 23 Jan 2005 18:54:16 +0000 (GMT)
From:  "Lyn St George" <lyn at zolotek.net>
Subject:  Re:  Perl help

On Sun, 23 Jan 2005 18:35:48 +0000, John Poltorak wrote:

> 
>> Question: does anyone know what 'uname -sv' returns on eCS?
>> Or why it returns a version of 0.00 on WSeB (or does this only 
>> happen on the Aurora beta, which I'm running because I've lost
>> my Warp 4 CD)? 
>
>Why not just run 'uname' by itself? That returns 'OS/2'. (I don't know 
>what eCS returns).

The standard Webmin setup asks for the platform first, then the version.
This allows for different flavours of Linux, and for us should allow for 
different flavours of OS/2 to be reported and then displayed on startup. 


-
Lyn



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

Date:  Sun, 23 Jan 2005 20:14:04 +0100
From:  Stefan.Neis at t-online.de
Subject:  Re:  Perl help

        Hi,
  
> > Question: does anyone know what 'uname -sv' returns on eCS?

Here, I get "OS/2 2.45" (eCS 1.0), I suppose it'll be the same for
later versions.

        Regards,
	        Stefan	







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

Date:  Sun, 23 Jan 2005 12:21:14 -0700
From:  Andy Willis <abwillis at comcast.net>
Subject:  Re:  coreutils with gcc 3.3.5 b4

Knut St. Osmundsen wrote:

>
>
> Dave Yeo wrote:
>
>> On Thu, 20 Jan 2005 14:58:09 -0700, Andy Willis wrote:
>>
>>
>>> I ran configure and edited the makefile in src to add -Zomf to
>>> cflags and -lsocket to hostname ldadd.  I also had to add for 5.3.0
>>>  #include <sys\syslimits.h>   in lib\chdir-long.c or you could
>>> define PATH_MAX to 260 which it is in syslimits.h.
>>
>>
>> I wonder where PATH_MAX is usually defined, I've had to add
>> <sys/syslimits.h> a few times now. 
>
>
> That's a bit weird. We're defining PATH_MAX in sys/syslimits.h like 
> BSD. We have the same limits.h as BSD, which at the end includes 
> sys/syslimits.h if the sources aren't strict ANSI or C99. Are 
> coreutils now built with -ansi perhaps?
>
Coreutils isn't using -ansi but as far as I know it is not defining 
__POSIX_VISIBLE either.  Where should __POSIX_VISIBLE get defined?
Andy


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

Date:  Sun, 23 Jan 2005 20:07:49 +0000
From:  John Poltorak <jp at warpix.org>
Subject:  Re:  Perl help

On Sun, Jan 23, 2005 at 06:54:16PM +0000, Lyn St George wrote:
> On Sun, 23 Jan 2005 18:35:48 +0000, John Poltorak wrote:
> 
> > 
> >> Question: does anyone know what 'uname -sv' returns on eCS?
> >> Or why it returns a version of 0.00 on WSeB (or does this only 
> >> happen on the Aurora beta, which I'm running because I've lost
> >> my Warp 4 CD)? 
> >
> >Why not just run 'uname' by itself? That returns 'OS/2'. (I don't know 
> >what eCS returns).
> 
> The standard Webmin setup asks for the platform first, then the version.
> This allows for different flavours of Linux, and for us should allow for 
> different flavours of OS/2 to be reported and then displayed on startup. 

AIUI Webmin tries to work out what platform it is running on by parsing 
os_list.txt comparing the results of the OS test in the fifth field of a 
line with the first field of the line, and it is only if it can't 
determine the OS that it asks.

I may have got that wrong but that is what it sounds like although I 
haven't figured out which bit of code does this...
 
> 
> -
> Lyn
> 


-- 
John



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

Date:  Sun, 23 Jan 2005 12:16:43 -0800
From:  "Steven Levine" <steve53 at earthlink.net>
Subject:  Re:  Perl help

In <20050123164959.O18360 at warpix.org>, on 01/23/05
   at 04:49 PM, John Poltorak <jp at warpix.org> said:

Here's an commented version.

