From: UnixOS2 Archive
To: "UnixOS2 Archive" <UnixOS2_Archive at unixos2.com>
Date: Tue, 3 Dec 2002 04:42:37 EST-10EDT,10,-1,0,7200,3,-1,0,7200,3600
Subject: [UnixOS2_Archive] No. 387

**************************************************
Monday 02 December 2002
 Number  387
**************************************************

Subjects for today
 
1  RE:  wxWindows-2.3.4 : Dave Webster <Dave.Webster at bhmi.com>
2   Quoting (was: wxWindows-2.3.4) : illya at vaeshiep.demon.nl
3  RE:  wxWindows-2.3.4 : Yuri Dario" <warp at os2power.com>
4  RE:  wxWindows-2.3.4 : Hakan" <agents at meddatainc.com>
5  RE:  wxWindows-2.3.4 : Hakan" <agents at meddatainc.com>
6  RE:  wxWindows-2.3.4 : Steve Wendt <stevew at shocking.com>
7  RE:  wxWindows-2.3.4 : Hakan" <agents at meddatainc.com>
8  RE:  wxWindows-2.3.4 : Dave Webster <Dave.Webster at bhmi.com>
9  RE:  wxWindows-2.3.4 : Hakan" <agents at meddatainc.com>
10  RE:  wxWindows-2.3.4 : Dave Webster <Dave.Webster at bhmi.com>
11  RE:  wxWindows-2.3.4 : Hakan" <agents at meddatainc.com>
12  RE:  wxWindows-2.3.4 : Yuri Dario" <warp at os2power.com>
13  RE:  wxWindows-2.3.4 : Adrian Gschwend" <ktk at datacomm.ch>
14   configure - problems, Quoting : Franz Bakan" <fbakan at gmx.net>
15  Re:  configure - problems, Quoting : John Poltorak <jp at eyup.org>
16  Re:  wxWindows-2.3.4 : Ken Ames <kenames at pacbell.net>
17  Re:  configure - problems, Quoting : Ken Ames <kenames at pacbell.net>
18  Re:  installpkg : Michael Zolk <o86 at ix.urz.uni-heidelberg.de>
19  RE:  wxWindows-2.3.4 : Yuri Dario" <warp at os2power.com>

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

Date:  Tue, 3 Dec 2002 08:24:46 -0600
From:  Dave Webster <Dave.Webster at bhmi.com>
Subject:  RE:  wxWindows-2.3.4

Hmmm....

Yes, the instructions on the Web Site are a bit confusing at this time.  The
problem is, the "development" branch is not currently the DEVELOPMENT
branch!!!  The administrators of the library branched to WX_2_4_BRANCH about
three months ago and ceased, temporarily, development on the main
development branch *&* at #^*&^ at *(!!!   So really, for now, the REAL
"development" tree is WX_2_4_BRANCH, not the main trunk.  YOu need to check
out wxWindows (or wxOS2) branch or revision tag WX_2_4_BRANCH.

Now if you're not completely confused as of yet, since they have release
2.3.4 (nothing more than a dev snapshot), they are merging WX_2_4_BRANCH
back into the main development tree.  From that point forward we will once
again be developing directly to the main trunk and back porting any
pertinent fixes to the WX_2_4_BRANCH.  WX_2_4_BRANCH, after today, will be
"frozen" meaning no new features or even class interface changes (no header
file changes), only fixes to implementation code.  Eventually, targeted by
year's end, we will release 2.4.0, the next "stable" release of wxWindows
and ALL active development on WX_2_4_BRANCH will cease.  We will then
perform "passive" development, i.e. user reported bug fixes only and
periodically release patch releases 2.4.1, 2.4.2,.....

So after tonight (midnight GMT) if you want the latest and greatest, but
buggy, feature sets, synch up with the main trunk.  If you want more stable
code that will be more representative of what 2.4.0 will be, use
WX_2_4_BRANCH.

Are totally confused now???  I am.....


-----Original Message-----
From: Ken Ames [mailto:kenames at pacbell.net]
Sent: Tuesday, December 03, 2002 1:37 AM
To: os2-unix at eyup.org
Subject: Re:  wxWindows-2.3.4


hi Dave,
  I am currently stumped. I do a fresh "cvs checkout wxOS2" and I get it 
but a LOT of the cvs files are older then what is packaged in the 2.3.4 
release. I think this is why I am not seeing the new files you 
committed. the cvs web page on wxwindows.org says doing the above should 
send the development branch but doesn't seem to send the latest stuff. I 
have even done a full checkout with no help either. I MUST be missing 
something. Maybe you have an explanation (I hope). I look forward to 
hearing back and thanks.

