[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Known MacOS programmer about WO/EOF
From: |
Riccardo |
Subject: |
Re: Known MacOS programmer about WO/EOF |
Date: |
Fri, 1 Apr 2005 13:16:13 +0200 |
Hello,
I give here some quick thoughts. Many of these themes have been
discussed in the past and opinion sometimes converge sometimes they
differ. Have a look ath the archives, Tabitha.
On Wednesday, March 30, 2005, at 02:30 AM, Tabitha McNerney wrote:
I wonder what it would take to turn around GNUstep's perception? To
take it from hobbyist perception to enterprise perception?
many would reply "new look" and I disagree. But I think many would agree
on "more applications".
But this is only a quick answer. I could detail it in:
* many bigger or smaller applications that are useful
- trying to port them from Cocoa, GTK and others. Why? because
people already know these applications, so this smooths a switch-over.
Of course often ported applications aren't good as freshly developed
one... especially if they are "alien"
- self-developed for GNUstep, but always port them to cocoa too.
Not only as a proof-of-concept but because they ease integration. Thin
about having an iBook and a non-apple workstation... the more you can
share, from addresses, bookmarks up to bigger application files, the
better you can live in a mixed environment.
* a change of attitude, of the way we are seen from the rest of the
world. if we are the "Think Different" part in the opensource world when
comparing to Gorm and KDE for example, we still need to provide a true
heritage of the NeXT and apple spirit not only in the GNUstep framework,
but in the main applications too. Quality, design, care for the detail,
portability...
- insist on portability, get a good *BSD exposure
- I think a windows port is a good thing (and the more integrated
and the less depending on MiniGW, extra GCC etc, the better. Ideally a
simple window installer for a single app...)
* some bug fixes in the backends and speedups here and there... OpenStep
run on a 486 and was fist like hell on a MicorSparc, 75Mhz. We currently
provide about the same look and set of features and are a lot slower. If
we have this "bare bones" look (call it whatever you wish) it should be
fast too. Ok, on a PIII, GNUstep IS fast if compared to sloppy gtk
stuff, but on slower machines it gets rapidly very slow, a strange thing
to explain. We even have some software that is the same (think of
TextEdit....) but runs slower...
* surely a lot of other things... like a good documentation of the
classes (like CocoaBrowser)
cheers,
Riccardo