FWIW, this is really terrible code and would fail miserably if run with -w
and/or use strict.

# generate dir name
$tmp_dir = $ENV{'tempdir'} || "/tmp/.webmin";

# try 10 times - I really hate uninitialized variables
while($tries++ < 10) {
	local  at st = lstat($tmp_dir);
       # if we are owner
       # and if this a directory (_ is a file handle)
       # and if xrw for owner and rw for everyone else
       # the exit should really be inside the if because if lstat
       # fails  at st will be an empty list
	exit(0) if ($st[4] == $< && (-d _) && ($st[2] & 0777) == 0755);
       # if stat found file/dir
	if ( at st) {
              # try to delete file and then directory the easy way
              # othewise let the system try the hard way
		unlink($tmp_dir) || rmdir($tmp_dir) ||
			system("/bin/rm -rf ".quotemeta($tmp_dir));
		}
       # try to create, loop if fail
	mkdir($tmp_dir, 0755) || next;
       # set owner/group
       # looks redundant to me
       # will fail if not already owner, unless root
	chown($<, $(, $tmp_dir);
       # set permissions - i guess just in case umask is not set to 0755
	chmod(0755, $tmp_dir);
	}
exit(1);

BTW, if you don't use perl regularly, you really need a cheat sheet to
keep special variables and data types straight.  I find perl extremely
useful, but I will never clain it's easy for the casual user to deal with.

HTH,

Steven


-- 
----------------------------------------------------------------------
"Steven Levine" <steve53 at earthlink.net>  MR2/ICE 2.60b #10183 Warp4/FP15/14.100c_W4
www.scoug.com irc.fyrelizard.com #scoug (Wed 7pm PST)
----------------------------------------------------------------------



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

Date:  Sun, 23 Jan 2005 20:59:00 +0000 (GMT)
From:  "Lyn St George" <lyn at zolotek.net>
Subject:  Re:  Perl help

On Sun, 23 Jan 2005 20:07:49 +0000, John Poltorak wrote:

>> >Why not just run 'uname' by itself? That returns 'OS/2'. (I don't know 
>> >what eCS returns).
>> 
>> The standard Webmin setup asks for the platform first, then the version.
>> This allows for different flavours of Linux, and for us should allow for 
>> different flavours of OS/2 to be reported and then displayed on startup. 
>
>AIUI Webmin tries to work out what platform it is running on by parsing 
>os_list.txt comparing the results of the OS test in the fifth field of a 
>line with the first field of the line, and it is only if it can't 
>determine the OS that it asks.
>
>I may have got that wrong but that is what it sounds like although I 
>haven't figured out which bit of code does this...

I've just revisited that on Linux, and on both a RH and an LFS system
if fails to do so automatically - it's a minor point so I'll leave it.

ftp://ftp.zolotek.net/os2 has  webmin-1.170.os2.diff and 
run_update_os.pl, which should do the job.  Apply the patch
and setup.sh will call run_update_os.pl. Setup still requires
human interaction, but it has a list of OS/2 and eCS versions to
pick from. It also updates selected module.info files with 
'Warp 4 WSeB eCS' appended to any existing string so that these
platforms are allowed to use those modules. You will need to add 
extra ones probably, but this file can be run by itself any time 
after the main installation. This patch works on my WSeB system
in a clean installation directory. The CPAN module is still a PITA 
though ...
 


-
Lyn



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

Date:  Sun, 23 Jan 2005 14:50:46 -0500
From:  lsunley at mb.sympatico.ca
Subject:   Locales and EURO qualifier

Hi all

I made some changes to the LIBCb4 code to handle the  at EURO qualifier on
locales (like de_DE at EURO) and did some fixups to the POSIX and C locales
to return the mon_decimal_point mon_thousands_sep and a couple of other
things.

