Today I read a news post about the KDE 3.5 fork Trinity and it was discussed that you should use it because of the bad performance of KWin:
the one constant is abysmal Kwin performance.
So to me it sounds like the author of said news article things it is a better idea to run the KDE 3.5 fork, because there KWin is without (OpenGL) compositing, instead of just turning off compositing in KDE 4. Well that’s interesting reasoning. (Btw. if you do not run at least 4.7.2 you should not complain about KWin performance).
Now to my actual blog post: I appreciate the idea of the Trinity developers to bring back the KDE 3.5 desktop experience to those users who really want it. This is a great offer. Personally I doubt that the Trinity project is doing the effort in the right approach. Instead of just porting Kicker and KDesktop to Qt4/KDE4 they forked everything and the kitchen sink. I rather doubt that a team which is smaller than the KWin team is able to maintain not only KWin but also every other part of KDE 3.5 and now also Qt 3.
The developers of Trinity offer the users something that their system cannot provide: a future for complete KDE 3.5. I did a lengthy analysis of Trinity for German readers in the freiesMagazin some time ago: Trinity – Desktop ohne Zukunft
Actually what you write about porting kdesktop and kicker to Qt4 has been done by the Trinity project. The current code can be compiled against both Qt3 and Qt4. However there have been multiple performance issues were Qt4 was performing significantly slower than Qt3. This is why there is less focus on the Qt4 port now.
It appears these performance issues are inherent to Qt4 and cannot be easily solved. If there is a way to let the Qt4 port perform equally to the Qt3 port, the Trinity developers are very interested in help with this.
Well we all know that a Qt 3 to Qt 4 port is not trivial and requires changing the way how to do your painting. Yes of course it is likely that the Qt 3 code renders slower with Qt 4 and the Qt 3 support classes or if the porting has been done badly. Don’t forget that the developers abandoned KDesktop/Kicker and implemented a new and better solution.
This seems like a valid reason to keep Qt3 for now then however, since it is obviously more work to get a fast kicker and kdesktop running on Qt4 than keeping Qt3 working for now. (And you already found yourself that Trinity is low on manpower 🙂
To make it clear, I do not agree with statements that kwin performance would be an argument to completely switch to another DE. I myself had issues in the past with newer KDE versions, especially on systems which had barely/no graphics acceleration (not even 2D). KDE 3 and Trinity ran more smoothly in those cases, so I stayed with that. Not wanting to bother with different DEs on different boxes, I ran Trinity everywhere.
I understand that graphics performance has significantly improved now however. Would you even recommend testing it on underpowered systems, which lack in the 2D & 3D graphics acceleration support? I would expect the KDE 4 developers to mainly focus on supporting modern systems with proper 2D & 3D graphics acceleration and caring less about older systems.
Right now, I still have the impression that KDE 3.5/Trinity is more suitable than KDE 4 on older hardware and that the code quality of KDE is so high due to all it’s great developers that even the older versions can be maintained better than the other light-weight DEs.
How Trinity did the “port” is that they implemented their own, much more extensive Qt3Support (http://websvn.kde.org/branches/trinity/dependencies/tqtinterface/) and scripted a mass Q?TQ class replacement.
In fact, that “tqtinterface” is a wrapper (kind of an abstraction layer) around basically all Qt classes, with both a Qt 3 and a Qt 4 version, where the Qt 3 version usually just forwards to the Qt 3 class.
I’m not all that surprised that reimplementing all of Qt 3 in terms of Qt 4 as they did can lead to performance issues.
yeah and by that they also nicely broke binary compatibility of e.g. KWin window decorations…
Right. I was initially considering using Trinity for our kdelibs3 compatibility package in Fedora, but their binary-incompatible and probably even source-incompatible TQ* changes made it completely useless for that purpose (and made me glad we hadn’t already made that move when this happened). (I doubt KWin is the only place where binary compatibility was screwed up, the libraries got infected with that TQ* craziness as well.)
Hey,
I’ve just seen your post and would like to note that the developers of trinity have not made the comparison for kwin on kde3 or kde4.
Kde3 is inherently faster, there is no argument. It performs better, there is no argument.
As long as there is an interest in the project, it will remain alive.
Of course there is an argument. First of all I would say that you compare things which cannot be compared. Another thing is that we would have to define “kde3” and “kde4” before doing the comparison. What are we talking about? Just the desktop? Or also the applications? What about the new applications in “kde4” and in general all the functionality not present in kde3?
