opencxx-users

[Opencxx-users] Re: [opencxx] Re: ANSI C++/Gcc 3.2 compatibility patches

[Alexandre T. <kt...@fr...>]

Le lundi, 10 mar 2003, =E0 19:20 Europe/Paris, Michael Hohmuth a =E9crit =
:

> Grzegorz Jakacki <ja...@he...> writes:
>
>> You have been added to the project. Currently Alexandre Tolmos is
>> finishing his commits wrt. gcc-3.x, there will be an annoucement once =
=20
>> he
>> is done. Before this happens you may want to contact me if you think =20
>> you
>> need any help on branching/merging.
>
> Thanks.
>
> I don't think I need any help with merging, but I would like to have
> some feedback.
>
> I observed that in the current CVS version (which supposedly already
> contains some or all of the previous GCC3 work), there appear some
> "using namespace std" declarations in header files.  This, to my eyes,
> is bad style.  I am working on a project that uses OpenC++ as a
> library, and there it potentially ``hurts'' if a header file imports
> the whole std namespace into the global namespace.

Sorry, I'm responsible for that global namespace pollution :)

I recognize it's brute force but it was the simplest (and dirtiest) way =20
of promoting symbols from the standard library. In the case of of stand =20
alone application it does not hurt but if you intend to use Oc++ as =20
library then it's not the correct way.

> So, once Alexandre is finished, I would like to undo these changes and
> instead revert to the style used in my patch:  The "using namespace
> std" declarations are only allowed to appear in .cc files, and .h
> files reference std symbols using qualified names.

Actually I should have put the "using namespace std" declarations =20
within the "Opencxx" namespace itself; then none of the symbols from =20
the standard library would have been promoted in the global namespace...

> There also is the issue of naming header files coming from standard
> C.  In my original patch, they are included as, for example, <cstring>
> instead of <string.h>, whereas the current CVS version uses the
> latter, older style.  Here, I am planning to leave it at the CVS
> version's style, because I think that it might be more portable.

I think we should move to the new standard headers but I don't know if =20
it would still compile on Win32.

Btw, Grzegorz has got problems compiling with Gcc 3.2.1 (errors in =20
"buffer.cc" pertaining to the "placement new" operator). On my Mac I =20
only have Gcc 3.1. Did you encounter such problems with Gcc 3.2?

Alexandre.

------------------------------------------------------------------------=20
-
Alexandre Tolmos
E-mail:=A0...@fr...
ICQ: 92964905
------------------------------------------------------------------------=20
-
"Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn."
------------------------------------------------------------------------=20
-

[Opencxx-users] Re: [opencxx] Re: ANSI C++/Gcc 3.2 compatibility patches

[Michael H. <ho...@sa...>]

Grzegorz Jakacki <ja...@he...> writes:

> You have been added to the project. Currently Alexandre Tolmos is
> finishing his commits wrt. gcc-3.x, there will be an annoucement once he
> is done. Before this happens you may want to contact me if you think you
> need any help on branching/merging.

Thanks.

I don't think I need any help with merging, but I would like to have
some feedback.

I observed that in the current CVS version (which supposedly already
contains some or all of the previous GCC3 work), there appear some
"using namespace std" declarations in header files.  This, to my eyes,
is bad style.  I am working on a project that uses OpenC++ as a
library, and there it potentially ``hurts'' if a header file imports
the whole std namespace into the global namespace.

So, once Alexandre is finished, I would like to undo these changes and
instead revert to the style used in my patch:  The "using namespace
std" declarations are only allowed to appear in .cc files, and .h
files reference std symbols using qualified names.

There also is the issue of naming header files coming from standard
C.  In my original patch, they are included as, for example, <cstring>
instead of <string.h>, whereas the current CVS version uses the
latter, older style.  Here, I am planning to leave it at the CVS
version's style, because I think that it might be more portable.

Do you (Grzegorz, Alexande, or anyone else) have any objections?

Michael
-- 
ho...@sa..., ho...@in...
http://www.sax.de/~hohmuth/

[Opencxx-users] [opencxx] ETAPS 2003

[Grzegorz J. <ja...@he...>]

Hi,

I am trying to find out if anybody on this list is comming to ETAPS 2003
conference.

Regards
Grzegorz

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] Re: [opencxx] Let's disable which that we do not use

[Grzegorz J. <ja...@he...>]

On Fri, 7 Mar 2003, Detlef Vollmann wrote:

> Grzegorz Jakacki wrote:
> > Below you find summary of open SF features. I suggest to shut down
> > everything except CVS (of course), News, File releases and Mailing Lists
> > (not to be confused with Forums). I Hope I did not miss anything.
> Sounds reasonable.
> If possible, it would make sense to add a link to this list
> in the Mailing Lists section.

I am not sure if we can do it this way, since the project cannot add
content to the Mailing List section (otherwise than by setting up new
mailing lists) but I think we can:

(1) Put the info on opencxx.sf.net
(2) Add footer with subscribtion instructions to distributed messages.