Now I need a good reference to what the  at EURO qualifier should return for

Numeric treatment (in locales like de_DE non euro decimal_point is "," and
thousands is ".") decimal_point (EURO) = ? (in Linux it is ".")
thousands_sep (EURO) = ? (in Linux it is ",")

Monetary treatment

mon_decimal_point = "." ?
mon_thousands_sep = "," ?
currency_symbol = (weird e character - does anyone know the numeric
value?) int_curr_symbol = "EUR" ?

Many thanks...

Lorne

-- 
-----------------------------------------------------------
lsunley at mb.sympatico.ca
-----------------------------------------------------------



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

Date:  Sun, 23 Jan 2005 13:26:41 -0800
From:  "Dave Yeo" <dave_yeo at paralynx.com>
Subject:  Re:  Perl help

On Sun, 23 Jan 2005 18:14:11 +0000 (GMT), Lyn St George wrote:

>
>Question: does anyone know what 'uname -sv' returns on eCS?
>Or why it returns a version of 0.00 on WSeB (or does this only 
>happen on the Aurora beta, which I'm running because I've lost
>my Warp 4 CD)? 

Don't know about eCS but I have various unames here on Warp v4 fp#15 and get 2 
different results depending on which uname
OS/2 2.45
OS/2 4.50
which are both correct. The question becomes which is right?
I understand that there are 2 OS/2 v3, Warp v3 and OS/2 NT v3 and when the divorce 
ahppened MS got to keep OS/2 v3+ and IBM was stuck at v2+ so officially I run 2.45. 
Also if you think about it there hasn't been any large changes since OS/2 v2, just a lot of 
incremental upgrades unlike the jump from 1.x (16 bit) to 2.x (32 bit) to NT (whole new 
kernel and architecture).
Dave
ps for those that argue that NT isn't OS/2 I have a couple of Byte Magazines around 
somewhere. One describes MS getting OS/2 NT v3 up and running on (IIRC) MIPs or 
on of those Rxxx processers. Another describes MS getting the Presentation Manager 
running on NT (3.51?) so they were ready incase Windows lost out to OS/2.



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

Date:  Sun, 23 Jan 2005 14:15:36 -0800
From:  "Dave Yeo" <dave_yeo at paralynx.com>
Subject:   Re: Xfree86 with libc

On Fri, 21 Jan 2005 10:57:47 -0800, Dave Yeo wrote:

>On Fri, 21 Jan 2005 01:30:41 -0500, Henry Sobotka wrote:
>
>>Dave Yeo wrote:
>>> 
>>> I wonder if it would better to go with _foo=foo or foo=_foo or just plain old foo.
>>
>>As I recall, the grammar is externalname=internalname so it has to be 
>>_foo=foo if the program you're linking from wants _foo.
>
>Here it is actually LD that wants _foo instead of foo. Anyways I changed makedef.cmd 
>to output foo=_foo and LD seems happy. Maybe this gothru will actually produce a 
>working XFree86. Unluckily it takes about 12 hrs for the build to happen here.
>Dave
>

Ok this is getting very frustrating. First problem was along the lines
of this
emxbind: symbol XSaveContext undefined (EXPORTS)
So I changed makedef.cmd to add an underline to all exports. This fixed
the above problem.
Now the problem is this
gcc -Zdll -Zmtd -Zmap -Zbin-files -o `cat ../../lib/Xext/dll.name`
globals.o extutil.o XMultibuf.o XShape.o XShm.o
		MITMisc.o XTestExt1.o XSync.o Xdbe.o XLbx.o
		XSecurity.o XAppgroup.o Xcup.o DPMS.o XEVI.o 
-L../../exports/lib  -lX11 -lshm -lsocket -lbsd -lgcc Xexto_os2.def
extutil.o: Undefined symbol _XFree referenced from text segment
... (lots more undefined symbols from libX11)

Now this has me stumped as all these symbols seem to be in X11.dll. eg
doing emximp -o X11.def X11.dll and looking at X11.def I see
  "_XFree"  at 419

