Home > X11 > Linux Desktop Developers Find Common Ground Linux Desktop Developers Find Common Ground Eugenia Loli 2005-12-08 X11 71 Comments A developer meeting strengthens a Linux desktop alliance that could lead to better leverage with hardware vendors and more cross-distribution efforts such as the newly created Portland Project. About The Author Eugenia Loli Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker. Follow me on Twitter @EugeniaLoli 71 Comments 2005-12-08 10:04 pm I got really motivated when I read this. Hopefully the Portland Project will be successfull. To sum things up things are looking good in the linux world today :> 2005-12-08 10:24 pm bytecoder If you think so. I’m rather depressed about the personal computing industry as a whole, though. It seems to be reverting into a dark age of sorts. 2005-12-08 10:30 pm how many desktop projects does there need to be? 2005-12-08 11:37 pm enloop Don;t fret. The project that develops a desktop that convinces people to buy Linux instead of Windows or Apple will be the last Linux desktop. 2005-12-09 3:00 am To add to enloop’s comment with a paraphrase that may sound familiar: In the end, there might be only one… 2005-12-09 10:40 pm Sphinx Would that be based on the, “Highlander Theory”? 2005-12-09 10:45 pm Sphinx Sorry, what I mean to say is that by open source’s very definition there should always be several to choose from. 2005-12-09 9:37 am raboof how many desktop projects does there need to be? Luckily, from the article: Portland’s goal is to make it possible for desktop software developers to write to the Linux desktop without having to worry about whether the distribution or user is using KDE, GNOME or a more obscure Linux desktop. Since the goals of Portland are very much in line with the mission of FreeDesktop.org, it was agreed that FreeDesktop.org was the proper place. It doesn’t become very clear to me how exactly they’re going to reach their goals, but they sure are nice goals 2005-12-09 12:52 pm Many! New ideas need testing and development. But should the distros focus on all of them? That might not be so. 2005-12-09 3:53 pm John Nilsson As many as there’s a demand for, and a few more. 2005-12-08 10:32 pm JCooper As DE’s become less disparate in terms of features and core libraries upon which they are based, linux+DE of choice will see more adoption by, and backing from, larger corporates. This, along with the recent news of GParts (permitting KDE KParts to be embedded in GTK apps and vice versa) is a good general movement towards unified systems. These appear much more professional to corporates and will encourage investment. 2005-12-09 9:53 pm g2devi In case anyone’s interested, here’s quite a bit more information about GParts: http://pvanhoof.be/blog/index.php/2005/12/08/78-gparts-and-a-shared… The basics appear to be simple. Gtk and Qt will share a common main loop so that they can be in the same process space. You’ll be able to embed either GParts or KParts into containter written for GTK+ or Qt. GParts should be trivially compatible with Bonobo, so in theory, you could create a KPart and have it hosted in a Bonobo object without too much (any?) work. Basically, the walls between Gtk+ and Qt can finally start breaking down. 2005-12-08 11:06 pm segedunum I sincerely hope something might come out of this, but I seriously doubt it. It looks like yet another wishy-washy, touchy-feely Gnome and KDE heart to heart that will deliver nothing that some people can’t understand will never work. However, there seems to be some confusion in here: In addition, developers realized that by working together they will have more of a consolidated voice in talking with hardware vendors Getting the needed APIs and ABIs (application binary interfaces) for many devices has long been a sticking point for Linux attempts to equal Windows’ support for many desktop devices. Is this merely about hardware? Hardware integration is already being done where KDE and Gnome use common software, and I don’t see a problem there. All Portland seems to be, from the description, is a layer sitting on KDE and Gnome that will supposedly allow ISVs to run things unchanged. Big wow. It’s another ill-advised attempt at integration where integration just isn’t possible or just creates more work than is actually required. Effectively, to make this work they’ll be trying to build a layer on top of each desktop that will effectively become a desktop itself. All this does is create about five times the huge amount of work that would otherwise have to be done, which people barely have the time and resources to complete as it is. We’ll probably have to accept that there will be two desktops, and that’s the way it is. We’ll see some collaboration on common areas of interest, but the notion we’re going to get some dog-eared Portland layer supposedly providing some bogus compatibility leaves me with my head in my hands shaking my head. Also, I have absolutely zero faith in the LSB. In terms of providing binary compatibility to ISVs it is next to useless. It’s a decent gravy train for some people and a nice logo for distributors to have on their boxes, but unfortunately, nothing else. If these people think that is what is required to get a working desktop to people then they’re well, well wide of the target. 2005-12-08 11:21 pm butters I agree somewhat with your sentiment, but I interpreted this news a little differently. Instead of a desktop layer sitting above KDE/GNOME, this seems to me to be a compatibility layer within KDE/GNOME to work with distribution-specific things. For example, yesterday’s link about BUM, the new runlevel manager for Ubuntu/Debian. I was tempted to write a comment about how this is a good idea, too bad it will only work for Debian-style SysV init systems. What if KDE and GNOME both had a standard interface to manipulate runlevels? The distributor would have a well-defined set of methods to implement for their specific setup, and no more questions about the matter on the forum sites. At least that’s what I hope this means. If it’s some garbage about needing both Qt/kdelibs and GTK+/libgnome on every linux desktop, then I agree with you completely. 