This should eventually become redundant if we manage to move the entire
list management to SF.

Best regards
Grzegorz

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] Re: [opencxx] Re: ANSI C++/Gcc 3.2 compatibility patches

[Grzegorz J. <ja...@he...>]

On Fri, 7 Mar 2003, Michael Hohmuth wrote:

> Grzegorz Jakacki <ja...@he...> writes:
>
> > Thank you very much for your contribution. The gcc-3.x compatibility issue
> > surfaces regularily. Recently Alexandre Tolmos has merged the old patch
> > prepared by Michael Buro addressing exactly this same problem, and is
> > currently in the phase of testing it on Compile Farm.
> >
> > This unfortunately means, that some work has been done twice.  [...]
> > Next time, I would suggest that you first contact the mailing list or me
> > directly to discuss the contribution you would like to make. The project
> > resources are scarce and I would hate to waste time of any contributor.
> >
> > I am not sure if you are subscribing to the OpenC++ mailing list, since
> > gcc-3.x patch was discussed here.
> > [...] archived at [sourceforge.net] since Sep 2002.
>
> I am, in fact, subscribed to the list (and I even have a complete
> mailing-list archive dating back to May 1997),

Could you make a tarball out of it and send it to me? I think we should
put it on the website. I have been trying to find the complete archive for
several times, but I failed.

> but I have not followed
> it closely during the few past months.  Now it turns out I should have
> (but then again, it wasn't that much work).
>
> > Nevertheless, it would be reasonable to try to merge your changes after
> > Alexandre is done, so that we are sure that no detail has been lost. If
> > you feel comfortable with CVS merge facilities I would encourage you to
> > merge yourself (send me your SF userid for CVS access). This however may
> > be tedious, so if you do not volunteer to do this part, I can do it.
>
> I'd be happy to do the merge.

Great.

> My SF ID is "rotofoil".

You have been added to the project. Currently Alexandre Tolmos is
finishing his commits wrt. gcc-3.x, there will be an annoucement once he
is done. Before this happens you may want to contact me if you think you
need any help on branching/merging.

[...]
> I do not have any experience with SF's compile farm and could use some
> help in testing.  Unfortunately, I do not have the time to manually
> configure and test OpenC++ on each and every architecture and
> compiler.  Is there an automated way to try a compilation on all
> architectures?

This is what I would love to have. I am working on a script, that would
just configure, build, compile and test on a given platform, possibly
sending the results in e-mail. There is a problem with making a script for
all the platforms, because of strange login procedure at CompileFarm, but
this eventually can be solved too.

The testing on CF has been simplified thanks to adoption of sandbox
branches methodology. You keep all your changes in your branch, (e.g.
sandbox_rotofoil), which you merge into main trunk after you test them.
This simplifies testing on remote hosts (on CF), because in fact the only
thing you need to supply is the name of sandbox to test. This is how
it looks today:

    (1) log on and select host
    (2) checkout files from 'sandbox_rotofoil'
    (3) cd opencxx; ./configure; make test

This is how I hope it is going to look in one-month timeframe:

    (1) log on and select host
    (2) ./build-and-test --tag=sandbox_rotofoil

This is what I would love to have

    (1) ./multihost-build-and-test --tag=sandbox_rotofoil

> Which architectures need to be considered as ``supported by OpenC++''?

Currently we have resources to support i86/Linux, Sun/Linux and
Sun/Solaris, hopefully we will be supporting MacOS X is the near future.
(I think the head revision still compiles on i86/Cygwin and HPUX, but I
have no way of verifying it.)

Feel free to contact me if you have any questions.

Best regards
Grzegorz

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] Re: [opencxx] Let's disable which that we do not use

[Detlef V. <dv...@vo...>]

Grzegorz Jakacki wrote:
> Below you find summary of open SF features. I suggest to shut down
> everything except CVS (of course), News, File releases and Mailing Lists
> (not to be confused with Forums). I Hope I did not miss anything.
Sounds reasonable.
If possible, it would make sense to add a link to this list
in the Mailing Lists section.

Detlef

-- 
Detlef Vollmann   vollmann engineering gmbh
Linux and C++ for Embedded Systems    http://www.vollmann.ch/

[Opencxx-users] Re: [opencxx] Re: ANSI C++/Gcc 3.2 compatibility patches

[Michael H. <ho...@sa...>]

Grzegorz Jakacki <ja...@he...> writes:

> Thank you very much for your contribution. The gcc-3.x compatibility issue
> surfaces regularily. Recently Alexandre Tolmos has merged the old patch
> prepared by Michael Buro addressing exactly this same problem, and is
> currently in the phase of testing it on Compile Farm.
>
> This unfortunately means, that some work has been done twice.  [...]
> Next time, I would suggest that you first contact the mailing list or me
> directly to discuss the contribution you would like to make. The project
> resources are scarce and I would hate to waste time of any contributor.
>
> I am not sure if you are subscribing to the OpenC++ mailing list, since
> gcc-3.x patch was discussed here. 
> [...] archived at [sourceforge.net] since Sep 2002.

