Submitted by Groklaw has interviewed KDE about some recent misconceptions about KDE 4. “There has been a bit of a dustup about KDE 4.0. A lot of opinions have been expressed, but I thought you might like to hear from KDE. So I wrote to them and asked if they’d be willing to explain their choices and answer the main complaints. They graciously agreed.”
<quote>
Socially speaking, it assumes redirecting development effort is effective. To a certain extent it is, but for quite a few developers not developing features often means not developing at all. Not everybody is good at low-level stuff (or willing to do it). This has bitten us and other projects in the past and present. The development of KDE 4 has been slower than it would’ve been if we would be able/willing to force developers to work on whatever some top-down managers think is good…but it’s just not how FOSS works.
Have you guys heard of the old 20/80 rule? You spend 20% of your time implementing your features. Then 80% of the time testing them and debugging them. What you’ve pretty much told me in this paragraph is that you guys only do the 20 part, and nobody wants to do the 80 part, and that’s how FOSS works. Thanks, really. You’ve explained a lot.
</quote>
He is right!
Of course I’m right. You can suck my dick any time, freetard.
No, he’s not. And he is, probably full knowingly, misinterpreting the quote. What is being said is that the time KDE developers spend on new features could not be spent on lower level stuff like the usual whiners keep pretending. He makes it appear to be about QA and how FOSS developers don’t test/debug their code.
I’m sorry, but the fact of the matter is that in the real world FOSS software is in that regard usually better off than its commercial counterpart.
Depends on what software. Low level infrastructure stuff, I would agree. High level GUI stuff however is usually buggy as hell, and it is like pulling teeth to get anyone in the project to listen to you.
And tell me, when was the last time that MS fixed a bug you reported to them? or added a feature you asked for?
And don’t tell me Word, Outlook, IE, Visio, etc. has never crashed on you, taking away your time or your work. There is this bug with MS Word, where it crashes with a ‘not enough memory’ when opening a document. It’s been there since Office 95, instead of fixing it, MS gives me a ribbon instead… Yes, Companies are so much better at listening to their users, sure.
I work for a microsoft gold partner, and our VP of development is an MVP. When we report bugs, they get addressed.
As for features, a month ago I made a good argument for supporting POCO objects in dynamic data, and with the latest push we got a DynamicObjectDataSource that does just that.
I’m not saying that. I am saying that the vast majority of linux software is pre 1.0, which by its version number means the DEVELOPERS WRITING IT do not believe it is production ready.
And just so you don’t think im just bashing linux, I also think by and large the GUI apps on windows leave a lot to be desired in terms of usability, consistency, and esthetic, especially when compared to mac apps. And that if you ignore tooling, free software really sets the bar on infrastructure stuff and protocol specs.
As an MS gold partner I’ve no doubt that you’ve got an attentive ear from MS. That, however, is not the case for the vast majority of MS customers.
As a “basic” user of FOSS you have access to developers, you can submit bugs and feature requests. Sure, you have no guaranty that those requests will be addressed but at least the communication channel is there.
And I’m sure that for the kind of money your privileged relation with MS is costing you you could get the same level of service from a commercial Linux distribution.
As for the quality of Linux desktop software, using it daily both professionally and at home, I must say that most applications don’t feel the least like beta software. Firefox, Konqueror, Amarok, Krita, OpenOffice, KDE in general, none of them feel shaky or amateurish to me.
Now on the other hand using Windows always feels like an adventure, the question not being “will it crash?”, but “what will crash and when?”
Agreed! Let’s face it, one customer (particularly an *individual*) is neither here nor there for MS, but individuals tend to get better attention from Linux distros. Not saying that always happens, but it’s been my experience.
Agreed.
Ohh, yes…. Boy, is this ever true (when I was using XP). You just never knew if the thing would boot successfully or not. As for the “snapshots” that it used to have (for system-restore purposes), I found them to be utterly useless. I found XP to be very poor quality in terms of robustness. The amount of anti-virus software needed was ridiculous too. MS would do well to learn from OpenBSD’s “secure by default” approach.
It depends on what you mean by “individuals” and how you quantify “attention”. Are you talking about developers or consumers?
“Individuals” meaning Joe Bloggs, as opposed to a company. “Attention” basically meaning “help”. So, my reply was from the view of consumers rather than devs.