So the question becomes why does one DLL not link correctly with
another DLL
Dave




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

Date:  Sun, 23 Jan 2005 14:15:04 -0800
From:  "Dave Yeo" <dave_yeo at paralynx.com>
Subject:  Re:  zlib

--_=_=_=IMA.BOUNDARY.IASAT4138764=_=_=_
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

On Sat, 22 Jan 2005 19:38:30 +0000, John Poltorak wrote:

>> Also I'd send the makefiles but I think the list is stripping attachments now
>
>If you have any patches or a build script and it is getting stripped off, 
>can you send it to the UX2BS list?
>
>I really need to get ZLIB included in UX2BS as a an uptodate working app 
>as it's so widely used by so many programmes.

I'm still in conflict about whether to use Z.DLL (with ords, backward
compatible with lots of apps, especially in X) or ZLIB1.DLL (without
ords, recommended by the Z developers). Anyways heres the makefile and
def to build the ord version. I need to get back to the zlib1 version,
got sidetracked with X <gr>
I'm also somewhat in conflict wether we should even go the DLL route.
DLLs are nice for saving disk space but that doesn't seem to be a
problem anymore. Static libs can actually save memory (at least in the
common arena) as the lib is in the private process space if I
understand correctly. For a static lib configure works fine if shm is
not installed. Probally shouldn't use shm here as it is limited to 4
MBs and am not sure how much ZLIB uses. On systems wiht shm installed
the configure script most likely needs to be hacked a bit to not try to
load shm. (it only tests for the header and we also need the dll
loaded)
Dave
ps the zlib1.dll makefile should be selfupdating so should work at
least until zlib2

--_=_=_=IMA.BOUNDARY.IASAT4138764=_=_=_
Content-Type: application/octet-stream; name="zos2.def"
Content-Transfer-Encoding: base64
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--_=_=_=IMA.BOUNDARY.IASAT4138764=_=_=_
Content-Type: application/octet-stream; name="Makefile.emx"
Content-Transfer-Encoding: base64
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--_=_=_=IMA.BOUNDARY.IASAT4138764=_=_=_--


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

Date:  Sun, 23 Jan 2005 23:36:10 +0100
From:  Stefan.Neis at t-online.de
Subject:  Re:  Locales and EURO qualifier

        Hi,

> Now I need a good reference to what the  at EURO qualifier should return for

IMHO, that should be the same as without the qualifier, the whole point
really being the difference in fonts (without EURO, you can use standard
ISO-8859-1 fonts (which is missing that "weird e character"), with EURO,
you want something containing that symbol (ISO-8859-15 ?).
But then, on OS/2 all those font considerations seem like a non-issue,
mostly (as long as you don't use X).

> Numeric treatment (in locales like de_DE non euro decimal_point is ","
> and thousands is ".")

Yes, and with euro, it should be exactly the same. And for monetary
stuff as well.

> decimal_point (EURO) = ? (in Linux it is ".")
> thousands_sep (EURO) = ? (in Linux it is ",")

I don't think I agree with Linux here. 

> Monetary treatment
>   
> mon_decimal_point = "." ?
> mon_thousands_sep = "," ?

No, same as above, i.e. the other way round.

> currency_symbol = (weird e character - does anyone know the numeric
> value?

213 for codepage 850, no idea about ISO fonts or other code pages.

> int_curr_symbol = "EUR" ?

Yes.

        Regards,
	        Stefan	




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

Date:  Sun, 23 Jan 2005 23:50:28 +0100
From:  Stefan.Neis at t-online.de
Subject:  Re:  Perl help

        Hi,

> Don't know about eCS but I have various unames here on Warp v4 fp#15 and get 2 
> different results depending on which uname
> OS/2 2.45
> OS/2 4.50
> which are both correct. 