I am, in fact, subscribed to the list (and I even have a complete
mailing-list archive dating back to May 1997), but I have not followed
it closely during the few past months.  Now it turns out I should have
(but then again, it wasn't that much work).

> Nevertheless, it would be reasonable to try to merge your changes after
> Alexandre is done, so that we are sure that no detail has been lost. If
> you feel comfortable with CVS merge facilities I would encourage you to
> merge yourself (send me your SF userid for CVS access). This however may
> be tedious, so if you do not volunteer to do this part, I can do it.

I'd be happy to do the merge.  My SF ID is "rotofoil".

>> I have verified that OpenC++ is still compilable with Gcc 2.95, but
>> I have not tested it with any other compiler.
>
> Why don't you join the project? There is quite reasonable testing
> environment on Source Forge Compile Farm.

I do not have any experience with SF's compile farm and could use some
help in testing.  Unfortunately, I do not have the time to manually
configure and test OpenC++ on each and every architecture and
compiler.  Is there an automated way to try a compilation on all
architectures?

Which architectures need to be considered as ``supported by OpenC++''?

Kind regards,
Michael
-- 
ho...@sa..., ho...@in...
http://www.sax.de/~hohmuth/

[Opencxx-users] [opencxx] 2.6.1 released

[Grzegorz J. <ja...@he...>]

Hi,

I released 2.6.1, which fixes the packaging error pointed out by Michael
Hohmuth (the observable effect is package size reduction by 400KB).

Best regards
Grzegorz

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] [opencxx] Let's disable which that we do not use

[Grzegorz J. <ja...@he...>]

Hi,

I reviewed the opencxx pages on SourceForge and I found out that we have a
problem with unused SF features, e.g. discussion forums.

There are a few dated messages in miscellaneous forums that have never
been responded to, because we are not monitoring the forums. Also we are
not really using bug tracker or patch manager.

I found out that SourceForge now enables hiding those features and I vote
just to hide them. There is nothing more pathetic than a discussion forum
with the latest message posted a year ago. My first impression when I see
something like that is "this project is an ex-project" :-) It also does
not give a clue, that there is a real mailing list.

Below you find summary of open SF features. I suggest to shut down
everything except CVS (of course), News, File releases and Mailing Lists
(not to be confused with Forums). I Hope I did not miss anything.

I am also working on a facelift of the webpage at opencxx.sf.net, where I
want to put the info e.g. on how to report bugs or request features.

Before I shut down the services I will review the old messages there and
reply to them.

I am not sure if there is a way of disabling forums associated with project
news, while still using the news. If not, I will create a new user who will
be monitoring the news forums and forward his SF e-mails to our mailing
list.

Please let me know ASAP if you have any objections. If not, I will shut the
unused services within one-week timeframe.

Best regards
Grzegorz

------ Summary ------

Forums | Open Discussion:
5 messages

Bugs Tracker:
6 bugs total, 1 open (gcc-3.x does not compile)

Patches:
one patch by Michael Dupont

Feature requests:
1 total (I will follow up)

Support requests:
none

Tasks | General
We are not using it since Jan 2001

Docs
none

There are single messages in discussion assiciated with News; I replied to
them already and I am now monitoring them.

------ End ------


##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] Re: [opencxx] comments problem, continued

[Stanislav K. <st...@uc...>]

Ok, i have reduced my metaclass files and in the process found out that
if I do not include iostream, comments work like a charm, but the minute
that I do include iostream I start getting my compilation errors

Here are my new files:

//--------------------------------------
FILE: Makefile
//--------------------------------------

small.o: small.cpp Meta.so
	occ  -C -g -SMeta -- small.cpp

Meta.so: Meta.mc Meta.h 
	occ  -C -m -g Meta.mc -I/home/wokls

// ------------------------------------------------
// FILE: Meta.h
// ------------------------------------------------

#include "openc++2.5.12/src/mop.h"

#ifndef Meta_H
#define Meta_H

class Meta : public Class {
	public:
		
		//void TranslateClass(Environment *);
		void TranslateMemberFunction(Environment* env, Member& m);
		static bool Initialize();
};

#endif


// -----------------------------------------------

// FILE: Meta.mc
// ------------------------------------------------

#include "Meta.h"
//#include <iostream.h>
#include <stdio.h>

bool Meta::Initialize() {
	Class::ChangeDefaultMetaclass  ("Meta");
	Class::SetMetaclassForFunctions("Meta");
	return true ;
}

void Meta::TranslateMemberFunction(Environment * e, Member& m)
{
	m.Comments()->Display();
}

// ------------------------------------------------
// FILE: small.cpp
// ------------------------------------------------

/* jfdklafdjaslkfdsa
* fdsajkflasjflksafjdsal;
* sadfjaskfjalskfjlas
*/

// hiu


/* hi how are you doing? */


int main(void)
{
	return 0;
}

// ------------------------------------------------

// ------------------------------------------------
// FILE: small.ii
// ------------------------------------------------