2005-12-09 10:39 am segedunum Instead of a desktop layer sitting above KDE/GNOME, this seems to me to be a compatibility layer within KDE/GNOME to work with distribution-specific things. Same difference as far as I’m concerned. You’re still going to have to maintain that layer, all the distribution and desktop specific changes and complexity therein. At least that’s what I hope this means. If it’s some garbage about needing both Qt/kdelibs and GTK+/libgnome on every linux desktop, then I agree with you completely. Well, if it doesn’t solve that issue somewhat for ISVs then it’s useless. However, that is something that is really impossible to solve. If people choose GTK then they’ve chosen GTK, and likewise with Qt and KDE. Integration between the different toolkits and desktops is possible, as there is with the qt-gtk engine etc., so I do have a bit of hope there. However, you’re still going to need Qt, GTK, KDE and the relevant Gnome libs installed to run various applications that need them (or they will have to distribute them), whatever the desktop, so I just don’t see what can be done there. Not that that really matters. You always need certain libraries installed whatever piece of software you use, and some libraries on your system will even do the same things, but differently. Most of the arguments over this generally tend to be religious ones i.e. “We don’t want xxx desktop libraries on our system”, “It wastes disk space”, “It’s unnecessary bloat”. 2005-12-08 11:36 pm Novell Redhat and Ubuntu use Gnome as default desktop it’s time to accept that and rid off one time for ever KDE. sorry guys qt is a commercial library… 2005-12-08 11:48 pm LinuxRocks Trolltech does offer a commercial offering, but they also have a contract with KDE to forever offer a “Free” version of their software. Even if they (Trolltech) is bought by another company. Besides, KDE is MUCH more popular that Gnome; even though those distros use Gnome by default, most users (according to surveys) switch right away to KDE… 2005-12-09 12:17 am “Besides, KDE is MUCH more popular that Gnome; even though those distros use Gnome by default, most users (according to surveys) switch right away to KDE…” all the usability studies so far have concluded the exact opposite. 2005-12-09 1:16 am rayiner Usability studies don’t attempt to find which desktop is more popular, only which desktop is more usable. 2005-12-09 2:45 am “Usability studies don’t attempt to find which desktop is more popular, only which desktop is more usable.” the userbility studies attempt to find the DE that people prefer using….and gnome comes out top in almost every case. 2005-12-09 6:18 am Tuishimi And this little thread is exactly why Portland cannot win unless it remains somehow generic and simply seeks to establish a canonical set of tools, and a cononical naming convention for access to those tools via menu in an attempt to minimize confusion when zipping from one Linux box to another… 2005-12-09 9:17 am “the userbility studies attempt to find the DE that people prefer using….and gnome comes out top in almost every case.” I think you’re quite wrong here A usability says nothing about what people prefer, it lets people know which DE is more usable, from a certain point of view. A test could for example reveal that MacOS X is more usable than Microsoft Windows, yet not everyone seems to prefer MacOS X. So dispite that GNOME might be more usable than KDE, I would not be suprised if the majority indeed DID prefer KDE. 2005-12-09 4:09 am all the usability studies so far have concluded the exact opposite Which usability studies? Any links to a useful comparison? All professional usability studies I’ve seen so far concentrated on one of the desktops. The only comparisons I’ve seen were those “usability expert”-fanboys that seem to be more prevalent within the Gnome community but plague both projects. 2005-12-09 10:45 am segedunum all the usability studies so far have concluded the exact opposite. Usability studies (depending on who has done them of course) count for absolutely nothing and certainly don’t consider popularity or usage. Not that this whole pointless thread will get anybody anywhere, but it is the attitude that some people at Novell seem to have taken and will yield them exactly the same results and progress as before. Nothing. 2005-12-09 7:58 pm unoengborg “Besides, KDE is MUCH more popular that Gnome; even though those distros use Gnome by default, most users (according to surveys) switch right away to KDE…” Most survey are self selecting. KDE is a much better fit for the advanced user than it is for beginners or office workers with no particular interest in desktop environments or operating systems. So it is no surprise that KDE looks more popular. It is also no surprise that companies that try to market desktops for corporat use usually have Gnome as their default. It is easier to use for the newbie, and it is always better to have a low threashold of knowledge. You sell more that way. Expert users, or users that like to tinker with their system have no problem switching from Gnome to KDE. 2005-12-09 1:06 am elsewhere sorry guys qt is a commercial library… *sigh* And GTK is LGPL. Which means that any of those big commercial developers we’re hoping to attract to make the desktop relevant cannot restrict reverse-engineering of their code under LGPL. Enforceability of EULA’s may be a whole seperate debate, but if you think the commercial developers are going to easily convince their legal departments to remove that long-standing restriction from their EULA’s just so they can save money on what amounts to a very incremental cost in a product development cycle, you are sadly mistaken. Qt is GPL/proprietary. GTK is LGPL. Neither license is perfect, and both have their intended developer markets. But I suppose there’s no point in debating this rationally, right? 