Submitted by The new 0.4 version of Etoile had just been released. Etoile intends to be an innovative, GNUstep-based, user environment built from the ground up on highly modular and light components.
It is created with project and document orientation in mind, in order to allow users to create their own workflow by reshaping or recombining provided Services (aka Applications) and Components. 0.4 is a developer-targeted release on its way towards this goal. As a developer-focussed release, this predominantly consists of frameworks.
A few demonstration applications are also included.
are working in conjunction with Etoile?
I have the trunk and with Debian Sid it’s completely broken. Hell, GNUstep is definitely not in a workable state with Debian Sid, let alone Lenny.
ProjectCenter is behind, Gorm is outdated, LLVM is at 2.2 with only a tacit build of 2.3. With LLVM 2.4 now stable and still a major work in progress, I’m still trying to figure out the focus on LLVM first when GNUstep has been languishing and broken in focus for years.
I’ve read the “we are a frameworks” not a desktop environment rant for 5 years. Either this project grows, stays a niche with promise or dies.
Which one is it going to become?
It’s not going to compete against GNOME or KDE and neither of those two prevalent environment gives a rat’s ass about offering bindings for ObjC so I can’t forsee them jumping for joy over LLVM when it’s C++ status finally is usable to the level matching GCC.
I’m now trying to figure out how adding Smalltalk to Etoile is somehow going to improve the visibility and thus allow Cocoa developers become more interested in developing their applications [which helps drive interest] to GNUstep when ObjC2.0 and Cocoa first don’t even fully support the Openstep API specification, let alone anywhere near complete support for OS X 10.4 nevermind 10.5. With OS X 10.6 around the corner and a lack of interest in Linux with ObjC I see this remaining a very tiny niche, at best.
I’d like to see it grow, but I’m not seeing any sort of bridge work to add these powerful frameworks into GNOME or KDE, ala Mono did for GNOME–not that I’m saying Mono is a win/win seeing as people either love or hate it.
while i agree with you on the stake of GNUstep, this project itself is still only at a 0.4 state, and isnt moving quickly (lack of time for development and man power). So to think of a bridge to KDE and Gnome and all that jazz at this point to them makes little sense so much as getting there back end stable and feature complete. of course i cant speak for the devs personaly, i just know where my priorities are when coding for my projects.
I’m going to reply to your posting slightly out of order. See below:
“GNUstep has been languishing and broken in focus for years.”
“I’ve read the “we are a framework[]” not a desktop environment rant for 5 years. Either this project grows, stays a niche with promise or dies. ”
I’m assuming you’re referring to GNUstep here. if you are, then I’m not sure what you mean by a lack of focus.
If the developers and the project leader, myself, are saying that it’s a framework, then it’s a framework. Where’s the lack of focus if we’re focused on being what we say we are?
The very first thing I did when i took over the project in late December 2006 was to refocus the project as a set of frameworks and development environment only. Please see my blog, here:
http://heronsperch.blogspot.com/2006/12/plans-for-change.html
For some reason, people seem to want it to be an environment, I’m not sure why. I think the real problem might be that what we are doesn’t necessarily mesh with what many people think we should be.
There’s no reason for GNUstep itself to focus on being a full desktop environment, since, as you can see.. the Etoile team has done an excellent job of creating a desktop for GNUstep.
“It’s not going to compete against GNOME or KDE and neither of those two prevalent environment gives a rat’s ass about offering bindings for ObjC so I can’t forsee them jumping for joy over LLVM when it’s C++ status finally is usable to the level matching GCC.”
There are already ObjC bindings for GTK, please see here:
http://directory.fsf.org/project/gtk/
Also, KDevelop has support for ObjC.
Even if they didn’t, gcc’s ObjC++ provides a way to call C++ functions from Objective-C without the need for a bridge framework.
“I’m now trying to figure out how adding Smalltalk to Etoile is somehow going to improve the visibility and thus allow Cocoa developers become more interested in developing their applications [which helps drive interest] to GNUstep when ObjC2.0 and Cocoa first don’t even fully support the Openstep API specification, let alone anywhere near complete support for OS X 10.4 nevermind 10.5.”