# 2 "small.cpp"
/* jfdklafdjaslkfdsa
* fdsajkflasjflksafjdsal;
* sadfjaskfjalskfjlas
*/

// hiu



/* hi how are you doing? */



int main(void)
{

        return 0;

}


# 23 "small.ii"

// ------------------------------------------------


// ------------------------------------------------
// FILE: Meta.ii
// ------------------------------------------------

Not included, because its huge



when I include iostream, i get the following errors:

occ  -C -m -g Meta.mc -I/home/wokls  
/usr/include/g++-3/iostream.h:31: parse error before `32 "/usr/include/g++-3/iostream.h"'
/usr/include/g++-3/iostream.h:272: parse error before `# 3'
Meta.mc:6: parse error before `bool Meta'
Meta.mc:10: parse error before `) ;'
Meta.mc:11: parse error before `true ;'
Meta.mc:14: parse error before `void Meta'
Meta.mc:21: parse error before `'
7 Error(s).  OpenC++ stops.
make: *** [Meta.so] Error 1



thanks again



Grzegorz Jakacki wrote:

>On Wed, 5 Mar 2003, stan wrote:
>
>  
>
>>How does one go about getting openC++ to retain comments?
>>
>>Gregorz Jakacki asked me to supply my *.occ
>>files, what are these?
>>    
>>
>
>I am sorry. I meant .ii files, the files which contain preprocessed and
>translated sources, which are shipped to your underlying compiler.
>Something like
>
>    occ -E -SMeta -- Test.cpp
>
>should leave Test.ii in you working directory. Sorry for my mistake. Also,
>if possible, send your source (Test.cpp) and the source of your metaclass.
>Before you post, please reduce the size of Test.cpp to minimum size that
>still gives you an error message.
>
>Best regards
>Grzegorz
>
>  
>
>>I am using:
>>OpenC++ 2.5.12
>>GCC 2.96-85 on RH7.1
>>
>>my make file looks like this:
>>
>>Test.o: Test.cpp Test.h Meta.so
>>	occ  -g -c -SMeta -- Test.cpp
>>
>>Meta.so: Meta.mc Meta.h XMLWalker.o
>>	occ -m -g Meta.mc -I/home/uusaa XMLWalker.o
>>
>>XMLWalker.o: XMLWalker.cpp XMLWalker.h
>>	occ -c -g XMLWalker.cpp -I/home/uusaa
>>
>>my XMLWalker, walks the ast spitting out xml as it goes.
>>
>>I tried inserting the -C option into the occ compile commands and when I
>>do this,
>>i get the following errors:
>>
>>occ -C -g -c -SMeta -- Test.cpp
>>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:35: parse error before `36 "/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new"'
>>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:44: parse error before `} ;'
>>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:54: parse error before `void *'
>>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:67: parse error before `# 34'
>>/usr/include/g++-3/iostream.h:31: parse error before `32 "/usr/include/g++-3/iostream.h"'
>>/usr/include/g++-3/iostream.h:272: parse error before `# 54'
>>/usr/include/g++-3/stl_alloc.h:344: parse error before `_Obj *'
>>/usr/include/g++-3/stl_alloc.h:351: parse error before `void *'
>>/usr/include/g++-3/stl_alloc.h:398: parse error before `: static'
>>/usr/include/g++-3/stl_alloc.h:406: parse error before `( __n'
>>10 Error(s).  OpenC++ stops.
>>make: *** [Test.o] Error 1
>>any ideas?
>>
>>thanks
>>
>>
>>
>>    
>>
>
>##################################################################
># Grzegorz Jakacki                       Huada Electronic Design #
># Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
># tel. +86-10-64365577 x2074               Beijing 100015, China #
># Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
>##################################################################
>
>
>  
>

[Opencxx-users] Re: [opencxx] comments problem, continued

[Grzegorz J. <ja...@he...>]

On Wed, 5 Mar 2003, stan wrote:

> How does one go about getting openC++ to retain comments?
>
> Gregorz Jakacki asked me to supply my *.occ
> files, what are these?

I am sorry. I meant .ii files, the files which contain preprocessed and
translated sources, which are shipped to your underlying compiler.
Something like

    occ -E -SMeta -- Test.cpp

should leave Test.ii in you working directory. Sorry for my mistake. Also,
if possible, send your source (Test.cpp) and the source of your metaclass.
Before you post, please reduce the size of Test.cpp to minimum size that
still gives you an error message.

