Обсуждение: solaris port

Hi guys!

first of all I would like to tell that the 2.4 release of wx builds ok
(that's a good point to start, I think).
But, did someone here succeeded to build wx 2.5 snapshots on Solaris ?
If so I would be interested in the configuration (compiler, os and so
on) used to do the build.

Here is the build error of our wx snap (I'll give the full configure
line later):
./bk-deps g++ -c -o monolib_intl.o  -D__WXGTK__  -I./include
-I./src/expat/lib -DwxUSE_BASE=1 -D__WXDEBUG__ -Ilib/wx/include/gtk2ud-2.5
-I./include -threads -I/home/testuser/usr/include/gtk-2.0
-I/home/testuser/usr/lib/gtk-2.0/include
-I/home/testuser/usr/include/atk-1.0
-I/home/testuser/usr/include/pango-1.0 -I/usr/openwin/include
-I/home/testuser/usr/include/glib-2.0
-I/home/testuser/usr/lib/glib-2.0/include -I
/home/testuser/usr/include -g -O0 -I /home/testuser/usr/include
-pthreads -D_REENTRANT -Wall src/common/intl.cpp
In file included from src/common/intl.cpp:49:
include/wx/intl.h:40: warning: `_N' redefined
/usr/include/iso/ctype_iso.h:38: warning: this is the location of the
previous definition
src/common/intl.cpp: In method `bool wxPluralFormsScanner::nextToken()':
src/common/intl.cpp:241: `_N' undeclared (first use this function)
src/common/intl.cpp:241: (Each undeclared identifier is reported only once
src/common/intl.cpp:241: for each function it appears in.)
gmake: *** [monolib_intl.o] Error 1

Andreas does it talk to you (I mean, something to do with one of your
patches ?) ?
In fact I had no time to look further in the source to see if I
understand a piece of line.

Regards,
Raphaël

Raphaël Enrici wrote:

> -pthreads -D_REENTRANT -Wall src/common/intl.cpp
> In file included from src/common/intl.cpp:49: include/wx/intl.h:40:
> warning: `_N' redefined
> /usr/include/iso/ctype_iso.h:38: warning: this is the location of the
> previous definition


> src/common/intl.cpp: In method `bool wxPluralFormsScanner::nextToken()':
> src/common/intl.cpp:241: `_N' undeclared (first use this function)
> src/common/intl.cpp:241: (Each undeclared identifier is reported only
> once
> src/common/intl.cpp:241: for each function it appears in.)

This doesn't talk to me, it's not caused by our fixes.
_N is one of the gettext macros, and there seems to be some conflict
between wx (intl.h) and system (ctype_iso.h).
I'd recommend you use the official wx cvs HEAD to get it compiled. If
problems persist, you (maybe) get some support on wx-hackers, or we'll
find out ourselves how to solve it. In the second case, we certainly
should post the fix to the wx patch manager.

After that, we upgrade our snapshot to wx cvs HEAD, apply our patches
and continue hoping that there might be some time in the distant future
where we can use plain vanilla wx....

Regards,
Andreas

Andreas Pflug wrote:

> Raphaël Enrici wrote:
>
>> -pthreads -D_REENTRANT -Wall src/common/intl.cpp
>> In file included from src/common/intl.cpp:49: include/wx/intl.h:40:
>> warning: `_N' redefined
>> /usr/include/iso/ctype_iso.h:38: warning: this is the location of the
>> previous definition
>
> > src/common/intl.cpp: In method `bool
> wxPluralFormsScanner::nextToken()':
>
>> src/common/intl.cpp:241: `_N' undeclared (first use this function)
>> src/common/intl.cpp:241: (Each undeclared identifier is reported only
>> once
>> src/common/intl.cpp:241: for each function it appears in.)
>
>
> This doesn't talk to me, it's not caused by our fixes.
> _N is one of the gettext macros, and there seems to be some conflict
> between wx (intl.h) and system (ctype_iso.h).

Ok,
I found this in docs/gtk/install.txt

--enable-permissive     Enable compilation without checking for strict
                            ANSI conformance.  Useful to prevent the build
                            dying with errors as soon as you compile with
                            Solaris' ANSI-defying headers.

Maybe it's the way to follow...

also found this:

- Solaris compilation with gcc: if the compiler has problems with the variable
  argument functions, try putting the gcc fixinclude file paths early in the
  include path.


> I'd recommend you use the official wx cvs HEAD to get it compiled. If
> problems persist, you (maybe) get some support on wx-hackers, or we'll
> find out ourselves how to solve it. In the second case, we certainly
> should post the fix to the wx patch manager.

ok, I'll try this too.

> After that, we upgrade our snapshot to wx cvs HEAD, apply our patches
> and continue hoping that there might be some time in the distant
> future where we can use plain vanilla wx....

"hope"... that's a nice title for a song, I'm going to write some lyrics :)

Thank you for your explanations, I'll tell you if I get something working.

Regards,
Raphaël

Raphaël Enrici wrote:

>
>> After that, we upgrade our snapshot to wx cvs HEAD, apply our patches
>> and continue hoping that there might be some time in the distant
>> future where we can use plain vanilla wx....
>
>
> "hope"... that's a nice title for a song, I'm going to write some
> lyrics :)


So gimme hope, wx-a
hope, wx-a
gimme hope, wx-a,
'fore frustration come
gimme hope, wx-a
hope, wx-a
hope before frustration come.


--> your lyrics here

Maybe we can persuade Vadim to contribute a verse...

Regards,
Andreas