Obviously, but it completely misses the point of my blog post.
Well, on my machine KWin3 (with XRender compositing) is more snappy than KWin4 (either with OpenGL or XRender compositing), on all of nvidia, fglrx or radeon. KWin3 doesn’t have wobbly windows or fancy shadows or zoom, but it does have transparent windows where needed, it allows moving/resizing windows without redrawing the windows below and it doesn’t suffer from sync issues with urxvt. In other words, it’s not fancy, but it’s fast and stable.
I know that KDE4 has its advantages and I WANT to switch, so I can finally use activities and better window switching and all the awesome stuff. So I’ve installed every single release since 4.1.x, but every single time KWin has been slow (including 4.7.2).
I know what you’re saying – “it doesn’t work that way any more, because it’s not comparable, and it’s not my fault.” Of course I must agree with that, and noone can rightfully fault you for the awesome work you’re doing with KWin.
But the part where “it doesn’t work that way any more” still stands, the responsiveness has regressed.
The same has happened to other aspects of KDE. There are a lot of improvements over KDE4, but there are also regression remaining. Depending on your specific needs and priorities, KDE3 may still be the better choice.
Again, I’m not faulting you for the inherent slowness of fancy effects, and I look forward to the day when *stable* 60fps have been achieved. But for now, there’s a desire for updated KDE3 versions that compile against the latest kernel/plumbing/X11/drivers and don’t force you into ancient distributions. I agree that the future is KDE4, but for now I’ll stick to KDE3.
The grass sure looks greener on the KDE4 side, but every time I try to cross the fence I get shocked :p
if you are on nvidia hardware try to use graphicssystem raster or disable the Oxygen animation. I rather doubt that Kwin3+XRender could be faster than a correctly configured KWin4
It will be nice to get some tips about improving Kwin’s performance. I’m running KDE 4.7.3 on Arch Linux and Ati x1600xt graphic card. Kwin is quite fast on my box, but there are some situations when it slow downs. In example, when there’s Kadu (client similar to Kopete, but focused on gg protocol) opened and I click on K menu the animation is choppy. I have v-sync turned off. Btw. big thanks for the ability to suspend compositions when playing full screen games! Saves a lot of FPS. 🙂
KDE3 is faster than KDE4 like Dos 1.0 is faster than Windows 7. Yes, it users less resources. Obviously, as it does far less.
Jos,
You have a hit a very very interesting point. And painful one.
It seems although computer hardware is much better, boot time and general perfomance does not improve.
This is especially through for WINDOWS Family OS’s, but a also for Linux.
And from my point of view, if KDE4 is slower or the same speed on my new Laptop comparing to my 5 years old laptop with KDE3, than I rather not use it.
that is probably one of the fundamental laws of computer science. If the computing power increases also the demand of the software increases.
> If the computing power increases also the demand of the software increases.
That’s a stupid proverb: software has no demand, only software developers make choices.
“It seems although computer hardware is much better, boot time and general perfomance does not improve.
This is especially through for WINDOWS Family OS’s”
That’s not true.
Nowadays a typical win32 PC (Windows 7) only takes a few seconds to boot.
In Windows 98 times, I remember it usually taking more than a minute.
I’ve never seen windows 7 booting in few seconds. It always lagged much behind Ubuntu which starts fast!
“And from my point of view, if KDE4 is slower or the same speed on my new Laptop comparing to my 5 years old laptop with KDE3, than I rather not use it.”
But even assuming that’s true (I’m not sure it is, e.g. according to my subjective feeling, application startup times have *vastly* improved with the KDE 4 on todays hardware combination compared to KDE 3 on slightly older hardware), is low-level performance really the most important factor?
Isn’t overall productivity what actually counts?
Is it really advantageous to use a system that has a 10 milliseconds latency when moving windows instead of 50 milliseconds, but on which the task you’re working on as a whole takes 2 hours to complete instead of 30 minutes, because it simply doesn’t offer as much convenience functionality and streamlined usability?
The thing that you constantly forget is that sometimes “less is more” especially if the “doing more” is buggy AND activated by default which seems to be KDE’s current policy.
I used to think so too. Esp the startup times of KDE4 were always behind that of KDE3 – even in today’s h/w KDE4 login takes up about as much time as kde3 did with 3-4 years old h/w.
I was always turning off desktop effects thinking that that would speed things up but I was wrong ! Turning on desktop effects allows kwin to use h/w openGL etc where possible, dropping the start up time to a third of the original. Now KDE4 starts up in 5 secs from cold boot on my h/w. It is definitely faster than xfce ( I haven’t tried E17 in a long time). All in all, the desktop is much more responsive than windows 7 on same h/w (both with effects turned on) and OSX on any h/w.