Best regards
Grzegorz

>
> I am using:
> OpenC++ 2.5.12
> GCC 2.96-85 on RH7.1
>
> my make file looks like this:
>
> Test.o: Test.cpp Test.h Meta.so
> 	occ  -g -c -SMeta -- Test.cpp
>
> Meta.so: Meta.mc Meta.h XMLWalker.o
> 	occ -m -g Meta.mc -I/home/uusaa XMLWalker.o
>
> XMLWalker.o: XMLWalker.cpp XMLWalker.h
> 	occ -c -g XMLWalker.cpp -I/home/uusaa
>
> my XMLWalker, walks the ast spitting out xml as it goes.
>
> I tried inserting the -C option into the occ compile commands and when I
> do this,
> i get the following errors:
>
> occ -C -g -c -SMeta -- Test.cpp
> /usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:35: parse error before `36 "/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new"'
> /usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:44: parse error before `} ;'
> /usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:54: parse error before `void *'
> /usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:67: parse error before `# 34'
> /usr/include/g++-3/iostream.h:31: parse error before `32 "/usr/include/g++-3/iostream.h"'
> /usr/include/g++-3/iostream.h:272: parse error before `# 54'
> /usr/include/g++-3/stl_alloc.h:344: parse error before `_Obj *'
> /usr/include/g++-3/stl_alloc.h:351: parse error before `void *'
> /usr/include/g++-3/stl_alloc.h:398: parse error before `: static'
> /usr/include/g++-3/stl_alloc.h:406: parse error before `( __n'
> 10 Error(s).  OpenC++ stops.
> make: *** [Test.o] Error 1
> any ideas?
>
> thanks
>
>
>

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] [opencxx] Re: ANSI C++/Gcc 3.2 compatibility patches

[Grzegorz J. <ja...@he...>]

Michael,

On Wed, 5 Mar 2003, Michael Hohmuth wrote:

> Dear OpenC++ maintainers,
>
> please find attached below a patch relative to OpenC++ 2.6.  This
> patch makes the OpenC++ source more ANSI-C++ compliant and fixes some
> minor compatibility problems for compilation with Gcc 3.2.

Thank you very much for your contribution. The gcc-3.x compatibility issue
surfaces regularily. Recently Alexandre Tolmos has merged the old patch
prepared by Michael Buro addressing exactly this same problem, and is
currently in the phase of testing it on Compile Farm.

This unfortunately means, that some work has been done twice.
Nevertheless, it would be reasonable to try to merge your changes after
Alexandre is done, so that we are sure that no detail has been lost. If
you feel comfortable with CVS merge facilities I would encourage you to
merge yourself (send me your SF userid for CVS access). This however may
be tedious, so if you do not volunteer to do this part, I can do it.

Next time, I would suggest that you first contact the mailing list or me
directly to discuss the contribution you would like to make. The project
resources are scarce and I would hate to waste time of any contributor.

I am not sure if you are subscribing to the OpenC++ mailing list, since
gcc-3.x patch was discussed here. If not, I would strongly suggest that
you do subscribe (see http://www.csg.is.titech.ac.jp/~chiba/openc++.html
for instructions on how to subscribe). The list is low-volume (typically
0-5 messages a day) and keeps you updated about the project news. It is
archived at

    http://sourceforge.net/mailarchive/forum.php?forum_id=11832

since Sep 2002.

> I have
> verified that OpenC++ is still compilable with Gcc 2.95, but I have
> not tested it with any other compiler.

Why don't you join the project? There is quite reasonable testing
environment on Source Forge Compile Farm.

> I have not included compatibility patches for the garbage collector.
> The version currently included with OpenC++ (4.14) has at least one
> problem: It does not provide an "operator new [](size_t, GCPlacement)".
>
> In fact, I am not using the GC version supplied with OpenC++, but a
> newer version of the Boehm GC (currently version 6.0), which works
> fine with Gcc 3.2.  Newer Boehm GCs already are libtool'ized and
> autoconf'd, and they should be easy to integrate into OpenC++'s
> configuration system, but I don't know enough about libtool to do the
> integration myself.  I currently just replace the contents of the gc
> directory with the new version and re-run configure in it.

I am deliberately delaying upgrade to new GC, because all currenly
available versions have problems with 'make dist'. I fixed it, Hans Boehm
will include it in 6.2alpha4 ("to be released in the next few days") and
this is probably what we will incorporate soon (also because of MacOS X
port by Alexandre Tolmos).

> (Also, I noticed that the OpenC++ 2.6 tarball also includes a second
> copy of the GC in src/gc, which is never used and should be deleted.)

This is my packaging error. I will fix it when I have a bit of time and
release as 2.6.1 . Thanks for spotting this.

Best regards
Grzegorz

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] [opencxx] comments problem, continued

[stan <st...@uc...>]

How does one go about getting openC++ to retain comments?

Gregorz Jakacki asked me to supply my *.occ
files, what are these?

I am using:
OpenC++ 2.5.12
GCC 2.96-85 on RH7.1

my make file looks like this:

Test.o: Test.cpp Test.h Meta.so
	occ  -g -c -SMeta -- Test.cpp

Meta.so: Meta.mc Meta.h XMLWalker.o
	occ -m -g Meta.mc -I/home/uusaa XMLWalker.o 

XMLWalker.o: XMLWalker.cpp XMLWalker.h 
	occ -c -g XMLWalker.cpp -I/home/uusaa

my XMLWalker, walks the ast spitting out xml as it goes.