2005-12-09 9:45 am raboof GTK is LGPL. Which means that any of those big commercial developers we’re hoping to attract to make the desktop relevant cannot restrict reverse-engineering of their code under LGPL. What exactly are you referring to? (not to be starting a thread full of trolls, i’m really wondering…) At least as long as they dynamically link, applications are never harmed by the fact that they link to LGPL libraries, right? 2005-12-09 6:32 pm elsewhere At least as long as they dynamically link, applications are never harmed by the fact that they link to LGPL libraries, right? I’ll admit I’m not a lawyer, so my interpretation is based on what I’ve read from various sources. If the applications dynamically link, assuming the libraries are pre-installed on the user’s system (which in fairness is a goal of initiatives like the LSB) then yes, the application is clean. But where it gets convoluted is if the application is statically linked, modifies the library or is linked to a work that is linked. At the end of the day, commercial developers would have to do due diligence to ensure that their applications are in compliance, not that it can’t be done but their legal departments will be concerned. My real point was that Qt shouldn’t be “written off” just because it’s not “free”, that license fee buys not only the right to use the framework, but the right to license their product as they see fit. The cost is really marginal for developers in the context of application development. I don’t think there’s anything inherently wrong with LGPL, any more than I think there’s anything wrong with Qt’s dual-licensing. I just get frustrated by the people that say LGPL is the only way because it’s “free”, it’s not necessarily “no strings-attached” and it’s a very narrow perspective to assume commercial developers will flock to it because it’s no-cost. 2005-12-10 9:07 am raboof assuming the libraries are pre-installed on the user’s system There’s no harm in bundling LGPL’d libraries with proprietary software modifies the library Indeed when the lib is modified those changes need to be ‘given back’ – reasonable, I think. At the end of the day, commercial developers would have to do due diligence to ensure that their applications are in compliance, not that it can’t be done but their legal departments will be concerned. True, though that’s true for any license. Given that the LGPL is a relatively widely used (and thus well-known) license that might make this easier rather than harder. My real point was that Qt shouldn’t be “written off” just because it’s not “free” To that I, of course, whole-heartedly agree. I just get frustrated by the people that say LGPL is the only way because it’s “free”, it’s not necessarily “no strings-attached” and it’s a very narrow perspective to assume commercial developers will flock to it because it’s no-cost. Indeed people should realize what exactly they’re choosing for when going with any license. I hope posts like yours and mine help people making a more informed choice. 2005-12-09 2:43 pm “And GTK is LGPL. Which means that any of those big commercial developers we’re hoping to attract to make the desktop relevant cannot restrict reverse-engineering of their code under LGPL” glibc is LGPL too, right ? 2005-12-10 9:14 am raboof if you think the commercial developers are going to easily convince their legal departments to remove that long-standing restriction from their EULA’s just so they can save money on what amounts to a very incremental cost in a product development cycle, you are sadly mistaken. I understand removing a restriction from the EULA is often unacceptable, but I don’t see how using an LGPL library would require that. Could you elaborate on that for a bit? 2005-12-09 1:43 am “sorry guys qt is a commercial library…” Don’t act so communistic… So what if it is a comercial libary… You don’t what commercial programs in linux… commercial games… You want everything free… so much that even the free stuff must be perfect for you, if not you damn it? I people can’t make money with some programs of tools then linux could be dead… And we don’t want that… Besides… you can use whatever you seem fit… don’t be like a vegetarian who wants to make everybody a vegetarian… Vegetables suffer to you know! -iMoron 2005-12-09 4:25 am dylansmrjones Well, the GPL is a commercial license. Don’t confuse proprietary with commercial. Freeware is proprietary yet not commercial. GPL is libre software, yet it’s very often a commercial product. Sooner or later people will learn. KDE seems to be more popular in Europe with Gnome more popular in USA, but personally I still have to meet KDE users in Denmark. So far all of them have been Gnome users, including all the linux users in my Computer Science class (including teachers). However, many polls seem to indicate that KDE is more widespread than Gnome, but there could be many reasons for the results of those polls. Gnome in general seems to have a wider range of applications, incl. games. This could indicate a more widespread use of Gnome, but then again. There could be several other reasons. 