discuss-gnustep
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: backart - default backend?


From: Guido Schimmels
Subject: Re: backart - default backend?
Date: Fri, 18 Feb 2005 00:23:06 +0100

On Wed, 16 Feb 2005 20:25:38 -0800 (PST)
Gregory John Casamento <greg_casamento@yahoo.com> wrote:

> MJ,
> 
> --- MJ Ray <mjr@dsl.pipex.com> wrote:
> 
> > MJ Ray <mjr@dsl.pipex.com> wrote:
> > > Any or all of the above may be my fault and I'll try to rebuild
> > > some things properly (following instructions carefully, taking
> > > logs and all that) later today to test, but I'd welcome hearing
> > > about advances since November on any of the above.
> > 
> > Further to the above: I tried, but the current back-art release
> > doesn't compile with the current freetype. Reported as bug
> > 12009. Looks like it might be fixed in CVS, but I guess it can't
> > be enough to affect release dates. After all, freetype 2.1.8 had
> > been out for about 5 months already when back 0.9.4 was released.
> 
> I'm not sure what you're implying here, but I believe that Alex Malmberg has a
> patch for the problem. :)   
> 
> Additionally, users CAN download the previous version of freetype and use that
> if they have a newer installation of Linux.
> 
> > So, I wonder how many non-CVS users use back-art? 
> 
> I have a couple of machines that I usually compile the release on, lets see..
> 
> Well, I know that the release compiles on my laptop which has SuSE 9.0 on it,
> and on my Sun box, which has Solaris 9 on it, and I do use back-art on both, 
> so
> I guess there's at least one. :)

Two :)

"""
Here's a post I recently sent to bug-gnustep@gnu.org:

From: Guido Schimmels <__guido__@web.de>
To: bug-gnustep@gnu.org
Subject: Patch to make art backend compile with freetype >=2.1.8
Date: Tue, 15 Feb 2005 12:20:22 +0100
X-Mailer: Sylpheed version 1.9.2 (GTK+ 2.6.2; i686-pc-linux-gnu)

gnustep-back-0.9.4 art backend doesn't compile against freetype >=2.1.8.
As I didn't want to downgrade I tried to adapt it to the latest freetype API.
I stopped working on it as soon as it compiled in the night at 4, so it's 
probably buggy.
So far it seem okay. No crashes at least. All my fonts get listed. [...]
I've only tried GNUstep shortly 2 years ago. And now wanted to have another 
look.
So I don't really know what I'm doing nor how GNUstep with the arts-backend is 
supposed to look like.
But it runs and I can play with the environment.
Probably you have been hacking on that on CVS already and my efforts are 
redundant.
But I'm inpatient. 
If you want something to get done, do it yourself :)
"""

I'm on modem dial-up. CVS checkout is sooo slow. So that was my "shortcut".
I know I'm crazy.

Btw. I'm not going to consider GNUstep for my apps, unless it has an EWMH mode.
I have no interest coding for such a limited userbase (unless you coun't MacOS 
X).
A theme engine is required also. And I'm glad it's seriously being worked on 
now.
But it's not enough. "My way or the highway" is a lousy ass attitude. And it 
doesn't
work, as the the GNUstep community hopefully finally understands.
Steve Jobs may get away with shipping high-end SMP Unix workstations with a 
single-button
mouse. The "you just don't know what is good for you, so daddy will tell you" 
sentiment
isn't appreciated in the open-source world though.

Incidentally, how about placing a link to Cenon prominently on the homepage?
Seems like the best example about what GNUstep can do already on X11 and
how well GNUstep apps can run on MacOS X natively. GNUmail maybe nice and all, 
but for few that is enough incentive to install the GNUstep runtime.

O, and 50%-75% of the GNUstep application links are dead. 
Not helpfull. Leaves a very bad impression :-(

I've read about concern how theming would make GNUstep invisible. Well sorry, 
it can't
get any more invisible than now. That is, people simply don't bother at all.

My interest in GNUstep is with the framework and developer tools. I have no 
interest
telling users which OS they are supposed to run. Especially telling Linux users 
they should be running WindowMaker is not going to happen, as far as I'm 
concerned - especially as I'm the author of the ROX desktop window-manager 
(OroboROX) :-p

For those who haven't plonked me at this point (shall see if I'm too cynical 
here):

So why am I interested in GNUstep?
I'm currently using pygtk for application development. Pygtk hides a lot of the 
GTK+ ugliness and
works quite well, until you try to create custom widgets, that is. Lack of 
introspection also is not so nice.
It's no good when programming mistakes result in cold and dry segfaults. 
Somewhat defeates the purpose of programming in a very-high-level programming 
language like python. I have hope pyobjc is better in this respect, though I 
have'nt tried. But objc's typing system seems like a naturally fit. Hoping the 
pyobjc crew stops treating the GNUstep port like a stepchild soon. Anyway, 
sometimes python doesn't cut it and rewriting critical parts in C isn't good 
enough always. With Gtk+ you are stuck then. There is no viable option IMHO. 
Porting to C is going to take too much time, and not much fun to begin with. 
C++ is just a horrible language. Well, there is Mono. In my case, 50 billion $ 
vs 500___ in the bank doesn't stand up too well in court. Please don't respond 
to that. Sorry for opening that can off worms. It's my conclusions. As Senior 
de Icaza has yet to give a remotely satisfactory response to that, nobody is 
going to make me think otherwise. Not that I expect anybody on this list would 
wan't to :)
Again sorry for getting there. As my excuse please consider the potential 
opportunity for GNUstep to give Gnome fugitives a new home.  I think with the 
Novell bullying, a Gnome a chism is imminent. The people in charge have long 
forgotten what that G in Gnome stands for and why Gnome came into exsistance, 
in the light of KDE, in the first place.

GNUstep has fumbled the ball in the past. Thanks to MacOS X there is another 
ball heading in your general direction. Please dear GNUstep community, don't 
drop the ball now!
GNUstep should be for everyone. Please don't try to keep reserving it for the 
truely enlightened!
They can have their GNUstep Live CD and Backbone. Who is trying to take that 
away?
But please don't try to force the whole widget on everyone  - that will fail.

And finally, it would be o so cool if GNUstep could beat the competition on the 
cairo/glitz thing to the punch.
I know, lack of developers. But a gnustep-backend should be a lot easier to do 
than hacking it into the thin layer over the Xlibs called Gtk+. Suddenly 
GNUstep wouldn't look like the Hurd of (GUI) frameworks anymore. Hey, that even 
makes sense when you read it literally, I noticed after writing it down :)

For the purists who kept reading until here, but know they really hate me and 
I'm a moron: don't mind me, if GNUstep doesn't deliver what I expect from it, 
I'll simply leave for good. Not going to fight on this battleground.
Been there - pointless - I'm doing it for fun - and that's not my idea of it.

Just another voice. You might consider it.

PS:
When responding to any of my numerous addendums (AKA "incoherent babble") it's 
prolly a good idea to open a new thread, right?






reply via email to

[Prev in Thread] Current Thread [Next in Thread]