Re: [Fwd: Re: GNUstep and Linux Fund]

[Hans Baier]

> GNUstep has lacked and continues to lack good PR, consistent development of
> applications. But for the core itself I think the following points would
> help us a lot (in no particular order)

I think the design Website is - in its current state - a bit outdated:
It still appeals to the NeXT era kind of look and feel, but now
the most exciting stuff seems to be going on in OS X.

Also navigating the website, it could use more simplicity and structure.

The following sites have excellent design:
http://cocoadevcentral.com/ (try one of the tutorials)
http://cocoawithlove.com/2009/11/creating-iphone-and-mac-icons-using.html
and - last but not least the apple website.

But more important the Top most two reasons which prevent me from using
GNUstep for more than hobby hacking are those:

*. Look and feel. Althought Qt is an excellent development
environment, I ditched it together
   with KDE, because it is so cluttered and grossly inconsistent in
its look and feel. Icons
   play an extremely important role there, and the KDE icons look childish.
   But consistency plays the most important role in look at feel.
   Here, the default GNUstep theme, though outdated, is pretty good.
   The master in this category here is OS X, but I dumped it for the
following reaon.

*. Usability: Having a fast, simple way of doing everyday work. For me,
   keyboard support is the most important, since my day job is that of a
   software developer. This is why I dumped OS X and ended up
installing Debian on
   my $4000k Mac in the past: The weird keyboard layout, mostly missing keyboard
   shortcuts for even the essential things (like maximizing /
minimizing windows),
   no keyboard configurability made OS X a nice toy, but nothing
serious I could use
   for day-to-day work. It seems to have improved since 10.3, but
   for today I am put off by Apples restrictive licensing, and the
fact that I was forced
   to spend EUR 150 for a new OS version, just to be able to run Java 1.5.

   GNOME is the desktop of choice for me ATM, because it has modern look and
   feel, is very consistent, and has a keyboard way of doing
   things (like incremental search in the Alt-F2-Dialog and lists,
TAB-completion
   in the File Dialog, configurable Keyboard shortcuts, being able to
use the cursor arrows
   for GUI element focus navigation), without enforcing it on the user.

Look and feel as well as keyboard usability are my main issues with GNUstep ATM.
Hopefully with the Gtk+-Theme I am developing ATM, this will be fixed
soon (If I have time).
Interestingly, since the Gtk+-Theme goes much deeper than alternate pixmaps,
it shows how GSTheme could be extended in the future (especially where I have
to use Categories to override widget drawing functions).

Apart from that, one thing about GNUstep annoys me very much:
The absence of context menus. I don't know currently how OS X handles that,
but often I would find myself right-clicking on something, but the
application menu pops up.

Lastly about the GNUstep community: My main problem when I started contributing
(writing pbxbuild in two weeks of vacation), was that I did get no
feedback. I therefore felt that
no one really was using it / or cared about it and that extinguished
my motivation at that point
and my interest therefore changed to other things.
The GNUstep community, tough, is pretty friendly and tolerant to
mistakes, and today
I still remember Fred Kiefers kind reactions when I messed up things
in the cairo backend or broke something
in GNUstep gui.

I spent most of the my open source time in Ardour during the last two
years. It has a very lively community (especially in IRC).
There are lots of users coming in on IRC giving lots of feedback.
That kept me motivated for a long time. (But I currently don't use it much...)

Lastly about funding: I was very glad to read the news from the Linux
Fund, and since I also watched the project since 1996,
I am glad someone pays attention to it that way. I didn't quite
understand whether the $1000-$5000 figure was per month
or one-time sums from the mails, and I will be glad to see it used to
help GNUstep advance. Cheers to the linux fund for that.

Kind regards,
Hans Baier