Ken


Dave Webster wrote:

>When did you do the update?  If it was this weekend then the files didn't
>yet make it.  It was late Sunday night (US Central time) that I finally got
>the last files updated.  Compiled just fine here this morning.  Can speak
to
>EMX but VA 3.0 FP8 is currently compilable.  Sorry for the confusion, but
>try another synch, you should get several OS/2 .h and .cpp plus a new
>intl.cpp file from /common at the very least.
>
>-----Original Message-----
>From: Ken Ames [mailto:kenames at pacbell.net]
>Sent: Monday, December 02, 2002 3:35 PM
>To: os2-unix at eyup.org
>Subject: Re:  wxWindows-2.3.4
>
>
>you did not explicitely but did imply or I was imagining. whatever the 
>case but my problem remains. vac 3.0 with fixpak 8 applied will NOT 
>compile wxOS2. I did a cvs update from the cvs server listed on 
>wxWindows home page and got 0 (zero) updated files so, cvs needs repair 
>on their end. either in the readme files or setup postings or whatever. 
>I repeat, I am trying to use VAC 3.0 with fixpak 8 applied and compile 
>does NOT work
>
>Dave Webster wrote:
>
>  
>
>>Where did I say gcc?  icc (.icc) are VA 4.0 build files not gcc (EMX).
>>Stefan has a working EMX (gcc) distrubution, but I personally use VA3.0
FP8
>>and now have a clean compiling version for both static and dynamic
>>    
>>
>libraries
>  
>
>>on the latest CVS tree.  You are free to use gcc/EMX or VA 3.0 FP8 against
>>Os2 toolkit V4.5 or later.  Both work just fine.  Watcom C++ support is
>>unknown since neither Stefan nor I use it and no one else has tried it.
>> 
>>
>>    
>>
>can you point to the latest cvs tree?U 
>mozilla/l10n/langpacks/en-GB/defaults/profile/en-GB/panels.rdf
>U mozilla/l10n/langpacks/en-GB/defaults/profile/en-GB/search.rdf
>
>checkout finish: Mon Dec 2 11:04:39 pst 2002
>Generating configures using autoconf
>SYS1079: << was unexpected at this time.
>configure.in:37: error: m4_file_append: cannot write: /tmp/forbidden.rx
>configure.in:37: the top level
>gmake.exe[1]: *** [real_checkout] Error 2
>gmake.exe[1]: Leaving directory `/'
>gmake: *** [checkout] Error 2
>
>
>  
>
>>-----Original Message-----
>>From: Ken Ames [mailto:kenames at pacbell.net]
>>Sent: Monday, December 02, 2002 2:00 PM
>>To: os2-unix at eyup.org
>>Subject: Re:  wxWindows-2.3.4
>>
>>
>>
>>
>>Dave Webster wrote:
>>
>> 
>>
>>    
>>
>>>There are icc files for an OS/2 build on the CVS tree, still.  You are
>>>welcome to download them and fiddle with them if you desire.  They worked
>>>quite well once upon a time.
>>>
>>>
>>>   
>>>
>>>      
>>>
>>vac 3 with fixpak 8 applied is what I have been trying to use. I guess 
>>you want me to use gcc so I will. :)
>>they seem to work rather inconsistent now. Yuri Dario says my latest 
>>error looks like an ilib bug and I agree.
>>
>>    
>>
>the last sentence above indicated you wanted me to use gcc. If it is not 
>an ilib bug please tell me.
>
>  
>
>> 
>>
>>    
>>
>>>The single biggest bug on VA4 is it's HUGE memory leak.  On a system with
>>>128MB ram the wx build generates and OS/2 swap file over 256MB large.  On
>>>smaller systems, it just crashes on every build.  There are, at time of
>>>      
>>>
>IBM
>  
>
>>>dropping support, over 1000 APARS against this version covering various
>>>issues.  And at $1800 for a non-supported, buggy product, who'd ever buy
>>>      
>>>
>it
>  
>
>>>that didn't otherwise get it for free?
>>>
>>>   
>>>
>>>      
>>>
>I have a copy of VAC4. I do NOT use it, buggy! no good. crashes, etc.... 
>everyone knows vac4 story - dead end.
>
>  
>
>>I have it, bought it from mensys, dont use it, crashes too much and 
>>generates some bad code.
>>
>> 
>>
>>    
>>
>>>It is also an enormous CPU and memory hog.  Don't even try using it on a
>>>system less than PIII650 with less than 256MB ram.  A lot of OS/2 users
>>>   
>>>
>>>      
>>>
>>have
>> 
>>
>>    
>>
>>>much smaller systems than that.  While I have no problem with VA 3.0 FP8
>>>      
>>>
>on
>  
>
>>>a old PII 266 with 64MB ram, VA4 won't even run.  And one of the biggest
>>>selling features left for OS/2 or eCS is it's very small footprint, even
>>>   
>>>
>>>      
>>>
>>for
>> 
>>
>>    
>>
>>>developers.
>>>
>>>VA4 is definitely on the cutting edge of C++ standards and development
>>>environment.  In fact, it probably too far ahead for comfort.  Its
>>>   
>>>
>>>      
>>>
>>technical
>> 
>>
>>    
>>
>>>superiority is not enough to overcome the enormous inertia behind the
>>>traditional header/makefile build paradigm.  And of course, it's most
>>>      
>>>
>fatal
>  
>
>>>flaw, is it is no longer supported by anyone.  At least VA 3.x is a
>>>      
>>>
>mature,
>  
>
>>>debugged, product, even if discontinued.  
>>>
>>>I just wish Watcom could get up to snuff.  As it is, it is not even as
>>>      
>>>
>good
>  
>
>>>as VA 3.0 FP 8 and EMX is just too "Unix-like" for me to bother with.
>>>      
>>>
>What
>  
>
>>>we need is the Watcom compiler complete with the latest OS/2 toolkit WITH
>>>      
>>>
>a
>  
>
>>>fully impelemented stl library.  But the codestore is a concept whose
time
>>>is still YEARS away from being accepted by the masses.
>>>
>>>   
>>>
>>>      
>>>
>>watcom is coming along. c++ compiler is a topic of discussion lately. 
>>seems they are going to start on it soon (I hope).
>>
>><snip>
>>
>>    
>>
>as it stands from here at the moment, wxOS2 is unbuildable with vac3.08 
>so I am hoping you will supply the missing pieces.
>please excuse anything that may seem like I am trying to make you angry 
>as I am most definitly NOT trying to do that. I am simply trying to get 
>a cross platform solution in place here. wxOS2 has been far to long in 
>coming but I realize you are a busy man and still must make a living.
>
>Ken
>
>
>
>
>  
>