2005-12-09 10:53 am dark child I wish people would stop this useless bickering about which is better/popular KDE or GNOME. Such lame arguements portray the opensource community as fragmented and make it harder to convince people to make the switch from the poprietary world because we are just seen as a bunch of zealots who can’t agree about anything. Opensource is all about choice and the presence of KDE, GNOME and other DEs or WMs provides healthy competition for developers of these projects and the community as a whole. GNOME is popular in the states and KDE is popular elsewhere, does it really matter? The KDE vs GNOME debate is really tired and people keep repeating the same old arguements. Lets be honest and admit that there will never be a single Linux/Unix desktop as long as there are people using opensource because we are so diverse and have different preferences. When the devs and other interested parties work together, then this is a good thing because it shows that people from various opensource projects can work together on common ground even though their ultimate objectives may be different. 2005-12-09 12:01 pm dylansmrjones Well, I agree. One unifying desktop is as close as one unifying car model 2005-12-09 1:31 pm segedunum Well, I agree. One unifying desktop is as close as one unifying car model Imagine if that unifying car model was a Ford Edsel :-). 2005-12-09 1:32 pm No it isn’t! A desktop is something on which you run PROGRAMS. Having a ‘standard’ desktop is like having a standard road, not a standard car. 2005-12-09 2:33 pm dylansmrjones Well, I don’t know about how the situation is where you live, but in Denmark we have all kind of roads, public and private, good shape and bad shape, concrete, asphalt and dirt tracks, and stone roads… 2005-12-08 11:57 pm moleskine I suspect the Linux desktop will sort itself out without manual intervention from a committee of worthies. It will only require one killer app or one killer software suite from the likes of Abobe to come out on either gtk or QT and a de facto standard will have been set. Actually, you could argue that one is already being set since the big distros seem to be moving inexorably towards Gnome and so before long Gnome will gain a majority mindshare. None of this invalidates KDE which is a darn fine piece of work and no doubt will continue to get even better for years to come (I hope). The LSB would go further to my mind if it did something like require desktop Linux distros to deliver a list of 100 apps guaranteed to run without crashing or avoidable, distro-specific problems. Linux needs some ruthless Q&A if the reality of trying to use it on the desktop is to get anywhere near the hype. It needs more bureaucrats like a hole in the head. 2005-12-09 11:29 am segedunum I suspect the Linux desktop will sort itself out without manual intervention from a committee of worthies. Possibly, but we’re the best part of a decade away from anything like that happening. There’s a huge amount of issues to get over. I see KDE having better technology to get over those issues, but Gnome, KDE and a Linux-based OS still have all the same problems which need to be solved to get widespread and critical mass usage of desktop Linux. Actually, you could argue that one is already being set since the big distros seem to be moving inexorably towards Gnome and so before long Gnome will gain a majority mindshare. That’s not happening I’m afraid, as much as many people jump up and down on the spot to try to convince you. The only real Gnome based distributions out there are Ubuntu and Fedora, desktop sales do not make up a small splash in Red Hat’s revenue, and some rather sad people at Novell are still trying to convince everyone that the NLD is the corporate desktop and it’s selling like hotcakes. The most popular desktop distribution at Novell is still Suse Linux Professional, and openSUSE (what is it with that case?), and the popularity of both of those still very much depends on Suse’s traditional treatment of KDE. Despite many people trying to paint a different picture Suse Linux (and Novell) is still a very, very long way from being a Gnome based distributor. Considering that Novell are becoming less relevant by the day (and if you want Gnome just go Ubuntu) I find that all a very mute point. The LSB would go further to my mind if it did something like require desktop Linux distros to deliver a list of 100 apps guaranteed to run without crashing or avoidable, distro-specific problems. You’re describing some sort of Microsoft certification bullshit there. I’d prefer it if things just ran. There’s no reason why they shouldn’t. Certainly, a list of apps is of no use to an ISV because it tells them nothing of the quality of the environment they’ll be using to run and test their apps. All it tells them is that if they fit a narrow band of criteria then they may get the applications to run. Linux needs some ruthless Q&A if the reality of trying to use it on the desktop is to get anywhere near the hype. It needs more bureaucrats like a hole in the head. With QA and the LSB comes yet more bureaucrats. 2005-12-09 12:12 am ma_d I think you’re right as far as the information age coming to a close (ie, being bought by media companies). But I think PC software is better today than ever by far. 2005-12-09 12:13 am ma_d As many as people choose to work on. How many news sites do there need to be? How many forums do we need? How man car companies do you need to have? How many people should be allowed to own restaurants.. Efficiency has never been capitalism’s finest trait, succeeding in the face of inefficiency might be though. 2005-12-09 12:17 am ma_d Acrobat 7 seems to use gtk . Gtk is more popular for most non-associated builders (ie those not working with Gnome or KDE) because it’s independent of Gnome (technically speaking) and KDE. Writing QT apps is less common, I think, because you have to write your own theming code (at least enough to load a theme and provide an interface for your users to choose a theme). The nice thing, for kde users, is that gtk software loads as well in kde as it does in Gnome. And even Gnome apps only have a small load performance cost for being a Gnome app under KDE: Where a kde app when launched for the first time on a Gnome system can take a good 30 seconds. So, using KDE apps for gnome users isn’t so good, but using gnome apps for kde users isn’t a big deal. 2005-12-09 2:11 am leos Acrobat 7 seems to use gtk . Yes. And it is disgustingly slow and doesn’t integrate with Gnome or KDE. I’m not saying that’s GTK’s fault, but it really shows that we’ve now reached a point where the open source apps of all kinds are just plain superior to those proprietary apps that used to be held up as the holy grail.. 