I suppose you're not going to see anything different with eCS - at least
not with eCS 1.0

> I understand that there are 2 OS/2 v3, Warp v3 and OS/2 NT v3 and when the divorce 
> ahppened MS got to keep OS/2 v3+ and IBM was stuck at v2+

This seems somehow wrong, given that that divorce happened between OS/2
1.3 and OS/2 2.0 ...

> so officially I run 2.45. 

IIRC, the only reason for that numbering scheme is that there is enough
software which would refuse to start on any version >2, since the logic
back at the time when that software was written was that the major
version number only changes on non-backward compatible changes of the
API, not at the whim of some marketing people.
Older versions of autoconf's config guess for example would not recognize
OS/2 if the version number wasn't 2.xy. I remember how much fun it was
to track down some build failures on other peoples' systems to a "bad"
version of uname (which returned 3.something or 4.something as version
number) and then find out what config.guess actually did do in such a
case. :-(


If you look at the output of "ver", that other form (4.50) seems to be
the one that's politically correct, though.

> Also if you think about it there hasn't been any large changes since OS/2 v2, just a lot of 
> incremental upgrades unlike the jump from 1.x (16 bit) to 2.x (32 bit) to NT (whole new 
> kernel and architecture).

> ps for those that argue that NT isn't OS/2

Well, at least the numbering seems _very_ strange.

> I have a couple of Byte Magazines around 
> somewhere. One describes MS getting OS/2 NT v3 up and running on (IIRC) MIPs or 
> on of those Rxxx processers. Another describes MS getting the Presentation Manager 
> running on NT (3.51?) so they were ready incase Windows lost out to OS/2.  

IIRC, that's similar to Win32(s) apps running on NT, i.e. they included
enough of the competing API to be able to run programs (at most after
some recompilation), but that doesn't necessarily imply that the basis
really is identical...

        Regards,
	        Stefan	




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

Date:  Sun, 23 Jan 2005 18:00:53 -0500
From:  Henry Sobotka <sobotka at axess.com>
Subject:  Re:  Re: Xfree86 with libc

Dave Yeo wrote:
>
> Now the problem is this
> gcc -Zdll -Zmtd -Zmap -Zbin-files -o `cat ../../lib/Xext/dll.name`
> globals.o extutil.o XMultibuf.o XShape.o XShm.o
> 		MITMisc.o XTestExt1.o XSync.o Xdbe.o XLbx.o
> 		XSecurity.o XAppgroup.o Xcup.o DPMS.o XEVI.o 
> -L../../exports/lib  -lX11 -lshm -lsocket -lbsd -lgcc Xexto_os2.def
> extutil.o: Undefined symbol _XFree referenced from text segment
> .. (lots more undefined symbols from libX11)
> 
> Now this has me stumped as all these symbols seem to be in X11.dll. eg
> doing emximp -o X11.def X11.dll and looking at X11.def I see
>   "_XFree"  at 419
> 
> So the question becomes why does one DLL not link correctly with
> another DLL

Usually that means libX11 isn't being found or is invalid (e.g. OMF when 
a.out is expected), but you should see a message to that effect just 
before the slew of undefined symbols.

h~


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

Date:  Mon, 24 Jan 2005 00:21:09 +0100
From:  "Knut St. Osmundsen" <bird at anduin.net>
Subject:  Re:  Locales and EURO qualifier

Lorne,

I already answered some of this in reply to a private mail. But since 
you also ask here and get different answer I'll repeate myself.

lsunley at mb.sympatico.ca wrote:
> Hi all
> 
> I made some changes to the LIBCb4 code to handle the  at EURO qualifier on
> locales (like de_DE at EURO) and did some fixups to the POSIX and C locales
> to return the mon_decimal_point mon_thousands_sep and a couple of other
> things.

In the C/POSIX locale everything is 127 or "" except for  decimal_point 
which is ".". Check with your favourite Unix or the SuS.

> Now I need a good reference to what the  at EURO qualifier should return for
> 
> Numeric treatment (in locales like de_DE non euro decimal_point is "," and
> thousands is ".") decimal_point (EURO) = ? (in Linux it is ".")
> thousands_sep (EURO) = ? (in Linux it is ",")
> 
> Monetary treatment
> 
> mon_decimal_point = "." ?
> mon_thousands_sep = "," ?