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

Date:  Tue, 3 Dec 2002 08:36:44 GMT
From:  illya at vaeshiep.demon.nl
Subject:   Quoting (was: wxWindows-2.3.4)

Ken:
>hi Dave,
>Dave Webster wrote:
>>When did you do the update?  If it was this weekend then the files didn't
>>From: Ken Ames [mailto:kenames at pacbell.net]
>>you did not explicitely but did imply or I was imagining. whatever the 
>>Dave Webster wrote:
>>>Where did I say gcc?  icc (.icc) are VA 4.0 build files not gcc (EMX).
>>>From: Ken Ames [mailto:kenames at pacbell.net]
>>>Dave Webster wrote:
>>>>There are icc files for an OS/2 build on the CVS tree, still.  You are
[Hundreds of lines chopped]

Could you guys stop the top-posting, quote according to netiquette and start reducing the
quoted text to only the relevant pieces / what you're responding to?
See for example http://www.xs4all.nl/~wijnands/nnq/nquote.html .
Thank you.



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

Date:  Tue, 03 Dec 2002 10:10:55 +0100 (CET)
From:  "Yuri Dario" <warp at os2power.com>
Subject:  RE:  wxWindows-2.3.4

Hi Hakan,

>relaunching it takes care of that problem) I have yet to come across
>any code generation problems.  I know that the version 4 of the
>compiler also comes with a command-line version of the compiler and
>other tools which I *assume* is what is being used by the IDE itself

I'm using VA4 for a big C++ project, and I never got code optimizer to work. Every try to add 
opt(level,2) ended with a crashing binary. Further look at optimizer options, showed that the 
code scheduler is the failing optimizer option. Unfortunately it can't be turned off once 
opt(level,2) is used. Actually I'm living with only size optimization, I'm (quite) happy with it.

Regarding the command line tools, the one installed with VA4 is only the C compiler, that is run 
as an external tool by the IDE (while C++ code is handled internally).


Bye,

	Yuri Dario

/*
 * member of TeamOS/2 - Italy
 * http://www.quasarbbs.net/yuri
 * http://www.teamos2.it
 * http://www.opera.com/os2/
 */

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

Date:  Tue, 03 Dec 2002 12:38:35 -0500 (EST)
From:  "Hakan" <agents at meddatainc.com>
Subject:  RE:  wxWindows-2.3.4

Is 3.6.5 the back-end for VAC++ 4?  IOW, are they at the same level
with respect to conformance with standard C++ etc?

On Mon, 2 Dec 2002 19:18:21 -0800 (PST), Steve Wendt wrote:

>On Mon, 2 Dec 2002, Hakan wrote:
>
>> Well, VA 3.x is no longer supported either and VAC++ 4. offers some
>> level of conformance with standard C++.
>
>VA 3.6.5 is still getting fixes, thanks to IBM's Mozilla team.
>
>
>
>



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

Date:  Tue, 03 Dec 2002 12:50:27 -0500 (EST)
From:  "Hakan" <agents at meddatainc.com>
Subject:  RE:  wxWindows-2.3.4

Hi Yuri,

I have not used the optimizer at all yet since I am in the early stages
of my project.  Re the command-line compiler, are you saying that it
cannot be used to compile C++ programs, only C programs?

Re APARs against VAC++, I visited
http://www-1.ibm.com/support/apar_search.html.  A search for VAC++
resulted in 144 hits while searching for vacpp resulted in 40 hits.  I
would think that not only is there overlap between the two, the APARs
also cover versions 3, 4 and 5 of the compiler for all platforms.  I
have a hard time reconciling this with Dave Webster's statement that
there are over 1000 APARs against VAC++ 4.