I tried inserting the -C option into the occ compile commands and when I 
do this,
i get the following errors:

occ -C -g -c -SMeta -- Test.cpp
/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:35: parse error before `36 "/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new"'
/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:44: parse error before `} ;'
/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:54: parse error before `void *'
/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:67: parse error before `# 34'
/usr/include/g++-3/iostream.h:31: parse error before `32 "/usr/include/g++-3/iostream.h"'
/usr/include/g++-3/iostream.h:272: parse error before `# 54'
/usr/include/g++-3/stl_alloc.h:344: parse error before `_Obj *'
/usr/include/g++-3/stl_alloc.h:351: parse error before `void *'
/usr/include/g++-3/stl_alloc.h:398: parse error before `: static'
/usr/include/g++-3/stl_alloc.h:406: parse error before `( __n'
10 Error(s).  OpenC++ stops.
make: *** [Test.o] Error 1
any ideas?

thanks

[Opencxx-users] [opencxx] Fw: ANSI C++/Gcc 3.2 compatibility patches

[Shigeru C. <ch...@is...>]

I forward this message since the mailing list did not accept it
due to non-member submission.

Chiba

[Opencxx-users] Re: [opencxx] errors in getting comments

[Grzegorz J. <ja...@he...>]

On Mon, 3 Mar 2003, Stanislav Kaushanskiy wrote:

>
> When trying to get comments from a source file,
> the appendix to the manual states to us the -C option
> to compile the program, when doing so I get the
> following errors, if I do not use the -C option, these errors
> do not appear.
>
> what am I doing wrong?

Please send:

  * .occ file you get without -C
  * .occ file you get with -C
  * version of C++ compiler you are using

Best regards
Grzegorz

>
> thanks
>
> >occ -C -g -c -SMeta -- Test.cpp
> >/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:35: parse error before `36 "/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new"'
> >/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:44: parse error before `} ;'
> >/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:54: parse error before `void *'
> >/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:67: parse error before `# 34'
> >/usr/include/g++-3/iostream.h:31: parse error before `32 "/usr/include/g++-3/iostream.h"'
> >/usr/include/g++-3/iostream.h:272: parse error before `# 54'
> >/usr/include/g++-3/stl_alloc.h:344: parse error before `_Obj *'
> >/usr/include/g++-3/stl_alloc.h:351: parse error before `void *'
> >/usr/include/g++-3/stl_alloc.h:398: parse error before `: static'
> >/usr/include/g++-3/stl_alloc.h:406: parse error before `( __n'
> >10 Error(s).  OpenC++ stops.
> >make: *** [Test.o] Error 1
> >
> >
>
>
>

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] [opencxx] errors in getting comments

[Stanislav K. <st...@uc...>]

When trying to get comments from a source file,
the appendix to the manual states to us the -C option
to compile the program, when doing so I get the
following errors, if I do not use the -C option, these errors
do not appear.

what am I doing wrong?

thanks

>occ -C -g -c -SMeta -- Test.cpp
>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:35: parse error before `36 "/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new"'
>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:44: parse error before `} ;'
>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:54: parse error before `void *'
>/usr/lib/gcc-lib/i386-redhat-linux/2.96/include/new:67: parse error before `# 34'
>/usr/include/g++-3/iostream.h:31: parse error before `32 "/usr/include/g++-3/iostream.h"'
>/usr/include/g++-3/iostream.h:272: parse error before `# 54'
>/usr/include/g++-3/stl_alloc.h:344: parse error before `_Obj *'
>/usr/include/g++-3/stl_alloc.h:351: parse error before `void *'
>/usr/include/g++-3/stl_alloc.h:398: parse error before `: static'
>/usr/include/g++-3/stl_alloc.h:406: parse error before `( __n'
>10 Error(s).  OpenC++ stops.
>make: *** [Test.o] Error 1
>  
>

[Opencxx-users] Re: [opencxx] getting comments

[Grzegorz J. <ja...@he...>]

Stan,

This is a general rule, that when you are looking for help with any kind
of error, you need to supply at least the error message, and preferably
the short code along with complete info that enables the recipients to
recreate your error.

Please repost with more info.

Best regards
Grzegorz

On Thu, 27 Feb 2003, stan wrote:

> In the appendix to the manual,
> under the Comments() functions, it says to run the compiler
> with the -C option when I do this I get parse errors
> in ../include/new
> and in ../stl_alloc.h
>
> I am running openc++ 2.5.12
> and gcc 2.96
>
> any ideas?
>
> thanks,
>
> stan
>
>
>
>
>

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] Re: [opencxx] Porting to MacOs X

[Grzegorz J. <ja...@he...>]

On Thu, 27 Feb 2003, Alexandre Tolmos wrote:

[snip]
> I created a new branch (sandbox_kthulhu) and checked in my changes. You
> can remove the previous branch.

Thanks.

> Now I have a problem for accessing "cf.sourceforge.net". It asks for my
> password and denies permission to login.
>
> Note: I used the following command: "ssh kt...@cf..."

The command is OK. I usually add '-v' option, so that I can see what goes
wrong.

> I assume I should use the same password I use for login to "cvs.
> sourceforge.net". Right?