2005-12-09 4:27 am dylansmrjones Well, Adobe Reader is slow on all platforms. On Windows I use FoxItReader (damn small and fast and all you could ever wish for in a PDF reader), and on linux I can recommend xpdf if speed is the main issue. Otherwise you could use gPDF or a third option. 2005-12-09 9:20 am Acrobat 7 seems to use gtk . Yes. And it is disgustingly slow So just use Evince, I prefer that to *any* version of Acrobat Reader ever released by Adobe, I hope it gets ported to win32 one day … 2005-12-09 3:21 pm anda_skoa I think, because you have to write your own theming code (at least enough to load a theme and provide an interface for your users to choose a theme) No, Qt can load and handle styles by itself. Unless an application specifically removes it, there is a commandline switch to specify an override style. If the application does not choose a style in its code or no override is provided on commandline, it will use the one that is configured globally for all Qt applications (see qtconfig) 2005-12-09 1:36 am Just in, Gnome-lusers have no life, but are reported sitting in their parents basement, compulsory fireing off flamebait into every thread remotely touching the subject of KDE/GNOME. How does the saying go? “With such friends you don’t need enemies?” 2005-12-09 1:46 am hobgoblin like they are going for a lowest common denominator lib of some sort. ie, i write against this lib and it will work under both QT/kde and GTK/gnome. complete with copy&paste, tray and other elements… ok so most of this have been coverd allready, but its up to the apps creator to implement the standards set. but with a common lib that one can just code against and get all the features one wants, one can realy speed up development of the app itself. 2005-12-09 2:21 am I think Linus’s joining the OSDL was the beginning of solidarity in the Linux world. The OSDL makes a great collaboration and focal point for Linux developers. 2005-12-09 3:04 am aseigo > Is this merely about hardware? Hardware integration > is already being done where KDE and Gnome use > common software, and I don’t see a problem there. it’s not about integrating hardware into the desktops (though that’s part of the equation, and you’re right that we’re getting better there rapidly). it’s about getting proper drivers from IHVs in the first place. it’s about getting decent APIs for classes of hardware. we have a long ways to go when it comes to printers, video cards and what not. > All Portland seems to be, from the description, is > a layer sitting on KDE and Gnome that will > supposedly allow ISVs to run things unchanged. Big > wow. It’s another ill-advised attempt at > integration where integration just isn’t possible > or just creates more work than is actually > required. actually, the integration is possible because we’ve got some proof of concept stuff kicking around. and we’ll have more as Portland continues. as for the amount of work necessary, it won’t require writing the equiv of another desktop. it’s pretty simple bindings to a set of common desktop services. there are a couple dozen things that ISVs really need access to, and the short list is just a subset of those things. martin konold has been working on this for a while and has determined that with a very manageable expenditure of resources it can be done. and then ISVs don’t have to worry that their toolkit selection means that they’ll look like crap or not on any given desktop. and then users won’t have to worry about whether or not FooApp is “written right” and gives them the proper file dialog or will launch the correct app for a given mime type. 2005-12-09 9:53 am Ookaze it’s about getting decent APIs for classes of hardware. we have a long ways to go when it comes to printers, video cards and what not. I don’t understand. Don’t we have some good API external to both desktops, but at the same time managed by both ? Like CUPS, DRI, HAL, DBUS, … it’s pretty simple bindings to a set of common desktop services. there are a couple dozen things that ISVs really need access to, and the short list is just a subset of those things and then users won’t have to worry about whether or not FooApp is “written right” and gives them the proper file dialog or will launch the correct app for a given mime type Well now I think I see what it is about, but as for the GUI, I think we’re talking about a layer on top of gnome’s libglade XML format, and the equivalent for KDE ? 2005-12-09 11:43 am segedunum it’s about getting proper drivers from IHVs in the first place. He, he, he. Well, the OSDL needs to encourage and tell IHVs how to do it. Most of the time all of this binary driver rubbish comes from the fact that IHVs are simply embarrassed about the state of their drivers. The hard line of many kernel developers on the licensing of drivers, binary compatibility layers and how to get involved in kernel development are well documented. And still hardware companies out there are coming up with proposals for illegal binary compatibility layers that will just be ignored. it’s about getting decent APIs for classes of hardware. we have a long ways to go when it comes to printers, video cards and what not. Well, we did the current state of video cards quite recently so we’ll leave that. All of it is pretty vendor specific right now as well, and there’s certainly nothing in the way of a common set of interfaces. If that works out it will certainly help. martin konold has been working on this for a while and has determined that with a very manageable expenditure of resources it can be done. Hmmmm. Do