Hakan

On Tue, 03 Dec 2002 10:10:55 +0100 (CET), Yuri Dario wrote:

>Hi Hakan,
>
>>relaunching it takes care of that problem) I have yet to come across
>>any code generation problems.  I know that the version 4 of the
>>compiler also comes with a command-line version of the compiler and
>>other tools which I *assume* is what is being used by the IDE itself
>
>I'm using VA4 for a big C++ project, and I never got code optimizer to work. Every try to add 
>opt(level,2) ended with a crashing binary. Further look at optimizer options, showed that the 
>code scheduler is the failing optimizer option. Unfortunately it can't be turned off once 
>opt(level,2) is used. Actually I'm living with only size optimization, I'm (quite) happy with it.
>
>Regarding the command line tools, the one installed with VA4 is only the C compiler, that is run 
>as an external tool by the IDE (while C++ code is handled internally).
>
>
>Bye,
>
>	Yuri Dario
>
>/*
> * member of TeamOS/2 - Italy
> * http://www.quasarbbs.net/yuri
> * http://www.teamos2.it
> * http://www.opera.com/os2/
> */
>
>





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

Date:  Tue, 3 Dec 2002 14:12:59 -0800 (PST)
From:  Steve Wendt <stevew at shocking.com>
Subject:  RE:  wxWindows-2.3.4

On Tue, 3 Dec 2002, Hakan wrote:

> Is 3.6.5 the back-end for VAC++ 4?  IOW, are they at the same level
> with respect to conformance with standard C++ etc?

No, I'm pretty sure 3.6.5 is a "normal" compiler like 3.0.8.  It's just
version 4 that does weird things.  I don't know how 3.6.5 compares to 3.0
and 4.0 with regards to C++ conformance, although I know the Mozilla team
has had a couple problems with people using newer C++ code that didn't
work (but it also failed on several other compilers; hence Mozilla has a
strict rule about NOT using recent C++ standards).  Personally, I expect
C++ to stop evolving, because there aren't too many vendors interested in
making compilers for it anymore.  C will live on, but Java will probably
replace C++ for a lot of things.


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

Date:  Tue, 03 Dec 2002 15:28:45 -0500 (EST)
From:  "Hakan" <agents at meddatainc.com>
Subject:  RE:  wxWindows-2.3.4

Yuri,

Ouch.  So, for C++ your  choices are VAC++ version 3.08 or version 4
using the IDE.  As for conformance with standard C++, I understand
Watcom C++ is far behind and I don't know how gcc fares.

I wonder what the possibility is that IBM would open-source their
compiler??

Hakan

On Tue, 03 Dec 2002 19:52:41 +0100 (CET), Yuri Dario wrote:

>Hi,
>
>>Re the command-line compiler, are you saying that it
>>cannot be used to compile C++ programs, only C programs?
>
>exactly: if you try to run iccv4.exe against a cpp file, it will tell you that module "xxx" cannot be 
>loaded.
>
>
>Bye,
>
>	Yuri Dario
>
>/*
> * member of TeamOS/2 - Italy
> * http://www.quasarbbs.net/yuri
> * http://www.teamos2.it
> * http://www.opera.com/os2/
> */
>
>



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

Date:  Tue, 3 Dec 2002 15:44:38 -0600
From:  Dave Webster <Dave.Webster at bhmi.com>
Subject:  RE:  wxWindows-2.3.4

I wonder, honestly, how many people are actually using VAC 3.0 that they
actually paid for?  This product is at least eight years old now and most
the original liscenses are unused now, probably given away to anyone who
cared.  My old employer abandoned OS/2 back in 1995 for Windows NT 3.5 and
just gave me the compiler because they no longer had any use for it.  

You can download the whole thing, at will from public ftp sites, mostly in
Russia and China and such.  In fact, many IBM employees world wide just give
the stuff to you if ask! A friend of mine here walked over to the local IBM
office and just asked if they had any spare, unused copies of VA 3.6 for
OS/2 lying around and they gave him TEN of them!  I don't know why they just
don't open it up, their own people are giving it away for free if just know
who to ask!

-----Original Message-----
From: Adrian Gschwend [mailto:ktk at datacomm.ch]
Sent: Tuesday, December 03, 2002 3:11 PM
To: os2-unix at eyup.org
Subject: RE:  wxWindows-2.3.4


On Tue, 03 Dec 2002 15:28:45 -0500 (EST), Hakan wrote:

>Ouch.  So, for C++ your  choices are VAC++ version 3.08 or version 4
>using the IDE.  As for conformance with standard C++, I understand
>Watcom C++ is far behind and I don't know how gcc fares.
>
>I wonder what the possibility is that IBM would open-source their
>compiler??

I tried to get in contact with some people about that. Looks like
absolutely no chance. It's even not possible to get free licenses of
VAC3.08 for netlabs.org.

To quote Mike Kaply (Mozilla for OS/2 maintainer): 

---snip---
I have had discussions with the VACPP about just getting a few copies
of 
the compiler to give to people and it was like pulling teeth.

IBM has NO interest in giving the compiler to anyone now that it is out
of 
support, and I don't think you will get anywhere trying to distribute
it.

I can give you names if you want, but I'm being totally honest with you

when I tell you that it will be a waste of time. The Toronto team
doesn't 
want anyone to have the compiler anymore.

If they had their way, the compiler would have self destructed when it 
went out of support :)
---snip---

That's not really about opensource it but I think it's hopeless.