I can’t speak to what the Etoile teams intentions are, but I don’t think that adding Smalltalk had anything to do with attracting Cocoa developers, at least that’s what I believe.
Also, some of the things you state are common misunderstandings regarding GNUstep. Generally, GNUstep follows Cocoa and I’m currently aware of no commonly used parts of the OpenStep specification that GNUstep does not implement. The only pieces missing are certain classes which are rarely used or those that were deprecated when OpenStep became Cocoa. GNUstep, in fact, currently implements a large portion of the Cocoa framework.
As for being 10.4 or 10.5 compatible, we do implement a lot of 10.4 functionality, but, alas, we’re a small team and we don’t have the capacity to keep up with every single release Apple makes the moment it comes out.
Also, to complicate matters there is currently no free/open source implementation of ObjC2.0 anywhere except LLVM. I’m not certain what the situation is with the patches that Apple has made on their gcc branch or even if they’re portable back to the main branch of gcc.
“I’d like to see it grow, but I’m not seeing any sort of bridge work to add these powerful frameworks into GNOME or KDE, ala Mono did for GNOME–not that I’m saying Mono is a win/win seeing as people either love or hate it.”
We are doing our best to make GNUstep applications interoperable with KDE and GNOME and things should continue to improve on this front. See my blog posting above for points on this as well.
As I said on the FLOSS Weekly podcast recently (#44 for those who are interested) GNUstep needs developers more than anything right now. If you’re interested in making GNUstep better, then please help out.
Sincerely,
Gregory Casamento
GNUstep Lead Developer
If I read your post correctly then someday in the future it would be possible to port Osirix to Linux.
That would be awesome because PowerMACS are far to expensive for us and iMACS have a 4Gb limit.
I can garantee you that a project like Osirix on Linux would have a big impact on developers for your framework.
Hello, Greg.
I’ve been following a few aspects of GNUStep Development from far away for a while, and it’s nice to see someone that’s immersed in this community so I can ask something:
How is the work going on porting WebKit into GnuStep (or any other good web rendering framework, actually)?
It’s actually the one thing that pushes me away from GnuStep: I know I don’t have the skills to port something like that, but if I could use it more often (and I’d need a good browser) I’d love to work on desktop applications using GnuStep, just because I enjoy ObjC
There have been some stalled efforts to port WebKit to GNUstep. It is a considerable, tedious undertaking.
That said, a functional Haiku port was created rather quickly, so it should also be possible for GNUstep in principle.
I would say the following Wishlist is an area which reveals how much GNUstep has to go to complete Openstep API compatibility, not to mention other wishes:
http://wiki.gnustep.org/index.php/Summer_Of_Code_Ideas
Thaks for the response, Greg. You’ve always been a class act.
The reason I brought this up was for visibility.
I used to get people pissed off at me when I worked at Apple and they could not give me one respectable answer, other than where is the business case/profits in it for us cocka on our PPC platform [still myopic as has been proven with the growth of skilled ObjC developers the growth of the developer community and future product development], and they’ve continued to act as if GNUstep is someone’s nice idea.
Part of my venting was actually towards GNOME and KDE. The GTK+ devs and the Qt devs could have done themselves a lot of legwork if they had worked well with you guys long ago. The work on WebKit would have gone much more rapidly, but then again I’m overly optimistic and altruistic that people will occasionally spread the wealth, so-to-speak, of knowledge when it’s a win/win all around.
Actually I believe it should be called ~A‰toil~A(c).
Edited 2008-11-18 23:25 UTC
Indeed, it is on ~A‰toil~A(c)-produced documents — the accents weren’t transmitted in this OSAlert story.
Edited 2008-11-19 06:08 UTC
Etoile is an awesome project. I don’t think it really matters what happens as long as they are able to get a few more applications written. At this point there’s an almost functional desktop in there. I would like to see an improved finder more along the lines of nautilus/Finder for mac. A completed Vespucci and Melodie would round out most of the desktop for me. I mainly use the web/chat/irc/mail/music. GNUMail.app is ready, TalkSoup is ready. Babbler/Gmplayer ports are progressing. The rest is coming along well. I’ll be interested in the 4.1 user release. It’s shaping up to be a very mac-alike desktop without the expensive price tag. I just checkout the code and build it myself. I’ve found binary distros are crap at keeping up with package versions. Unfortunately both gentoo and Sabayon appear to suffer from the same rot.
