[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality contr
From: |
Helge Hess |
Subject: |
Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control) |
Date: |
Thu, 23 Oct 2003 02:40:21 +0200 |
On Wednesday, October 22, 2003, at 08:06 AM, Philip Mötteli wrote:
I dare adding my 2c… For me, Gnustep are first of all the libraries.
Not a filemanager, nor a desktop, nor a system, nor development tools
(apart of the libraries, obviously). So I personally regret a lot,
that so much effort is spent in things, that are not especially
unique, but are already implemented, maintained and evolved by so many
others.
I mean something like GDL2 is unique and just "insanely great", isn't
it?
Well, I know I'm going to get killed, but no, it isn't. I really spend
at least about 5 years with working around GDL, thinking about OO
database technologies etc. In about the last two years I would say I
have come to the conclusion that if you want RDBMS, use RDBMS, if you
want an OO database, use one.
GDL (or EOF) is impressive for demonstrations purposes. You can rapidly
constructs very nice apps since you don't have to think about the
storage. But as soon as you have actual data in the database, the whole
idea behind EOF collapses and you start working "around" EOF to get the
work done. EOF just doesn't scale and for the simplistic problems it
solves, you are better of using simple OO database - like ZODB.
So JDBC sucks, but DBKit (if anybody still knows *that* ;-), was pretty
near at the thing which makes sense is still as usable as possible.
BTW: similiar issues arise in the context of WebObjects.
But ProjectCenter compared to KDevelop is just way, way behind and
unfortunately will probably never be even half the way of KDevelop.
Well, IMHO the same is true in the library area. OpenStep/Cocoa was
unique 10 years ago, it is still very much up to date, but it isn't
anymore significantly better than other solutions. That is also due to
new technologies, eg scripting, which make stuff which was previously
hard easy.
So in my eyes, the most important things would be:
1. Complete the already offered libraries. Especially in GDL2, where a
lot is missing.
Well, GDL alone is useless, you need some application around which
either has to be:
a) pure Web or middleware, aka gstep-web or NGObjWeb
b) AppKit, which more or less implies the AppKit desktop as well
c) some other, eg GToolKit
You need at least one of those options.
2. Debug them.
3. Improve the Windows port! (This is the second unique part of GS:
Cross platform)
Makes the above points even harder and isn't really important in
practice. Windows users use Windows software/technology (say ADO.NET,
C#, ...). Hugh.
4. Implement some plug-ins or other ways to make KDevelop or any other
well maintained development tool, compatibel with Gnustep.
Yup, would be cool.
5. Dito for a desktop environment. Preferably KDE.
Well, I would prefer GNOME. But this sounds like option c), GToolkit.
But I already know that this will definitely be no point the GNUstep
community can agree on ;-)
Or to attack the same problem from another point of view: Why might
somebody move to GS?
Well, the only real "customer" I can see are MacOSX developers which
need to do work on Linux/(BSD ;-). But even those are not interested in
ObjC WebObjecs/EOF anymore! (it's Java now, you know ;-)
I don't think many people do that because of a filemanager or a
desktop.
I would do if GNUstep would be a complete desktop. But I do not believe
in that the GNUstep community can reach that goal in the next ten years
- no offense intended.
Now of course, why I just don't do it myself? I would love to do it,
but how do I survive? I'm doing a PhD without a boursery or any
savings – and I just can't find a job in the GS world.
There are no GS jobs because GS provides no "solution". And you won't
find someone who jumps in "sponsoring" yet-another Linux desktop
project when we have two, pretty good (though not perfect!), ones,
available.
regards,
Helge
--
Reich wird man nicht durch das, was man verdient.
Reich wird man durch das, was man nicht ausgibt.
-- Henry Ford
- Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), (continued)
- Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), Rogelio M . Serrano Jr ., 2003/10/22
- Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), thisguyisi, 2003/10/22
- Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), Rogelio M . Serrano Jr ., 2003/10/22
- Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), Björn Giesler, 2003/10/22
- Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), Rogelio M . Serrano Jr ., 2003/10/22
- Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), Chad Hardin, 2003/10/22
- Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), Dennis Leeuw, 2003/10/22
- Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), Björn Giesler, 2003/10/22
- recruiting (was Re: GNUstep roadmap), Pete French, 2003/10/22
- Environment? (was: Re: GNUstep roadmap), Stefan Urbanek, 2003/10/22
- Re: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control),
Helge Hess <=
- Re[2]: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), Manuel Guesdon, 2003/10/23
- Re: Re[2]: GNUstep roadmap (was Re: [Suggestion] GNUstep-test for quality control), Helge Hess, 2003/10/23
- GDL2/EOF scaling (was: Re: GNUstep roadmap), David Ayers, 2003/10/28
- Re: GDL2/EOF scaling (was: Re: GNUstep roadmap), Helge Hess, 2003/10/28
- Re: GDL2/EOF scaling (was: Re: GNUstep roadmap), Patrick Coskren, 2003/10/28
- Re: GDL2/EOF scaling (was: Re: GNUstep roadmap), Helge Hess, 2003/10/28
- Re: GDL2/EOF scaling (was: Re: GNUstep roadmap), Patrick Coskren, 2003/10/28
- Re: GDL2/EOF scaling (was: Re: GNUstep roadmap), Helge Hess, 2003/10/28
- Re: GDL2/EOF scaling (was: Re: GNUstep roadmap), Patrick Coskren, 2003/10/28
- Re: GDL2/EOF scaling (was: Re: GNUstep roadmap), Helge Hess, 2003/10/28