One of the reasons why KDE4 (out-of-the-box without any tweaks) may startup slower then KDE3 did is that the new Plasma desktop is heavily plugin based. That gives us lot of flexibility (something like Plasma Active or Plasma Netbook was never possible with kicker+kdesktop for example) but also has the drawback that loading tons of plugins can be slower then loading no or very few plugins.
I guess that confirms what Jos just wrote. More flexibility and a changed architecture cause of new requirements (different form-factors, different interfaces, all plugin based to make it easy to change parts while reusing others, …) so it does more and can be (and already is) used for more but that has also drawbacks.
Let me add that the new plugin architecture doesn’t only provide more flexibility but makes it imho also much easier to add new functionality or maintain existing functionality. Looking at the number of plugins Plasma has this days compared to what we had in kicker+kdesktop at KDE3 and from where KDE4 move to (e.g. Plasma Active on NVidia Tegra 2) seems to confirm that.
> One of the reasons why KDE4 (out-of-the-box without any tweaks) may startup slower then KDE3 did is that […]
I actually don’t consider a valid reason, but an excuse. Look, I love KDE 4, and use it every day. Even on my MacBook.
I just read the story about the boottimes of the first Macintosh. There is a lesson there.
It booted slowly, with all the valid reasons. Then Steve came in, demanding the boot time to be shortened by 10 seconds. The tech guys said it was impossible, yet Steve mentioned the numbers in a different way. It was something like take 10sec multiplied by the how many years all people in the world combined would be waiting. The guys got to work, looking through all options. They managed to cut to boot time with 28 seconds!!!
It makes this look just quite funny: http://www.youtube.com/watch?v=tmRJ649ICPU 😀
Rather then saying why we can’t do it, I’m more curious for ways how it can be done.
There has to be a way. Even with my SSD drivers today, I still see the bouncing cursor in KDE 4. Dolphin and Gwenview take time to start. I don’t see that in Windows 7, where all the apps are instantly on the screen as I start them. So, please tell me, how the startup time can be improved!
I hope you are joking because every time I click on an app icon in windows 7, it takes 5-10 secs before a dialog pops up, asking if you WANT to do something with it, assuming you can even FIND the icon/menu item to click.
windows 7 snappier, NO WAY. It is dog slow, like vista.
My KDE4 apps start immediately. On windows they’re a lot slower.
I somehow don’t like these kind of posts from KDE4 devs. Belittling other projects and saying they have no future. Do you have a magic eye that can look in the future and say what project will get traction and which will wither? I think it is great that the code is out there and somebody cares to work on it.
Thom from OSnews will always complain about KDE4 composite performance, maybe he just does something wrong, why don’t you offer him a debug session. I guess you can tell him which distro to use and how to configure his onboard ATI card. That would definedly make good news for KDE4.
(KDE4 works perfectly well and fast on my box)
I recommend you to read the article I linked (Google Translate should be good enough). It shows the arguments why I think it has (with the current manpower) no future.
Martin,Look man, KDE4 has a damaged reputation still sadly. And still has issues please fix those issues before you diss others making a desktop that works better then yours.
Nothing bad in your comment, however, the last line made me feel as if KDE doesn’t consider KDE 3.x as a part of it. No, KDE3 is “ours” too.
I imagine that Linus Torvalds might start using it. That alone would ensure it has a future.
“just turning off compositing in KDE 4”
and have black triangles around the shutdown dialog 😉 your last advice to that was…. enabling composition, I’m happy with composition cause kwin is getting so much love nowadays and is getting better and faster with every release. I don’t know how are other composition-disabled-glitches in 4.7 era but you shouldn’t recommend solution which is not a entirely polished
PS those black corners are there from kde4.0 and nobody fixed that, I know this is trivial bug and it doesn’t crush anything but it’s ruining kde experience!
Other than that, great work kwin team and martin especially!
And are black borders around the shutdown dialog a valid reason to use a fork?
If you know that it is a trivial bug, why don’t you fix it? If you now say that you cannot develop or do not have the time, than you cannot know that it is a trivial bug.