ISVs import Portland’s libraries, or does this just happen automatically? and then users won’t have to worry about whether or not FooApp is “written right” and gives them the proper file dialog or will launch the correct app for a given mime type. To be honest, I though this stuff was already happening. There does seem to be some basic stuff that can be done there right off the bat without too much effort though, but I just wonder what the limit is to all this. You can integrate > 90% of applications tomorrow, simply because most of them have very little low level integration with a particular desktop environment. I’m giving this a pretty cautious welcome as long as people don’t get carried away and try to integrate stuff that just can’t be done. 2005-12-09 4:48 am According to John Cherry, OSDL’s Initiative manager for the Linux desktop, people finally understood that “we’ve got to stop acting like a bunch of disparate organizations and work together on the Linux desktop for it to be successful.” But what about choiiiicceeeee that the fanboys are so fond of screaming? Getting the needed APIs and ABIs (application binary interfaces) for many devices has long been a sticking point for Linux attempts to equal Windows’ support for many desktop devices. Sorry, but that’s out of the hands of the desktop developers. Better go talk to the kernel guys who have always said no to a stable ABI. In other words, Portland’s goal is to make it possible for desktop software developers to write to the Linux desktop without having to worry about whether the distribution or user is using KDE, GNOME or a more obscure Linux desktop. Great, should have happened 6 years ago. The desktop developers plan to get together again in May of 2006 to continue with their work of finding common ground for the best possible Linux desktop That’s the author’s commentary. But I guess even core KDE and Gnome developers have finally come to the conclusion that all this duplication of effort has hurt desktop linux. At this point it’s a little late. Mac will be out on Intel chips soon, with most likely cheaper prices and good virtualization speeds. Open source is its own worst enemy when it comes to getting anything into the mainstream. Some people don’t care about mainstream, others obviously do. 2005-12-09 10:45 am Ookaze But what about choiiiicceeeee that the fanboys are so fond of screaming? That the desktop work together does not eliminate choice. What I understood is that you still have the choice of desktop, but ISV apps can use a generic layer that integrates with your current desktop, be it Gnome, KDE or another. And next time, try not to call people fanboys just because they have a different opinion than yours. Getting the needed APIs and ABIs (application binary interfaces) for many devices has long been a sticking point for Linux attempts to equal Windows’ support for many desktop devices I alwasys hear this, but IMHO both efforts are not in the same league. Windows only have to support devices for itself. Talking of “Linux attempts” is reducing what the desktops are doing ion Linux. They actually try to unify hardware on several architectures and several OS, not only Linux, but also Windows, BSDs, Mac OS, embedded, … API and ABI are there, but time is needed for some to go above the rest. I think we can safely say that CUPS is ahead for printers, gphoto for cameras, SANE for scanners, V4L2, … Better go talk to the kernel guys who have always said no to a stable ABI I disagree. They have already stated very strong reasons for what they are doing. Proponents of the opposite have not given any strong reason. And the principal problem is with 3D support in graphic cards. Even with V4L, vendors would not support webcams on Linux, so kernel stable ABI looks like a straw man to me. Great, should have happened 6 years ago Come on, 6 years ago, there was no Gnome 2, so it would have needed a big overhaul, and could have slowed the desktops tremendously. Anyway, the fundations were not there for such a project 6 years ago. But I guess even core KDE and Gnome developers have finally come to the conclusion that all this duplication of effort has hurt desktop linux And I doubt it. I rather think the competition has improved both desktops. Given the very different philosophy of both desktops, I doubt there is so much duplication of effort. Perhaps you’re talking about apps ? I still wonder why the fact that there are several programs doing the same thing is seen like a weakness on Linux, and like a force on Windows. Especially when I see lots of comments of fans of softwares, saying they prefer one or another, given the fact that they do the same thing, but differently. At this point it’s a little late. Mac will be out on Intel chips soon, with most likely cheaper prices and good virtualization speeds I think you missed the point of FOSS software and desktops … Open source is its own worst enemy when it comes to getting anything into the mainstream That’s their problem. It all started with Free Software. That Open Source is flawed would not surprise me, RMS warned them from the start. Some people don’t care about mainstream, others obviously do The only reason why I care about going mainstream, is that when (if) we represent 10+% of users, driver manufacturers will have to actually make drivers for us. Otherwise, I would actually not care. 2005-12-09 5:36 am cendrizzi with linux developed disperate for so long this is a huge untertaking. LSB so far hasn’t been everything some of us had hoped but it’s slowly coming together(except notable distros like ubuntu aren’t even going to try). I’m very curious to see what comes of this project, I don’t really understand what they are attempting to do besides maybe just push the freedesktop.org project harder. While toolkits will never merge it will be really nice seeing things with the two desktop environments merge. I think the nice thing about this meeting is it got some commercial vendors like Adobe involved too. I don’t believe freedesktop.org had up until this point. Before you are quick to think nothing will come of any of this, just look at freedesktop.org and stuff and see how far things have come. I remember the first time I used Linux and how horrible even simple things like copy and paste were. Thing are getting standardized, we can only that this will push it quicker. After all, whether you’re a fan of gnome or KDE, it’s in everyones best interests. 