IMHO we should go for GCC for ports and OpenWatcom for "native" stuff
and abandon the rest.

cu

Adrian


-- 
Adrian Gschwend
 at  netlabs.org

ktk [a t] netlabs.org
-------
Free Software for OS/2 and eCS
http://www.netlabs.org


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

Date:  Tue, 03 Dec 2002 16:42:52 -0500 (EST)
From:  "Hakan" <agents at meddatainc.com>
Subject:  RE:  wxWindows-2.3.4

Adrian,

That's very disappointing to hear, of course and I do not understand
the reason for it.  BTW, is not the current version of the AIX compiler
version 6?  IOW, IBM has continued to improve their compiler technology
(substantially?) and it might not be that hard to port the AIX version
to OS/2?

I wonder if Serenity might have more clout in getting access to the
compiler technology (and releasing an updated/bug fixed version for
OS/2)?  It would certainly be very much in their interest to do so. 
Have you been in contact with Serenity?

Hakan

On Tue, 03 Dec 2002 22:11:03 +0100 (CET), Adrian Gschwend wrote:

>On Tue, 03 Dec 2002 15:28:45 -0500 (EST), Hakan wrote:
>
>>Ouch.  So, for C++ your  choices are VAC++ version 3.08 or version 4
>>using the IDE.  As for conformance with standard C++, I understand
>>Watcom C++ is far behind and I don't know how gcc fares.
>>
>>I wonder what the possibility is that IBM would open-source their
>>compiler??
>
>I tried to get in contact with some people about that. Looks like
>absolutely no chance. It's even not possible to get free licenses of
>VAC3.08 for netlabs.org.
>
>To quote Mike Kaply (Mozilla for OS/2 maintainer): 
>
>---snip---
>I have had discussions with the VACPP about just getting a few copies
>of 
>the compiler to give to people and it was like pulling teeth.
>
>IBM has NO interest in giving the compiler to anyone now that it is out
>of 
>support, and I don't think you will get anywhere trying to distribute
>it.
>
>I can give you names if you want, but I'm being totally honest with you
>
>when I tell you that it will be a waste of time. The Toronto team
>doesn't 
>want anyone to have the compiler anymore.
>
>If they had their way, the compiler would have self destructed when it 
>went out of support :)
>---snip---
>
>That's not really about opensource it but I think it's hopeless.
>
>IMHO we should go for GCC for ports and OpenWatcom for "native" stuff
>and abandon the rest.
>
>cu
>
>Adrian
>
>
>-- 
>Adrian Gschwend
> at  netlabs.org
>
>ktk [a t] netlabs.org
>-------
>Free Software for OS/2 and eCS
>http://www.netlabs.org
>
>
>



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

Date:  Tue, 3 Dec 2002 16:57:22 -0600
From:  Dave Webster <Dave.Webster at bhmi.com>
Subject:  RE:  wxWindows-2.3.4

Some of the newer things in the C++ standard simply will not work with VA
3.0 or 3.6 for that matter.  Templates are still a big problem and forget
the entire concept of the codestore.  That is the really radical standard
that I doubt ANYONE will EVER bother to implement in a compiler.  I know MS
has no intent of EVER supporting that standard.  Anyone foolish enough to
opt for projects using codestores better not ever expect to produce a
development library they expect anyone to actually use.

However C++ is not going away and MS and the GNU folks are intent on
supporting C++ for many many years to come as it is now the defacto standard
for development of business software and like COBOL before it, ain't going
away for DECADES to come.

-----Original Message-----
From: Steve Wendt [mailto:stevew at shocking.com]
Sent: Tuesday, December 03, 2002 4:13 PM
To: os2-unix at eyup.org
Subject: RE:  wxWindows-2.3.4


On Tue, 3 Dec 2002, Hakan wrote:

> Is 3.6.5 the back-end for VAC++ 4?  IOW, are they at the same level
> with respect to conformance with standard C++ etc?

No, I'm pretty sure 3.6.5 is a "normal" compiler like 3.0.8.  It's just
version 4 that does weird things.  I don't know how 3.6.5 compares to 3.0
and 4.0 with regards to C++ conformance, although I know the Mozilla team
has had a couple problems with people using newer C++ code that didn't
work (but it also failed on several other compilers; hence Mozilla has a
strict rule about NOT using recent C++ standards).  Personally, I expect
C++ to stop evolving, because there aren't too many vendors interested in
making compilers for it anymore.  C will live on, but Java will probably
replace C++ for a lot of things.


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

Date:  Tue, 03 Dec 2002 18:39:16 -0500 (EST)
From:  "Hakan" <agents at meddatainc.com>
Subject:  RE:  wxWindows-2.3.4

Dave,

I don't know where you got the idea that the codestore is a "standard",
it certainly is not part of standard C++ but was/is of course IBM's
attempt at making the programmer more productive.  Further, I have
already told you in a previous message that is possible to retain your
traditional source file structure with #include statements yet still be
compatible with/not be penalized by the codestore concept.  Templates
of course, are not only extremely useful but required by the STL.

At the risk of repeating myself -- can you please give us a reference
supporting the statement  you made re "1000's of APARs" against VAC++
4?