Right in general. Although if you deposited your SSH keys with SF, then
possibly your ssh connection is being autenticated via "passphrase", not
"password"? In such a case the client should clearly ask for "passphrase"
and you respond with the word that you supplied when creating your SSH ke=
ys.

There may be also one more issue. The page

    http://sourceforge.net/docman/display_doc.php?docid=3D10472&group_id=3D=
1

says

     Compile Farm services are now provided on an opt-in basis to project
     developers. To opt-in for compile farm access, check the 'Opt-in for
     Compile Farm' check box on your Account Maintenance page. After
     opting-in for Compile Farm access, it will take approximately 24 hou=
rs
     for your account to become available on cf.sourceforge.net

Is this box checked on your account?

If none of the above is solves the issue, then we have to contact SF
support. Login, go to www.sf.net, click "Contact SF.net Support" (left co=
lumn), and
then click "submit a Support request".

Hopefully we can get it right soon. Let me know if any issues.

Best regards
Grzegorz

PS: I added the info about checkbox to the docs.

>
> Alex
>
> -----------------------------------------------------------------------=
-
> -
> Alexandre Tolmos
> E-mail:=A0...@fr...
> ICQ: 92964905
> -----------------------------------------------------------------------=
-
> -
> "Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn."
> -----------------------------------------------------------------------=
-
> -
>
>
>
>

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] [opencxx] getting comments

[stan <st...@uc...>]

In the appendix to the manual,
under the Comments() functions, it says to run the compiler
with the -C option when I do this I get parse errors
in ../include/new
and in ../stl_alloc.h

I am running openc++ 2.5.12
and gcc 2.96

any ideas?

thanks,

stan

[Opencxx-users] Re: [opencxx] Porting to MacOs X

[Alexandre T. <kt...@fr...>]

Le jeudi, 27 f=E9v 2003, =E0 03:00 Europe/Paris, Grzegorz Jakacki a =E9cr=
it :

> [...]
>
> Great. However could you change your sandbox tag name to
>
>     sandbox_kthulu
>
> or
>
>     sandbox_kthulu_<ANYTHING>
>
> ?
> (See the doc, the guidelines are there).
>
> It is neccessary, because there is a script generating daily snapshots =
=20
> of
> branches, and it has to be able to distinguish between sandboxes from
> other branches.

I created a new branch (sandbox_kthulhu) and checked in my changes. You =20
can remove the previous branch.

Now I have a problem for accessing "cf.sourceforge.net". It asks for my =20
password and denies permission to login.

Note: I used the following command: "ssh kt...@cf..."

I assume I should use the same password I use for login to "cvs. =20
sourceforge.net". Right?

Alex

------------------------------------------------------------------------=20
-
Alexandre Tolmos
E-mail:=A0...@fr...
ICQ: 92964905
------------------------------------------------------------------------=20
-
"Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn."
------------------------------------------------------------------------=20
-

[Opencxx-users] Re: [opencxx] Porting to MacOs X

[Grzegorz J. <ja...@he...>]

On Wed, 26 Feb 2003, Alexandre Tolmos wrote:

[...]
> Nope, I was using "ext" but the first ":" was missing (it's also
> missing in your doc).

Sorry about that. Fixed.

> >> Moreover, when executing the command "cvs rtag -b <sandbox>", a module
> >> name must be provided. What should it be ? There's nothing about it in
> >> your Html doc...
> >
> > Look for "Browse CVS Repository" on OpenC++ CVS page. This is an entry
> > point to the web cvs browser. From there you can find out the module
> > names
> > (the module you need is "opencxx"). With web interface you will also be
> > able to verify if you succeeded in creating a sandbox branch.
> >
> > Let me know if you need any more info. I would also suggest that you
> > write
> > down the information that you think was missing from docs, so that
> > other
> > developers can avoid the problems you have.
> >
> > Contact me if you need further info.
>
> I successfully created my own branch and checked out the tree

Great. However could you change your sandbox tag name to

    sandbox_kthulu

or

    sandbox_kthulu_<ANYTHING>

?
(See the doc, the guidelines are there).

It is neccessary, because there is a script generating daily snapshots of
branches, and it has to be able to distinguish between sandboxes from
other branches.

> but Cvs
> fails when I try to commit my changes (the Sf Cvs server times out...)

This may be a temporary problem. However, if things still do not work
today, put a line

    LogLevel DEBUG

in your .ssh/config and try to commit. You should get lots of debugging
output from ssh --- look into it and/or send it to me.