2005-12-09 10:39 am My first though was … not another projects … but then I read what it was about and I think it’s a great idea. It’s about time that there is some commonality amond *nix desktop environments and some sound standards. After all there is a lot of knowledge in the community but stupid decisions and bias behaviour constanty drive both users and developers away. Ever wonder why Open Solaris seems to be such a success and why Ubuntu of all distributions somehow got over 50% of the linux users? It’s about time that the OSS developers got back to their sences and started working together. After all it was only why the concept of open source was created … 2005-12-09 10:53 am Ookaze Ever wonder why Open Solaris seems to be such a success and why Ubuntu of all distributions somehow got over 50% of the linux users? No, I know the answer : the thing you’re smoking is definitely not so good. It’s about time that the OSS developers got back to their sences and started working together You’re mistaken, OSS is not a company. After all it was only why the concept of open source was created So it was really flawed like RMS implied ? 2005-12-09 11:24 am End of the war A preview release of the LSB Desktop specification doesn’t mention libqt or Kde… http://www.linuxbase.org/spec/book/Desktop/Desktop.html 2005-12-09 11:42 am raboof To prevent rehashing an old discussion all over again: 1) read http://www.linuxbase.org/futures/ideas/issues/libqt/ 2) this doesn’t mean libqt will not be installed/used, merely that it isn’t part of LSB, which is not irrelevant but not such a big deal, either. 2005-12-10 10:05 am There is no war, and Qt 4 will be in there: http://dot.kde.org/1134088861/1134125881/1134137563/ 2005-12-09 12:03 pm The desktop IS the common ground. This should have been done ages ago. Linux and its ilk will never be a -serious- contender if it cannot deliver to the desktop what everybody in the Windows and Aqua world is already used to: a stable environment that doesn’t play tricks on you. If an outfit like Microsoft can get something like a desktop environment to work [kind of], wouldn’t you expect the people with a little more savvy to be able to build something that is so vital to the computing experience and make it robust enough to be actually usable? 2005-12-09 9:33 pm John Nilsson This should have been done ages ago. Linux and its ilk will never be a -serious- contender if it cannot deliver to the desktop what everybody in the Windows and Aqua world is already used to: a stable environment that doesn’t play tricks on you. THERE IS NOT COHERENT MOVMENT BEHIND “LINUX”! Linux is a kernel, NOTHING ELSE! What you and I refers to as “Linux” is a group of related systems. They are SEPERATE PROJECTS, by DIFFERENT PEOPLE with DIFFERENT GOALS. So you are quite right, “it” will never be a serious contender, becase there is no “it”. Edited 2005-12-09 21:37 2005-12-09 12:17 pm bringmewater Will they make it so newly installed programs create icons and so Printers are easy to install? Then I’m all for it. Let’s get the basics please? 2005-12-09 1:33 pm It will make desktop linux even slower… How about building a X from the ground, speed focused, build a very configurable, themable, skinable and fast DE, all the on top of a standart framework, easy and acessible to all kind of developers ( somethind .net/java like), with point and click, visual development tools? We need convergence not competition, linux DEs competitors should be OSX/WIndows not thenselves. My english is shi* i know. 2005-12-09 2:42 pm amadeo See: http://mail.freestandards.org/pipermail/lsb-futures/2005-December/0… This, combined with the fact that the standard on Qt is being worked actively in the desktop group, probably means that the restrictions are lifted, and Qt is clear to go. Or not? 2005-12-09 3:31 pm segedunum, this time you are clearly missing the boat. AFAICS, Project Portland was wholly inspired by Martin Konold’s previous “RUDI” proposal [see http://www.kdedevelopers.org/node/1398 for details]. It is by far not “another ill-advised attempt at integration where integration just isn’t possible or just creates more work than is actually required”. This time you are clearly off target. 2005-12-09 9:44 pm segedunum this time you are clearly missing the boat. I’m afraid not, and here’s why. Both you and Martin are looking at something that draws a huge circle, and you both believe that you’ve ended up in a different place when you’re actually back where you started. I am all for compatibility layers when they make sense. When you’re creating desktop environments, for example, there are going to be people who have written software for older environments and that software still needs to work in KDE 15 with all the latest wizz bang features. An example is when people moved from Windows 3.1 to 95, and the applications written for 3.1 when run on 95 inherited new UI features amongst other things. ABI compatiblity in that manner is perfectly sensible and required if desktop Linux is to go anywhere. Martin hints at this ABI problem between different versions, but he describes totally the wrong solution to tackle it. What you have to do is take responsibility for the interfaces in the previous versions and continue to provide support for them, or you sensibly deprecate interfaces from the main libraries and continue support