Hakan

On Tue, 3 Dec 2002 16:57:22 -0600, Dave Webster wrote:

>Some of the newer things in the C++ standard simply will not work with VA
>3.0 or 3.6 for that matter.  Templates are still a big problem and forget
>the entire concept of the codestore.  That is the really radical standard
>that I doubt ANYONE will EVER bother to implement in a compiler.  I know MS
>has no intent of EVER supporting that standard.  Anyone foolish enough to
>opt for projects using codestores better not ever expect to produce a
>development library they expect anyone to actually use.
>
>However C++ is not going away and MS and the GNU folks are intent on
>supporting C++ for many many years to come as it is now the defacto standard
>for development of business software and like COBOL before it, ain't going
>away for DECADES to come.
>
>-----Original Message-----
>From: Steve Wendt [mailto:stevew at shocking.com]
>Sent: Tuesday, December 03, 2002 4:13 PM
>To: os2-unix at eyup.org
>Subject: RE:  wxWindows-2.3.4
>
>
>On Tue, 3 Dec 2002, Hakan wrote:
>
>> Is 3.6.5 the back-end for VAC++ 4?  IOW, are they at the same level
>> with respect to conformance with standard C++ etc?
>
>No, I'm pretty sure 3.6.5 is a "normal" compiler like 3.0.8.  It's just
>version 4 that does weird things.  I don't know how 3.6.5 compares to 3.0
>and 4.0 with regards to C++ conformance, although I know the Mozilla team
>has had a couple problems with people using newer C++ code that didn't
>work (but it also failed on several other compilers; hence Mozilla has a
>strict rule about NOT using recent C++ standards).  Personally, I expect
>C++ to stop evolving, because there aren't too many vendors interested in
>making compilers for it anymore.  C will live on, but Java will probably
>replace C++ for a lot of things.
>
>
>



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

Date:  Tue, 03 Dec 2002 19:52:41 +0100 (CET)
From:  "Yuri Dario" <warp at os2power.com>
Subject:  RE:  wxWindows-2.3.4

Hi,

>Re the command-line compiler, are you saying that it
>cannot be used to compile C++ programs, only C programs?

exactly: if you try to run iccv4.exe against a cpp file, it will tell you that module "xxx" cannot be 
loaded.


Bye,

	Yuri Dario

/*
 * member of TeamOS/2 - Italy
 * http://www.quasarbbs.net/yuri
 * http://www.teamos2.it
 * http://www.opera.com/os2/
 */

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

Date:  Tue, 03 Dec 2002 22:11:03 +0100 (CET)
From:  "Adrian Gschwend" <ktk at datacomm.ch>
Subject:  RE:  wxWindows-2.3.4

On Tue, 03 Dec 2002 15:28:45 -0500 (EST), Hakan wrote:

>Ouch.  So, for C++ your  choices are VAC++ version 3.08 or version 4
>using the IDE.  As for conformance with standard C++, I understand
>Watcom C++ is far behind and I don't know how gcc fares.
>
>I wonder what the possibility is that IBM would open-source their
>compiler??

I tried to get in contact with some people about that. Looks like
absolutely no chance. It's even not possible to get free licenses of
VAC3.08 for netlabs.org.

To quote Mike Kaply (Mozilla for OS/2 maintainer): 

---snip---
I have had discussions with the VACPP about just getting a few copies
of 
the compiler to give to people and it was like pulling teeth.

IBM has NO interest in giving the compiler to anyone now that it is out
of 
support, and I don't think you will get anywhere trying to distribute
it.

I can give you names if you want, but I'm being totally honest with you

when I tell you that it will be a waste of time. The Toronto team
doesn't 
want anyone to have the compiler anymore.

If they had their way, the compiler would have self destructed when it 
went out of support :)
---snip---

That's not really about opensource it but I think it's hopeless.

IMHO we should go for GCC for ports and OpenWatcom for "native" stuff
and abandon the rest.

cu

Adrian


-- 
Adrian Gschwend
 at  netlabs.org

ktk [a t] netlabs.org
-------
Free Software for OS/2 and eCS
http://www.netlabs.org


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

Date:  Tue, 03 Dec 2002 23:04:09 +0100 (CET)
From:  "Franz Bakan" <fbakan at gmx.net>
Subject:   configure - problems, Quoting

Hi,

since the SANE-Team updated to autoconf to 2.54 compiling
the SANE-CVS-Code for OS/2 now doesn't need
the os2unix.cmd helper application any more :-)
nor running autoconf here locally.

I can set the needed things in config.site.

sh configure

now works 'out of the box'

but there remains a small 'problem'

After sh configure has finished I now have 264 sh-thd-??????? files.
Not a real problem, but does someone know how to turn
this 'feature' off?

( I'm using sh.exe from pdksh-5.2.13-emx.zip 07.01.99 )

Another question:

Is there any switch to turn on 'debugging-output' for
sh configure?

And last but not least, I also dislike all those useless
full-quote-lines below postings.
Please only quote the relevant lines.

Ken, Hakan, Dave are you listening?

I'm connected with a modem and would like to keep
traffic low.