Take these from the base locale, i.e de_DE in the example. Unless 
someone can document that there is a standardized EU way of doing 
grouping, decimal and separators. (They're probably the same in most of 
EU anyway - as set above.)

More important are probably the char fields describing fractional 
digits, and negative/positivi stuff. I guess it would be safe to just 
take the LANG=de_DE values from linux in this case.

> currency_symbol = (weird e character - does anyone know the numeric
> value?)
 > int_curr_symbol = "EUR" ?

"U+20AC EURO SIGN" is the unicode point you're thinking of. But, Linux 
de_DE is using "EUR" in *both* cases - that certainly simplyfies 
codepage conversions. Good idea to stick with that.

Kind Regards,
  knut


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

Date:  Mon, 24 Jan 2005 09:43:27 +1000
From:  Andrew MacIntyre <andymac at bullseye.apana.org.au>
Subject:  Re:  wildcard expansion, was: coreutils with gcc 3.3.5 b4

John Poltorak wrote:

> I don't know what the scope of programmes handling EAs might be, but it 
> would be great if we could automatically have it in tar, bzip2, gzip, 
> rsync, wget among others besides cp and mv.

Realistically I don't expect to see EA support in either bzip2 or gzip,
as they are both stream compressors rather than archivers.

EA support really needs some archive format, eg tar/cpio (which can be
then fed to a stream compressor) or zip.

I suspect that you are also unlikely to get EA support added to wget 
either, as the protocol (http/ftp/gopher) has no notion of EAs in its 
handling of "files".  New/extended protocols will be needed to add 
metadata support, which could then be picked up by all web browsers 
(wget/cURL basically being headless web browsers).

-------------------------------------------------------------------------
Andrew I MacIntyre                     "These thoughts are mine alone..."
E-mail: andymac at bullseye.apana.org.au  (pref) | Snail: PO Box 370
         andymac at pcug.org.au             (alt) |        Belconnen ACT 2616
Web:    http://www.andymac.org/               |        Australia


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

Date:  Sun, 23 Jan 2005 16:43:40 -0800
From:  "Dave Yeo" <dave_yeo at paralynx.com>
Subject:   search order for libs with Innotek_libc

Given both libfoo.a and foo.a and something like gcc bar.c -lfoo which lib gets found first?
Dave



**= Email   29 ==========================**

Date:  Sun, 23 Jan 2005 17:12:31 -0800
From:  "Dave Yeo" <dave_yeo at paralynx.com>
Subject:  Re:  Re: Xfree86 with libc

On Sun, 23 Jan 2005 18:00:53 -0500, Henry Sobotka wrote:

>> So the question becomes why does one DLL not link correctly with
>> another DLL
>
>Usually that means libX11 isn't being found or is invalid (e.g. OMF when 
>a.out is expected), but you should see a message to that effect just 
>before the slew of undefined symbols.

The problem seems to be my laziness and perhaps a change in search
order for libfoo.a and foo.a. I hadn't finished fixing up makedef.cmd
so only X11.a got the underlines and not libX11.a. X11.a is the import
library with ordinals and libX11.a with no ordinals. The dlls are built
with ordinals
Now to try to decide if its better to use foo=_foo or just _foo. Also
whether residentname should be added
Dave



**= Email   30 ==========================**

Date:  Sun, 23 Jan 2005 21:27:11 -0500
From:  Henry Sobotka <sobotka at axess.com>
Subject:  Re:  search order for libs with Innotek_libc

Dave Yeo wrote:
> Given both libfoo.a and foo.a and something like gcc bar.c -lfoo which lib gets found first?