Edited 2008-11-19 00:51 UTC
This project is currently a moving target, so it is available through the GNUstep overlay of Gentoo, not in the main tree. When the project matures this will change.
You’ll find that the GNUstep packages in the main tree are fairly up-to-date.
one more thing. Do you guys dointerface usability testing? I mean stuff like how responsive the UI is? I’ve noticed when I try to highlight text gnustep programs will often sieze up for a few seconds. It gets really annoying and it’s one of those bugs that makes using programs like Gorm horrible. If you guys want bug reports I’m happy to provide them as well but there’s no debug output. If you can guide me to where I can report usability issues I’ll make a few reports for you.
On my Gentoo+GNUstep install I have no such issues, and this is the first I’ve heard of it. So it’s probably something specific to your system. There is one possibly related bug I could find that is Etoile-specific:
https://gna.org/bugs/?8752
Running a particular GNUstep exhibiting this problem under gdb may be useful; tips here:
http://www.gnustep.org/resources/documentation/Developer/Base/Gener…
If you are running Gentoo, set the “debug” USE flag first.
Also, someone on the GNUstep mailing list or IRC channel (#gnustep @ freenode) may have some insight.
Please report any bugs you find at:
http://bugs.gnustep.org
You will need to create a login id (helps to reduce spam from spambots).
GC
Yes, we do test things and usually selection is quite responsive.
If you don’t start up the pasteboard server, the first time you do a selection there will be a delay since the system starts one up for you automatically, but subsequent selections are just fine (or should be).
MidnightBSD???
I fail to see the point of this project other than to slap on a plug-ugly (aqua-inspired – what else)theme on a still great desktop.
The original NEXT/OPEN/GNUstep effort is still the benchmark in desktop design with its utilitarian and unadultered approach to usability, not least having been created by some of the finest minds in the industry. And to have a bunch of amateurs undermime it with such a derivitive eye-sore is just sad
Etoile is _far_ from being a theme put on a “still great desktop” (by this I guess you mean GNUstep, and/or NeXTSTEP).
1) GNUstep is not a desktop, it’s a programming framework. There are GNUstep apps (check the gap project, gworkspace) that together can be used to form a desktop. Etoile is a desktop, and is composed of both programming frameworks to simplify writing apps and of desktop applications.
2) Etoile does NOT want to create a NeXTSTEP-like desktop, nor a MacOS-like desktop. Although we obviously are inspired by both.
3) This 0.4 release is aimed to developers. Wait for the 0.5 for a user-oriented release. What 0.4 brings is new programming tools and frameworks. Notably, as the music player Melodie shows, you can write a complex application in Smalltalk + Etoile frameworks, in a very simple manner.
4) There is far more to a desktop than the theme it’s using.
Well, isn’t your comment a bit harsh? You are not forced to theme your desktop and you are not forced to use Etoile, if you wish, you can just use GNUstep and install the applications you like. Etoile is a projects which builds on GNUstep and which adds new powerful frameworks and applications. The choice is yours.
For me, the stunning about ~Af^A‰toil~Af^A(c) & GNUstep is, they can almost compete with the big KDE and Gnome while having only about a dozen active developers. This shows how great the initial (OpenStep-) design is and how well it was implemented by GNUsteppers. Bugs appear, but unlike in the Linux-kernel-like world, they’re almost always fixed within hours. Major rewrites rarely occur, as the fundamental design still proves to be well thought.
On the not so shiny side, neither GNUstep nor ~Af^A‰toil~Af^A(c) obviously see the high value of packaging for the major distributions, especially Ubuntu. Packages exist, but they’re dated. I’d love to see those two to provide their own package repository, so interested users can enjoy an up-to-date experience.
no matter which (etch, lenny or sid). have you actually tried the http://livecd.gnustep.org/ ?
Building Etoile with Debian Sid is broken, period.
Edited 2008-11-21 16:02 UTC