Regards,
Franz



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

Date:  Tue, 3 Dec 2002 23:10:45 +0000
From:  John Poltorak <jp at eyup.org>
Subject:  Re:  configure - problems, Quoting

On Tue, Dec 03, 2002 at 11:04:09PM +0100, Franz Bakan wrote:
> Hi,
> 
> since the SANE-Team updated to autoconf to 2.54 compiling
> the SANE-CVS-Code for OS/2 now doesn't need
> the os2unix.cmd helper application any more :-)
> nor running autoconf here locally.
> 
> I can set the needed things in config.site.
> 
> sh configure
> 
> now works 'out of the box'
> 
> but there remains a small 'problem'
> 
> After sh configure has finished I now have 264 sh-thd-??????? files.
> Not a real problem, but does someone know how to turn
> this 'feature' off?
> 
> ( I'm using sh.exe from pdksh-5.2.13-emx.zip 07.01.99 )


Personally, I would suggest this version as a standard:-

http://www.math.ohio-state.edu/~ilya/software/os2/pdksh-5.2.14-bin-2.zip


A lot of problems disappeared when I started using this one.


> Regards,
> Franz

-- 
John
 


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

Date:  Tue, 03 Dec 2002 23:30:44 -0800
From:  Ken Ames <kenames at pacbell.net>
Subject:  Re:  wxWindows-2.3.4



Dave Webster wrote:

>Hmmm....
>
>Yes, the instructions on the Web Site are a bit confusing at this time.  The
>problem is, the "development" branch is not currently the DEVELOPMENT
>branch!!!  The administrators of the library branched to WX_2_4_BRANCH about
>three months ago and ceased, temporarily, development on the main
>development branch *&* at #^*&^ at *(!!!   So really, for now, the REAL
>"development" tree is WX_2_4_BRANCH, not the main trunk.  YOu need to check
>out wxWindows (or wxOS2) branch or revision tag WX_2_4_BRANCH.
>
>Now if you're not completely confused as of yet, since they have release
>2.3.4 (nothing more than a dev snapshot), they are merging WX_2_4_BRANCH
>back into the main development tree.  From that point forward we will once
>again be developing directly to the main trunk and back porting any
>pertinent fixes to the WX_2_4_BRANCH.  WX_2_4_BRANCH, after today, will be
>"frozen" meaning no new features or even class interface changes (no header
>file changes), only fixes to implementation code.  Eventually, targeted by
>year's end, we will release 2.4.0, the next "stable" release of wxWindows
>and ALL active development on WX_2_4_BRANCH will cease.  We will then
>perform "passive" development, i.e. user reported bug fixes only and
>periodically release patch releases 2.4.1, 2.4.2,.....
>
>So after tonight (midnight GMT) if you want the latest and greatest, but
>buggy, feature sets, synch up with the main trunk.  If you want more stable
>code that will be more representative of what 2.4.0 will be, use
>WX_2_4_BRANCH.
>
>Are totally confused now???  I am.....
>
>
>-----Original Message-----
>From: Ken Ames [mailto:kenames at pacbell.net]
>Sent: Tuesday, December 03, 2002 1:37 AM
>To: os2-unix at eyup.org
>Subject: Re:  wxWindows-2.3.4
>  
>
><snip>
>
hi Dave,
  sure I am. actually it all makes sense. compile went much further this 
time. got all the way to png where it bombed with this error

..\png\pngwrite.c(400:44) : error EDC3292: The "->" operator cannot be 
applied t
o undefined class "tm".
..\png\pngwrite.c(414:11) : error EDC3013: Identifier "gmtime" is undefined.
NMAKE : fatal error U1077: 'G:\OS2\CMD.EXE' : return code '12'
Stop.

I hope this is not an error on my end, it may be. I will look at it a bit.

so the real  WX_2_4_BRANCH is gonna move to where it belongs soon then eh.
thank you for you help and insight in this trouble of mine.

Ken



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

Date:  Tue, 03 Dec 2002 23:36:45 -0800
From:  Ken Ames <kenames at pacbell.net>
Subject:  Re:  configure - problems, Quoting


hi Franz,
  I am sorry but I do not know the answers to your questions.

  This is to mere say that the full text quote helps developers keep 
some important notes in mind from one post to the next, otherwise some 
things get forgotten and confusion ultimately reigns. I remember my 
modem days and would much rather see extra quoted text then a temper 
eruptionfrom anyone.  That happens easily but I will keep this in mind 
and trim as much as I can. Thanks for bearing with me.