[C:\temp]gcc -Zexe whichlib.c -Lc:/temp -lfoo

[C:\temp]whichlib
libfoo.a found

h~


**= Email   31 ==========================**

Date:  Mon, 24 Jan 2005 09:13:10 +0100 (CET)
From:  "Yuri Dario" <warp at os2power.com>
Subject:  Re:  search order for libs with Innotek_libc

Hi,

>Given both libfoo.a and foo.a and something like gcc bar.c -lfoo which lib gets found first?

see release notes.


Bye,

	Yuri Dario

/*
 * member of TeamOS/2 - Italy
 * http://www.os2power.com/yuri
 * http://www.teamos2.it
 */


**= Email   32 ==========================**

Date:  Mon, 24 Jan 2005 09:27:19 +0100
From:  Holger Veit <holger.veit at ais.fraunhofer.de>
Subject:  Re:  Re: Xfree86 with libc

Dave Yeo wrote:

>On Sun, 23 Jan 2005 18:00:53 -0500, Henry Sobotka wrote:
>
>  
>
>>>So the question becomes why does one DLL not link correctly with
>>>another DLL
>>>      
>>>
>>Usually that means libX11 isn't being found or is invalid (e.g. OMF when 
>>a.out is expected), but you should see a message to that effect just 
>>before the slew of undefined symbols.
>>    
>>
>
>The problem seems to be my laziness and perhaps a change in search
>order for libfoo.a and foo.a. I hadn't finished fixing up makedef.cmd
>so only X11.a got the underlines and not libX11.a. X11.a is the import
>library with ordinals and libX11.a with no ordinals. The dlls are built
>with ordinals
>Now to try to decide if its better to use foo=_foo or just _foo. Also
>whether residentname should be added
>Dave
>  
>
Warning: there are symbols in X11 which regularly already have a leading 
underscore. IIRC, this holds for Xfree()
in particular - there is also an _Xfree(). Consequently, this has to be 
modified to __Xfree, otherwise you'll get an insane mess of _Xfree and 
Xfree which leads to a recursion in the application code. BTDT.

Holger



**= Email   33 ==========================**

Date:  Mon, 24 Jan 2005 09:46:12 GMT
From:  Johannes Fahrenkrug <jfa at gebeco.de>
Subject:   please REMOVE me from this list.

Hi everyone.

I wrote a buttload of emails to this list with all kinds of remove 
commands and pleas... but I'm still on it. I'd really like to be 
removed from the list, I just don't have time to read it anymore.

Thanks a lot!

Johannes.





**= Email   34 ==========================**

Date:  Mon, 24 Jan 2005 09:11:38 +0000 (GMT)
From:  "Dave Saville" <dave at deezee.org>
Subject:  Re:  Perl help

On Sun, 23 Jan 2005 20:14:04 +0100, Stefan.Neis at t-online.de wrote:

>        Hi,
>  
>> > Question: does anyone know what 'uname -sv' returns on eCS?
>
>Here, I get "OS/2 2.45" (eCS 1.0), I suppose it'll be the same for
>later versions.

Same on 1.14

-- 
Regards

Dave Saville



**= Email   35 ==========================**

Date:  Mon, 24 Jan 2005 09:23:36 +0000 (GMT)
From:  "Dave Saville" <dave at deezee.org>
Subject:  Re:  Perl help

On Sun, 23 Jan 2005 16:49:59 +0000, John Poltorak wrote:

> >	local  at st = lstat($tmp_dir);
>> >	exit(0) if ($st[4] == $< && (-d _) && ($st[2] & 0777) == 0755);
>
>^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
>
>I want an exit(0) here but I'm not sure what all the conditions mean. I 
>doubt whether they apply under OS/2...

The first line puts the result of lstat on tmp_dir into an array.
The second checks two values from that array 5th & 3rd (zero offset).
 
$st[4] = uid.
 $st[2] = mode access flags.
$< = Effective user id.
-d _  Not sure what's going on here - -d is a test on the last 'stat'ed file
for directory but that underscore is confusing me - sure it was not $_ ?

So: exit code zero if userid = effective user id and something to do with a
directory and the access permissions are 0755.
 