Best regards
Grzegorz


##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] Re: [opencxx] Porting to MacOs X

[Alexandre T. <kt...@fr...>]

Le mardi, 25 f=E9v 2003, =E0 02:00 Europe/Paris, Grzegorz Jakacki a =E9cr=
it :

> On Mon, 24 Feb 2003, Alexandre Tolmos wrote:
>
>> Grzegorz,
>>
>>> You have been added to the project.
>>
>> Great :)
>>
>>>
>>>> I'll soon be ready to checkin Michael's Gcc 3 fix to Occ. I need =20
>>>> some
>>>> time to read the dox on Sf about Ssh and Cvs, I haven't used that
>>>> stuff
>>>> for a while.
>>>
>>> Let me know if I can be of any help.
>>
>> I tried to login with cvs and execute the rtag command but it fails
>> when I enter my password. I assume this is the same password I used
>> when I created my account on Sf, right ?
>
> The fact that you tried to login suggests that you are using :pserver:
> access method, while you should be using :ext:. The easiest way to go =20
> is
> to browse www.sf.net/projects/opencxx and click "CVS". You will find =20
> the
> explanation of developer's access setup at the bottom of the page. It =20
> also
> has links to FAQs on SF CVS and ssh access (in particular how to upload
> ssh key, so that CVS does not ask you for a password each time).

Nope, I was using "ext" but the first ":" was missing (it's also =20
missing in your doc).

>
>> Moreover, when executing the command "cvs rtag -b <sandbox>", a module
>> name must be provided. What should it be ? There's nothing about it in
>> your Html doc...
>
> Look for "Browse CVS Repository" on OpenC++ CVS page. This is an entry
> point to the web cvs browser. From there you can find out the module =20
> names
> (the module you need is "opencxx"). With web interface you will also be
> able to verify if you succeeded in creating a sandbox branch.
>
> Let me know if you need any more info. I would also suggest that you =20
> write
> down the information that you think was missing from docs, so that =20
> other
> developers can avoid the problems you have.
>
> Contact me if you need further info.

I successfully created my own branch and checked out the tree but Cvs =20
fails when I try to commit my changes (the Sf Cvs server times out...)

------------------------------------------------------------------------=20
-
Alexandre Tolmos
E-mail:=A0...@fr...
ICQ: 92964905
------------------------------------------------------------------------=20
-
"Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn."
------------------------------------------------------------------------=20
-

[Opencxx-users] Re: [opencxx] Porting to MacOs X

[Grzegorz J. <ja...@he...>]

On Mon, 24 Feb 2003, Alexandre Tolmos wrote:

> Grzegorz,
>
> > You have been added to the project.
>
> Great :)
>
> >
> >> I'll soon be ready to checkin Michael's Gcc 3 fix to Occ. I need some
> >> time to read the dox on Sf about Ssh and Cvs, I haven't used that
> >> stuff
> >> for a while.
> >
> > Let me know if I can be of any help.
>
> I tried to login with cvs and execute the rtag command but it fails
> when I enter my password. I assume this is the same password I used
> when I created my account on Sf, right ?

The fact that you tried to login suggests that you are using :pserver:
access method, while you should be using :ext:. The easiest way to go is
to browse www.sf.net/projects/opencxx and click "CVS". You will find the
explanation of developer's access setup at the bottom of the page. It also
has links to FAQs on SF CVS and ssh access (in particular how to upload
ssh key, so that CVS does not ask you for a password each time).

> Moreover, when executing the command "cvs rtag -b <sandbox>", a module
> name must be provided. What should it be ? There's nothing about it in
> your Html doc...

Look for "Browse CVS Repository" on OpenC++ CVS page. This is an entry
point to the web cvs browser. From there you can find out the module names
(the module you need is "opencxx"). With web interface you will also be
able to verify if you succeeded in creating a sandbox branch.

Let me know if you need any more info. I would also suggest that you write
down the information that you think was missing from docs, so that other
developers can avoid the problems you have.

Contact me if you need further info.

Best regards
Grzegorz

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] Re: [opencxx] translating global declarations with openc++

[Grzegorz J. <ja...@he...>]

On Mon, 24 Feb 2003, stan wrote:

> is there a way to translate global-scoped declarations such as:
> enums, typedefs, global variables, etc  using OpenC++?
>
> for global functions there is a SetDefaultMetaclassForFunctions()
> is there a similar system for global declarations?

As far as I know: no. However it should not be difficult to extend OpenC++
to support those. Contact me for more info if you decide to go for it.

>
> what about function prototypes in global scope?

Similarly.

Best regards
Grzegorz

>
> thank you,
>
>
> stan
>
>
>

##################################################################
# Grzegorz Jakacki                       Huada Electronic Design #
# Senior Engineer, CAD Dept.              1 Gaojiayuan, Chaoyang #
# tel. +86-10-64365577 x2074               Beijing 100015, China #
# Copyright (C) 2002 Grzegorz Jakacki, HED. All Rights Reserved. #
##################################################################

[Opencxx-users] [opencxx] a walking strategy

[stan <st...@uc...>]

I am using a customized MyWalker class which is derived from Walker and 
overloads all the Translate*() virtual functions.  This walker goes 
through the ptree and spits out an xml representation of the parse tree.

In my metaclass I overload the TranslateClass() and 
TranslateMemberFunction() methods,
in these methods I basically call Translate(MyWalker) on the respective 
ptrees

I have not explored the OpenC++ source code, just the documentation.
My question is:
is this the best way to get at all of the parse tree for a source code?


thank you,

stan