round 2, OSX

Hello again everyone.

Here is my second round with the OS/X code. This time its
not much
different but it plays nicer with everything else and makes
use of the
autoconf/make setup. But I did slip in gtkspell support with
the
native osx GTK build script and app bundle.

This time it has two binary files that svn diff didnt want
to pick up.
gajim.icns goes into gajim/data/pixmaps/gajim.icns and
keyedobjects.nib goes into gajim/data/nibs/MainMenu.nib.
That is used
to setup the OS/X style menu.

gtkspell uses aspell which has a different distribution file
for the
dictionary for each language. Right now I just pull in the
en
dictionary. Please forgive the english centric view. Does
anyone have
any thoughts on how to distribute compiled app bundles with
particular
languages? The dictionaries aren't terribly large, the
english one is
3.5M uncompressed. The package is already 66M uncompressed
and 19M
compressed. That is almost all the GTK runtime.

James

_______________________________________________
Gajim-devel mailing list
Gajim-develgajim.org

https://lists.gajim.org/cgi-bin/listinfo/gajim-devel

James Newton a écrit :
> Hello again everyone.
> 
> Here is my second round with the OS/X code. This time
its not much
> different but it plays nicer with everything else and
makes use of the
> autoconf/make setup. But I did slip in gtkspell support
with the
> native osx GTK build script and app bundle.
> 
> This time it has two binary files that svn diff didnt
want to pick up.
> gajim.icns goes into gajim/data/pixmaps/gajim.icns and
> keyedobjects.nib goes into
gajim/data/nibs/MainMenu.nib. That is used
> to setup the OS/X style menu.
> 
> gtkspell uses aspell which has a different distribution
file for the
> dictionary for each language. Right now I just pull in
the en
> dictionary. Please forgive the english centric view.
Does anyone have
> any thoughts on how to distribute compiled app bundles
with particular
> languages? The dictionaries aren't terribly large, the
english one is
> 3.5M uncompressed. The package is already 66M
uncompressed and 19M
> compressed. That is almost all the GTK runtime.
> 
> James

Hi,

First aof all, thanks for your port !
But you forgot to attach the diff 

Is there still a TODO list for your package ? Is it small
enought so I 
can commit your diff ?

-- 
Yann
_______________________________________________
Gajim-devel mailing list
Gajim-develgajim.org

https://lists.gajim.org/cgi-bin/listinfo/gajim-devel

On 5/15/07, yann <asterixlagaule.org> wrote:
> James Newton a écrit :
> > Hello again everyone.
> >
> > Here is my second round with the OS/X code. This
time its not much
> > different but it plays nicer with everything else
and makes use of the
> > autoconf/make setup. But I did slip in gtkspell
support with the
> > native osx GTK build script and app bundle.
> >
> > This time it has two binary files that svn diff
didnt want to pick up.
> > gajim.icns goes into gajim/data/pixmaps/gajim.icns
and
> > keyedobjects.nib goes into
gajim/data/nibs/MainMenu.nib. That is used
> > to setup the OS/X style menu.
> >
> > gtkspell uses aspell which has a different
distribution file for the
> > dictionary for each language. Right now I just
pull in the en
> > dictionary. Please forgive the english centric
view. Does anyone have
> > any thoughts on how to distribute compiled app
bundles with particular
> > languages? The dictionaries aren't terribly large,
the english one is
> > 3.5M uncompressed. The package is already 66M
uncompressed and 19M
> > compressed. That is almost all the GTK runtime.
> >
> > James
>
> Hi,
>
> First aof all, thanks for your port !
> But you forgot to attach the diff 
>
> Is there still a TODO list for your package ? Is it
small enought so I
> can commit your diff ?
>
> --
> Yann

Yann,

oops! Here's the patch.

The TODO is still relatively long. I think the only thing
that is
preventing me from using Gajim on my mac now is the GTK cut
and paste
bug. Its definitely alpha, but it wont hurt the other
platforms. I've
tested the autoconf/make changes on linux but have not
tested on
windows. But I leave it up to you.  I can
continue to work as I am
for now, that is no problem.