Look up lstat to see the full list of returned values.

HTH

-- 
Regards

Dave Saville



**= Email   36 ==========================**

Date:  Mon, 24 Jan 2005 11:16:03 +0000
From:  John Poltorak <jp at warpix.org>
Subject:  Re:  Perl help

On Sun, Jan 23, 2005 at 08:59:00PM +0000, Lyn St George wrote:
> On Sun, 23 Jan 2005 20:07:49 +0000, John Poltorak wrote:
> 
> >> >Why not just run 'uname' by itself? That returns 'OS/2'. (I don't know 
> >> >what eCS returns).
> >> 
> >> The standard Webmin setup asks for the platform first, then the version.
> >> This allows for different flavours of Linux, and for us should allow for 
> >> different flavours of OS/2 to be reported and then displayed on startup. 
> >
> >AIUI Webmin tries to work out what platform it is running on by parsing 
> >os_list.txt comparing the results of the OS test in the fifth field of a 
> >line with the first field of the line, and it is only if it can't 
> >determine the OS that it asks.
> >
> >I may have got that wrong but that is what it sounds like although I 
> >haven't figured out which bit of code does this...
> 
> I've just revisited that on Linux, and on both a RH and an LFS system
> if fails to do so automatically - it's a minor point so I'll leave it.

It's this bit of Perl code in oschooser.pl:-

	foreach $o ( at list) {
		if ($o->[4] && eval $o->[4]) {
			# Got a match!
			$ver = $o;
			last;
			}
		}



Can't figure out the condition taht 'if' is checking at all...
 
> ftp://ftp.zolotek.net/os2 has  webmin-1.170.os2.diff and 
> run_update_os.pl, which should do the job.  

Great. I've got this now and am testing it out.

> Apply the patch
> and setup.sh will call run_update_os.pl. Setup still requires
> human interaction, but it has a list of OS/2 and eCS versions to
> pick from. It also updates selected module.info files with 
> 'Warp 4 WSeB eCS' appended to any existing string so that these
> platforms are allowed to use those modules. You will need to add 
> extra ones probably, but this file can be run by itself any time 
> after the main installation. This patch works on my WSeB system
> in a clean installation directory. The CPAN module is still a PITA 
> though ...

Does the Perl CPAN module work for you? I tried one with an older version 
of Perl and it seemed to work, so I'm wondering whether there is something 
wrong with my Perl build rather than any CPAN module.   
 
 
> -
> Lyn
> 


-- 
John



**= Email   37 ==========================**

Date:  Mon, 24 Jan 2005 11:45:52 +0000 (GMT)
From:  "Dave Saville" <dave at deezee.org>
Subject:  Re:  Perl help

On Mon, 24 Jan 2005 09:23:36 +0000 (GMT), Dave Saville wrote:

>-d _  Not sure what's going on here - -d is a test on the last 'stat'ed file
>for directory but that underscore is confusing me - sure it was not $_ ?

Sorry - Wrong again :-) file test operators work as in -test <file>. It is the
underscore that refers it to the last file used in stat, lstat or file.

-- 
Regards

Dave Saville



**= Email   38 ==========================**

Date:  Mon, 24 Jan 2005 12:20:56 +0000 (GMT)
From:  "Dave Saville" <dave at deezee.org>
Subject:  Re:  Perl help

On Mon, 24 Jan 2005 11:16:03 +0000, John Poltorak wrote:

>It's this bit of Perl code in oschooser.pl:-
>
>	foreach $o ( at list) {
>		if ($o->[4] && eval $o->[4]) {
>			# Got a match!
>			$ver = $o;
>			last;
>			}
>		}
>
>
>
>Can't figure out the condition taht 'if' is checking at all...

It would help if we knew what  at list contained - but it would appear to be a
list of (anonymous) arrays and is testing the 5th element of same. If it
exists, first test, it runs it as code and the return from the eval is whatever
it did last.  

-- 
Regards

Dave Saville