Ken

>
>And last but not least, I also dislike all those useless
>full-quote-lines below postings.
>Please only quote the relevant lines.
>
>Ken, Hakan, Dave are you listening?
>
>I'm connected with a modem and would like to keep
>traffic low.
>
>Regards,
>Franz
>
>
>
>
>  
>


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

Date:  Tue, 3 Dec 2002 23:40:49 +0100
From:  Michael Zolk <o86 at ix.urz.uni-heidelberg.de>
Subject:  Re:  installpkg

On Fri, Nov 22, 2002 at 12:08:25AM +0100, Andreas Buening wrote:
 
> > Yes. According to the FHS, such "internal binaries" should be placed in
> > /usr/lib/<pkg> (or /lib/<pkg> in this case), but this would require another
> > addition to PATH. So the right place would indeed be /bin or /sbin.
> 
> No, only /bin is in PATH. All essential binaries and scripts that
> are of "general interest" for the user like installpkg.cmd itself
> belong to /bin and /bin is in PATH. All essential binaries and scripts
> that are NOT of "general interest" for the user (but maybe for the
> admin) belong to /sbin (Currently, /sbin is empty) and /sbin is NOT
> in PATH. If you have some special installation scripts that are
> necessary to activate a certain package put them into /var/lib. 
> If the same script is required by two or three packages put it
> into each of them (-> /var/lib/unixos2/scripts/package1-foo.cmd,
> /var/lib/unixos2/scripts/package2-foo.cmd). But if it's required
> by several packages it's better to put that script into /sbin.

These "helper scripts" are not intended to be executed by the user, but 
are run by installpkg. The FHS says: "/usr/lib includes object files, 
libraries, and internal binaries that are not intended to be executed 
directly by users or shell scripts.", so /usr/lib/unixos2 would be the
right place if installpkg were in /usr/bin. I don't know if /lib/unixos2
is the correct location of scripts like mod.cmd, but I think I should put
them into /bin anyway to avoid having to add lots of directories to PATH.

As far as I understand it, the /var hierarchy is only for variable data that
can change during the execution of a program.
 
> > [ install scripts that require additional packages ]
 
> > I know that the Debian policy requires that all programs that are marked
> > 'Essential' must work properly even without configuration.
> 
> We can require the same. However, I think at least postinstall
> scripts won't be a problem.

Only if the execution of the postinstall scripts is deferred until all
of the essential packages are installed.

> How safe is doinst.cmd? E.g. if anbody installs the UnixOS/2 stuff
> several times will the directories be added several times to config.sys?
> Can the backup config.sys accidentally be deleted? 

doinst.cmd probably needs more error checking.

> Is it guaranteed
> that those paths are available when the first install scripts
> are executed? 

Yes, the ux2_base package should include all the "standard" directories 
from the FHS.

> > > SCRIPT {PREINST|POSTINST|POSTDEL} scriptname
> > > At least the POSTINST scripts can be sh scripts. Where are these
> > > scripts stored? How do you avoid name collisions?
> > 
> > The scripts are copied to /var/lib/unixos2/scripts/. The scripts there
> > are named <pkgname>.postinst<number>. The <number> at the end is used so
> > that it's possible to have more than one script of the same type.
> 
> What happens if a package has two scripts (foo1, foo2) and foo1
> runs foo2? 

This won't work at the moment since only the script that is specified in the
PKGINFO file is extracted from the zip file.

> Would it be possible to use <scriptname>.<pkgname>.postinstall
> instead so that the script gets a fixed name? (and an error if there
> are two scripts with the same name)


> > Of course it would be good to have a 'mkpkginfo' command that adds all the
> > files in a directory tree to a PKGINFO file.
> 
> The problem is it won't be possible to create PKGINFO file for
> large packages without a "mkpkginfo" if we don't have "FILE ALL".
> It might be possible to read the zip file content by "unzip -Z -1"
> (e.g. into a queue). This command prints all file names to stdout.
 
IIRC Rexx can search recursively for files in a directory tree. This is all
that's needed to write such a script :) It could of course also be a shell or
perl script.
 
> As long as we don't have those features it has no effect but if
> this ever happens we have a problem: Assume a file is installed
> for a user "joe". There might be no such user on the system
> (unless "joe" is "root"), the current user (i.e. you) may have
> no access to "joe"'s files, i.e. you might be able to install these
> files but the postinstall scripts may fail because they can't even
> modify a config file.

Of course package maintainers have to choose sensible defaults for
file owners, which probably means "root". 

> Btw. do we want to have another file extension for the package
> files instead of ".zip"? This would allow to use wps associations
> for that file extension.

I don't know... Which extension, maybe "ux2"?

> > > CONFFILE pathname [owner group mode]
 
> What happens if the whole distro with dozens of config files is
> installed at the same time? Is there an option to overwrite all
> config files (-f option?) or to rename the old ones to *.backup?
> Pressing "y" or "n" several hundred times can become very annoying. ;-)

No, not at the moment. The most elegant solution would probably be to check
if the configuration file has been modified by the user and only prompt
the user if this is the case. However, I don't know what's the best way
to implement this.

Michael
-- 
Who's
Taking
Over

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

Date:  Tue, 03 Dec 2002 23:56:11 +0100 (CET)
From:  "Yuri Dario" <warp at os2power.com>
Subject:  RE:  wxWindows-2.3.4

Hi,

>the reason for it.  BTW, is not the current version of the AIX compiler
>version 6?  IOW, IBM has continued to improve their compiler technology
>(substantially?) and it might not be that hard to port the AIX version

(I think) they dropped the IDE in version 5.0, so actually it is only a traditional command line 
compiler.


Bye,

	Yuri Dario

/*
 * member of TeamOS/2 - Italy
 * http://www.quasarbbs.net/yuri
 * http://www.teamos2.it
 * http://www.opera.com/os2/
 */