James

_______________________________________________
Gajim-devel mailing list
Gajim-develgajim.org

https://lists.gajim.org/cgi-bin/listinfo/gajim-devel

James Newton wrote:
> 
> +#include <CoreFoundation/CoreFoundation.h>
> +#include <CoreServices/CoreServices.h> 

If there's a way NOT to use CoreFoundation and Core* in
principle, by
all means please try to.  It is not implemented in GNUstep
(yet).
Some time ago I spent few restless nights trying to port
AbiWord's
MuckOS port to GNUstep and failed miserably mainly because
of this.  I
find it outrageous that free programs are being ported to
proprietary
OSes anyway, instead of extending them to work better on
variants of
the GNU system.  But I'm not a Gajim developer so my opinion
doesn't
count here.

Frustratingly, 85% of the traffic on this list was
Windoze-oriented.
Now it'll be 85% Windoze and 10% Muck OS X.

_______________________________________________
Gajim-devel mailing list
Gajim-develgajim.org

https://lists.gajim.org/cgi-bin/listinfo/gajim-devel

On 5/15/07, Yavor Doganov <yavorgnu.org> wrote:
> James Newton wrote:
> >
> > +#include <CoreFoundation/CoreFoundation.h>
> > +#include <CoreServices/CoreServices.h>
>
> If there's a way NOT to use CoreFoundation and Core* in
principle, by
> all means please try to.  It is not implemented in
GNUstep (yet).
> Some time ago I spent few restless nights trying to
port AbiWord's
> MuckOS port to GNUstep and failed miserably mainly
because of this.  I
> find it outrageous that free programs are being ported
to proprietary
> OSes anyway, instead of extending them to work better
on variants of
> the GNU system.  But I'm not a Gajim developer so my
opinion doesn't
> count here.
>
> Frustratingly, 85% of the traffic on this list was
Windoze-oriented.
> Now it'll be 85% Windoze and 10% Muck OS X.

If you know of a way to support both OS/X and GNUstep then I
will help
out from my side. I dont own a machine with GNUstep
installed and I
probably never will. But I am all for portability. You say
the
CoreFoundation stuff is not implemented in GNUstep yet. The
only
reason I am using that is for integration with OS/X's
'window manager'
and desktop environment. I'd guess there are other methods
for doing
the same in GNUstep. Incidentally the GTK+ port uses Quartz
and CF.

I'm not gonna get into pissing contest on the philosophy of
free
software. Beating people with an idea will never make them
accept it.
I for one believe in free software. But you have to bridge
the gap to
get people move from one camp to another.
_______________________________________________
Gajim-devel mailing list
Gajim-develgajim.org

https://lists.gajim.org/cgi-bin/listinfo/gajim-devel

James Newton wrote:
> 
> If you know of a way to support both OS/X and GNUstep
then I will
> help out from my side.

99% of the GNUstep apps run on OS/X and on all platforms
that GNUstep
supports (all variants of GNU, *BSD, Windows and
proprieatary Unix
variants).  So basically porting an application for GNUstep
means
porting to OS/X as well.  GNUstep is a free implementation
of the
OpenStep specification (which now exists as Cocoa).  In that
sense,
GNUstep is a free Cocoa and more (some may say
"less").  Typically,
Cocoa developers' interest in GNUstep is only
portability-wise, as
making their programs work for GNUstep is the only way to
port them to
Windows (which is, generally, very attractive for them).

> I dont own a machine with GNUstep installed and I
probably never
> will.

I can say the same thing for OS/X, except by
s/probably/definitely/,
at least until it is released as free software.

> Incidentally the GTK+ port uses Quartz and CF.

So there's no way to make it work.  Your approach seems to
be the
only possibility in that case, I guess.

> I'm not gonna get into pissing contest on the
philosophy of free
> software. Beating people with an idea will never make
them accept
> it.

No surprise.  Ethical considerations are off-topic on nearly
every
mailing list as they make people uneasy.  This
mini-discussion is no
exception.

> But you have to bridge the gap to get people move from
one camp to
> another.

My observations are different.  By porting powerful free
software
(such as Emacs, BASH, grep, sed, GTK+, GIMP, Dia, Gajim,
etc.) to say,
Windows, it makes that operating system more compelling.  So
Windows
users see no reason to migrate to GNU when they have nearly
all the
features that those free programs provide.  Either way, the
popularity
of free software is a shallow goal.  It will inevitably
happen even
without our help.  What's important is the philosophy of the
Free
Software Movement, which doesn't propagate to people's minds
as easily
as the programs themselves do.

_______________________________________________
Gajim-devel mailing list
Gajim-develgajim.org

https://lists.gajim.org/cgi-bin/listinfo/gajim-devel