I meant trivial like: not doing any crushes or other harm to system other than just esthetic inconsistency and there are far more urgent things to fix. If it was a big issue I suppose it would be visible on other rounded corners also, not just 3 out of 4 corners in shutdown/logout dialog (which means one of them is working properly)
You’re right, I’m no developer and I admire you guys, I thought developers should be open to what people have to say (I know it must be hard sometimes). Some users love x, some hate y, and I don’t like little papercuts as much as big bugs, we all have the same goal – make kde superb
I keep hearing how Trinity doesn’t have a future, and it reminds me of the time when I was hearing that a fork of KDE3 wasn’t going to happen. Why not? Big job, not enough developers, sha na na shboom shboom. Same song, different verse.
Timothy Pierson has been working on preserving KDE 3 ever since Kubuntu dropped it. He’s never stopped. The releases tend to be late, but they work better all the time I loved the last Kubuntu release, and the beautiful way it integrated KDE4 applications into the KDE3 desktop.
Maybe it doesn’t have a future. Fortunately for me, this is the present.
Given the bugs Trinity introduced for example in their KWin fork, I would not even say that it has a present.
Sorry about that last messed up attempt at reediting. I just wanted to add that my issues with KDE4 have nothing to do with Kwin.
Don’t really know about kwin, but qt3 surely is faster at painting windows, for example comparing dolphin (qt4) with konqueror (qt3).
Try switching to the raster engine of Qt4, for me it’s considerably faster.
>>Given the bugs Trinity introduced for example in their KWin fork, I would not even say that it has a present.
Well, MEOW! Looks like Trinity introduced a bug up someone’s ass!
I find the responses very odd.
As I only started using Linux when KDE4 was introduced, I am not qualified to make any comparisons. However, I cannot understand how kwin performance is being criticised.
Yes, for the first few KDE4 releases, kwin was slow and most KDE apps were rather bare (not as much as Gnome 3 atm) so I did move away for a while but since 4.4, I have not had any issues with kwin. Rather, I have had problems with plasma-desktop which is excused as it is far more advanced than KDE3’s (and most of the problems are due to the Plasmoids).
The argument that one should use Trinity due to KDE4’s low performance is rather odd. If this argument was to be extrapolated, all Windows users would have to install XP. As we all know, XP has some great futures lying ahead.
Anyways, kwin for the win!
I complete agree with this article. This projects goes to nowhere.
does exists a KDE 2.x fork today? I think not.
Could Trinity offer a solution for a mobile or netbok devices? No way. KDE4 does. Next year, we have probably qt5, and this folks scarcely have migrated to qt4.
IMHO, this project is a wasted time project, if they think that qt4 need improvements, go ahead, if they think that kwin could be better, try to improve the new version. No look back anymore. KDE4 is here for stay, and a KDE3 fork doesn’t matter anymore. If you want some light desktop environment, people have other choices. Or even Razor-Qt!
If it’s on my Desktop right now, it’s not looking back.
Actually, I’m not using Trinity right now. I’ve got a live system based on Slax 6.1.2 and Slackware 12.2
http://www.kiaragnulinux.blogspot.com
I installed it to my hard drive as a live system that doesn’t update, so there are fewer security issues. The system resets itself with a reboot, and all the software is legacy from the same period, so it runs really smoothly, lets just say a lot less buggy than Kubuntu 11.10.
On my other computer, I’m running Kanotix (based on Debian Stable), mostly KDE applications, managed with Fluxbox. Also kickass. I have a lot of respect for the great work that went into KDE4, and the applications are the best, but the Desktop is shapeless and distracting for me.
I used KDE4 for about a year and a half, from 4.4 to 4.6, before I went back to KDE3, and I discovered that after all that time, KDE3 was just a pure pleasure to use. KDE4 has usability issues that KDE3 never had. KDE4 makes me think about the desktop all the time.
Nobody wants a fork of KDE2, and that’s why there isn’t one. Maybe in two years nobody will want a fork of KDE 3. I hope that in the meantime, you can find the strength to live in a world where people DO want to use KDE4.
I hope that in the meantime, you can find the strength to live in a world where people DO want to use KDE3.
There.
> Btw. if you do not run at least 4.7.2 you should not complain about KWin performance
If it takes 7 major releases for developers to do basic performance profiling of paint code, then why don’t people have the right to complain? You’re only validating what the Trinity developers are saying
Who says that there was “basic performance profiling” involved? Who says that we did not do other profiling before? Btw. I did the profiling, I did implement an optimization for 4.8, but I did not notice that there was a bug in the first place.
No not at all, because in case of KWin we talk about a feature which can be turned off and is not available in Trinity. This is like comparing apples to pears.