for them in a thin compatibility layer. You don’t run away and hide in a new system, and in RuDI you still have to program a translation in for them. The ABI problem for C++ of GCC has an obvious solution – ABI stability of GCC needs to stabilise. You never, ever create a whole new layer just to get around a problem that shouldn’t exist in the first place. The problem with Martin’s RuDI approach is that he does not even really go into the problems above for coming up with what he’s done. He goes off on a huge, totally unrelated, tangent talking about not having a monoculture and supporting XFCE in a bank when there are already 400 KDE workstations. Is the security risk that great from a monoculture of KDE that this bank needs to consider putting XFCE in, together with a RuDI layer on top of that which allows the same applications to be run? No it isn’t, it’s just total lunacy and no one out there thinks like that. Martin then goes on to describe a whole service type architecture, of some complexity, and because of that complexity it really is a whole desktop enviroment in itself – all to insulate ISVs, and even whole programming environments, from different desktop environments! And what happens if there are differences between the environments (which there definitely will be) such that RuDI or Portland cannot translate for an application? You get very poor integration into the desktop environment, especially for people like IHVs, to the point where it becomes useless. Interestingly, you also have to come up with a common and agreed dialect which can be translated into each desktop environment or you simply get too many differences, which leads us right back to the arguments this architecture is supposed to solve. You want several hundred XML schemas and committees describing translations for different desktop environments? Yay! Why not just make RuDI or Portland into a desktop environment, can the underlying desktop and get rid of the complexity? Why, you might ask. Because more effort for zero return will go into this than actually improving the desktop environments, Portland will become a roadblock, too restrictive in terms of its functionality and people will continue to port to different desktop environments regardless, as they have always done. It is by far not “another ill-advised attempt at integration where integration just isn’t possible or just creates more work than is actually required”. Since RuDI (or Portland) hasn’t got off the ground or been proved in the real world you can’t make that statement. This time you are clearly off target. I am still not convinced otherwise. I’m sorry, but this whole thing is a touchy, feely, nicy, nicy, cross-platform, cross-desktop integration thing that makes zero sense in the real world. If you look at this whole sorry thing you find it makes more sense to an open source developer than it does to an ISV, and it makes absolutely zero difference to a user. It certainly doesn’t make the applications any better or the development tools any better. That’s what needed to be discussed at these OSDL meetings – stuff that actually makes a difference to ISVs and users, not stuff to solve the open source world’s problems. 2005-12-09 4:31 pm ### “GNOME is popular in the states and KDE is ### popular elsewhere, does it really matter?” —– Yes, it does. Given the fact that Desktop Linux is more popular in Europe too, I wonder if there is a a context? 2005-12-11 12:58 pm dylansmrjones Blah…. 2005-12-10 10:54 am Torsten Rahn > Most survey are self selecting. KDE is a > much better fit for the advanced user than it > is for beginners or office workers with > no particular interest in desktop environments > or operating systems. KDE is a good fit for both of these target groups. It’s a good fit for beginners and office workers and a very good fit for the average user. That’s e.g. why – it usually gets chosen by desktop distributions as the default (unless there are political or historical reasons to decide otherwise) – it’s usually the preferred desktop of Linux newbie magazines (like TUX or the newbie edition of Linux Magazin) – KDE Apps is still a lot more popular than e.g. Gnome Files. In average it does not only offer more applications per day and in total but also has higher download numbers from what I’ve seen. – it also gets chosen often in polls that target newbies and received lots of awards for that By offering (usually well hidden) functionality like KIOSK, DCOP, kioslaves, advanced printing and features that make life easier for the advanced user it’s often also the choice for power users as well as sysadmins. But of course this resembles the smaller (but not less important) part of our user base. > It is easier to use for the newbie, Sorry but this claim is unfounded. KDE as well as GNOME are both actively working on usability since years. Both of them have their strong points when it comes to usability — as well as their weak ones. KDE’s strong point is that it usually is configured in a way that it looks and acts very close to the Windows environment (while trying to get around the weak points of that interface). It’s a well known fact that one of the biggest obstacles migrations are facing in companies is the acceptance by its users. The biggest fear of the user usually is that he can forget about all of his previously collected knowledge and start from zero. Most users did already use a computer before they were offered Linux and for most users the OS previously used was Windows. The fact that out of the box KDE is configured in a way that makes it look and act quite similar to Windows pleases those users very well. Being involved with migrations in companies as well as public authorities I know that there are a lot of companies which just choose KDE for the reason of ease of use (but of course I also know that still there’s also a lot of work ahead when it comes to usability).