> Who says that we did not do other profiling before?
In your own words, you were surprised at how heavy the effect chain was. It was “way more heavy than expected”. If you’ve done the profiling before, how could this suddenly be a surprising result?
Well that is obvious that we would have spotted this bug before if we had done this profiling before. Obviously we did not do the profiling for that part of KWin till that moment. But this does not conclude anything about that we did not do profiling before, we profiled other parts.
In fact I expected that the effect chain is heavy but I was not sure if it is so heavy that it is worth to spend the time on to optimize it. That’s why I did the profiling. Before there were other things to optimize (think of the OpenGL 2 compositor), so I did not see the need to go for a small gain in the effect chain.
> Obviously we did not do the profiling for that part of KWin till that moment. But this does not conclude anything about that we did not do profiling before, we profiled other parts.
Sure, that’s why I specifically said you didn’t profile the “paint code”. I’m still surprised this was never profiled before considering it’s central to the window manager and is the most obvious place to profile. When people have always complained about the heaviness of effects, profiling the effects chain would seem obvious to me.
of course we profiled the paint code. Just not that part of the paint code (in fact the effect chain does not contribute to the painting at all). It is an extreme difference if you profile e.g. just the painting of a specific feature (e.g. a window) or the interaction of all effects.
If calling out to the effects chain is part of the repaint loop, are you saying you profiled other aspects of painting, but not the repaint loop itself?
Anyway, it seems ridiculous to me that you’re belittling the performance complaints from Trinity developers when clearly there are still performance issues to address. Stating that people can’t complain if they’re not running 4.7.2 simply implies that things were relatively bad before. Even if that’s untrue, you’re shooting yourself in the foot.
once again: Trinity does not offer OpenGL based compositing and KWin offers OpenGL based compositing which can be turned off. There is no way to compare these two products and you just cannot state that Trinity is faster of performs better because it does not use OpenGL compositing. In KWin you can turn it off and you have the same window manager just the KWin one has less bugs, a more modern code base, more features, more window decorations…
Please try to get to understand what I criticize about Trinity which is completely orthogonal to any performance issues in KWin’s OpenGL compositor.
I may be late to the party (last comment was dated yesterday if I am not mistaken) but I think many people are comparing apples and oranges. You can’t compare KWin3 and KWin4. They’re just different beasts under the same name. But IMHO you can compare KWin4 with Compiz or whatever wm Gnome is using (talking about Gnome 2 here, haven’t tried Gnome 3 yet). I’m a KDE user at home but at the office I have to use Gnome also and Compiz wins hands down on the very same machine with obviusly the same drivers to the point that I’m logging in Gnome far more often than in KDE. KWin4 often gets choppy, compiz never. It’s not just the effects. Turning them off speeds things just a little. At times it’s a true PITA. I can’t tell what Compiz devs have done right in comparison to KWin devs but there must be something.
No you cannot even compare KWin with Compiz as Compiz is mostly OpenGL 1, while KWin has an XRender, OpenGL 1 and OpenGL 2/OpenGL ES 2 backend. You can just switch the backend.
And again: update to at least 4.7.2
And you cannot compare KWin with the WM of GNOME 2 as that is Metacity which does not do OpenGL based compositing.
I use Gnome 2 at the office with effects enabled. Not very professional maybe but I like it this way. IIRC this puts Compiz in the equation at the expense of Metacity. I might have been clearer tough. As for OpenGL1 versus 2 you may have a point here. I didnt’t know. Regarding 4.7.2 I didn’t upgrade yet because each time I do so the profile goes bad and I have to delete the .kde directory losing all my personalizations. But, even if I did, since the system is slow even with compositing disabled would it make a difference? I was under the impression that the major fix in 4.7.2 is the fact that it doesn’t call inactive effects. Does it provide perceptible speedups even without compositing?
If the system is slow without compositing it is completely unrelated to the window manager/compositor. Then something else is causing the slowness.
To the author. It seemed just as hopeless and pointless back in the day when people first made plans to fork kde3 desktop. They forked and even released. To say there is no future is a little sensationalism for trinity. There is progress on the trinity project, there is releases, the future i would say be somewhat indeterminant.
But, with linux users being up in arms with gnome 3, unity, and at times kde4; there are plenty who are moving over to lxde, xfce, and other more traditional desktop environments that are also speedy. Trinity may be added to that least for all we know.
Please read the article I linked which explains why I doubt that there is any future for Trinity. This has exactly nothing to do with the number of users, but is technical.
The german magazine post that you made a while ago? Or the kwin performance thing that is not linked? I did read that one a while ago through a different source.
You’re post isn’t that technical. It’s mostly a rant in response to that stupid kwin performance of trinity and an old kde4 version comparison article. You still ultimately end with talking about the future of trinity which is what i was responding to.
A matter of a number of users? No, I was saying that trinity may be considered an alternative to gnome 3 and kde4 a lot like lxde and xfce have been. Not number of users, that it is an alternative choice (it already is, but become a noticeable alternative the main thing).
I mean the German magazine post.
And if you read my German magazine post, than you will notice that this is exactly my critics. Users think it is an alternative, but Trinity just cannot deliver what they want to deliver.
Seriously, two years ago we had a bunch of posts explaining why a forkm of KDE will never happen. Now we’ve got people explaining why it will never last. Wel;l, it’s a shame that it’s not going to last, because I sure do enjoy using it right now!
Anybody who wants to try Trinity, I recommend the Ubuntu CD.
The question is not whether the fork exists or not, but whether they can deliver something that satisfies the users. And that’s where I criticize Trinity. Why for everything if it would be enough to maintain Kicker/KDesktop? How will a patch to KWin from KDE 4 get to the KWin fork in Trinity? Is it worth to spend the time again to fix the same bugs? To do the same porting efforts to CMake? The manpower is limited and I think that they concentrate on the wrong parts and due to that don’t have a future as they promise a future for what used KDE 3.5 but that is just impossible to maintain with their manpower in the quality known from KDE 3.5.
Well, maybe you guys should have thought of that before you turned KDE into a big sprawling mess. If you can’t respect the KDE3 community for pursing it’s own vision, what respect does your big slow shapeless dirigible of a desktop that still isn’t quite done yet deserve?
It worked fine, and you broke it. And now you’re being arrogant about it. This is not the way to win the people you have lost.
You are aware that I started to contribute to KDE around the 4.0 release? So I did not break anything, I worked completely on an application already present in KDE 3.5
Wow, that’s literal! Thank you for not pointing out that KDE4 is not a dirigible.
Look, my little box is closing up fast, so I’m just going to promise you that that no one anywhere who read your words decided to stop supporting trinity. They all redoubled their efforts, because you are being infuriating. Trust me on that.
I aplogize for my anger. KDE4 is a very impressive effort that I respect, even if I find it too shapeless and distracting to be very usable personally, and I enjoy running KDE4 applications on Fluxbox. The are the best Linux applications to be found anywhere. As far as Kwin is concerned, i don’t really have a clear idea what it is.
So you are not able to understand any of my arguments. Fine. I explain a little bit.
The Desktop Shell which got rewritten for KDE 4 is called “Plasma”, it replaced Kicker and KDesktop. KWin is the window manager and has been introduced with KDE 2. It is a fundamental part of both the KDE 3.5 desktop and the Plasma Desktop.
So you want to keep the KDE 3.5 desktop experience aka KDesktop and Kicker. That is just fine and I appreciate any effort in that are. My critics concerning Trinity are not that they keep KDesktop and Kicker alive, I am concerned that they also keep those applications “alive” which have not been rewritten for Plasma. I hope it gets clearer now.
Actually , yes it does. Now let me make MY point.
I don’t know dick about development, but it’s just bad manners. I feel kind of foolish, but let me tell you in the broadest terms that people who are connected with KDE4 who criticize trinity have very little to gain by it. If you are wrong, you will piss people off, and if you are right, you will enrage them. And they’re not going to do what you think they should do.
KDE4 is an impressive project with a lot of good ideas, and the most advanced Desktop applications I’ve ever seen. After Gnome 3 and Trinity I especially respect KDE for focusing on the Desktop, and not on gadgets. I want to respect KDE for its achievements, and you make it difficult. Jesus, I hadn’t badmouthed KDE4 in a long time, and there I was, calling it a dirigible, as if that made some kind of sense. No matter what your intentions, telling people they’re wasting their time and their project is doomed is the REAL lost cause here. Timothy Pierson has been working on Trinity for three years, is he going to stop because he read your article?
Well, that’s my point, and whether you agree or not, I’ve got nothing to be gained by repeating it over and over. A sincere good day to you, sir, and thank you for your hard work.
No and that is not my intention. But maybe he sees that all which is needed is the KDE 3.5 desktop experience which is KDesktop and Kicker and he concentrates on just supporting those instead of forking everything and the kitchen sync.