Also, Linux has been around for long enough now that very many answers to problems can be found via Google.
( Not *all* answers, but many of them.)
If that doesn’t work, a detailed problem description on a forum usually gets very quick results, I’ve found.
On a similar theme, I’ve been doing a lot of app-testing for the Wine project recently, and my experiences with them have been *excellent*! They’re a pleasure to help.
If MS’ paid support is better than the free help I’ve got from Google and forums, it’ll have to be pretty darned good….
Edited 2008-07-14 08:01 UTC
You must not be looking very hard. I’ve gotten some great help with various Microsoft MVPs, and the forums on msdn.microsoft.com are excellent.
You haven’t heard that 99% of software development rules are only treated as rules to be followed in La-la land?
It’s a noble sentiment, no doubt.
That being the case it isn’t even remotely true, look at the love 3.5 received, the maintenance releases for 4.0, etc. 4.1 has found a healthy balance between additions & fixing.
Edited 2008-07-13 21:21 UTC
Unfortunately, he’s completely wrong and has probably been getting screwed over with every .0 release of software he’s ever bought into. Beta testing of new versions is always done on early adopters (an unwritten and accepted rule), even when you live in a world where money is handed over and you expect to get a finished piece of software with all the features promised. In the open source world, it’s the only way things can work. People use a piece of open source software, and distros tend to ship it, when it’s good enough and the vast majority use it on that basis, and that’s the way the layered system tends to work.
For some reason, Apple did exactly this for about two or three revisions after OS X 10.0 (commercial software, at that), and with hindsight everyone seems to think that Apple’s approach is brilliant and there are no complaints. Go figure. All KDE did was admit up front that this is the way things are in the software world.
Edited 2008-07-13 21:34 UTC
Apple was heavily criticized for this approach in 10.0 and the free 10.1 release. 10.2 reduced the criticism significantly but it was not until 10.3 that the very loud criticism died out, and for good reason. People do not much bother complaining about 10.0 now because it is past and memories in computer culture are short.
By the same token once the KDE team releases a decent version of KDE4 (and 4.1 is not it) then over a bit of time the furor over 4.0 will die down simply for the fact there is no point in continuing to complain.
“.0 means its not ready for real world use yet” is plain not true, ESPECIALLY in the free software world. .0 means “Code and feature complete for a major iteration”.
It took other projects (WINE, Gaim, etc) YEARS to get to the point where they were willing to call themselves 1.0.
The reason they released 4.0 when they did is because they were horribly late and looking worse every month that rolled by (kinda like Vista and leopard). 4.0 was beta quality at best, not release quality code (kinda like Vista and Leopard).
Not only that, but people by and large plain do not like many of the changes made (kinda like Vista and Leopard). The fact that they went through anyways shows a serious problem in the feedback cycle of the projects process (kinda like Vista and leopard). Not only that, but now that 4.0 has been a raging catastrophe by pretty much any metric, the dev team is sticking their figures in their ears and telling each other that people are just wrong about 4.0, and that it actually was a success!.
What I don’t understand is this. According to the article yesterday about how you cant compare FLOSS to commercial operating systems, since one is bound by things like revenue and shareholders, and the other is primarily driven by hobbyist devs. So that being said, why the hell didn’t the KDE team just tell everyone to get off their backs until they had a proper release?
I find KDE 4 a raging disappointment for a whole bunch of reasons. GNOME has obviously lost all sense of vision, and nobody really knows what to do about it (other then tabbing every interface imaginable). KDE actually has boatloads of vision, but they have obviously lost all connection with their user base, and have become both politically motivated (if its not ready then dont bloody release it, no matter how bad you look), and downright arrogant (screw the users, we are going to build what we want).
What I am waiting for is a desktop team that has not lost its way to come up with something clever that will effectively destroy both existing projects.
Really? It seems to me that on the contrary most complaint are dealt with. When the first betas where released people complained about the new kick-off menu, and you can now use the old style menu. Then people complained about the desktop icons, and with the new folder view you’re getting it back. KDE developers show all signs of listening to their user base.
How has this anything to do with politics? You make it sound as there are internal fights for power or control. I’ve seen no sign of this.
The reason for the release has been clearly explained I think. 4.0 is not a full desktop, it’s a basic desktop with a guarantied stable binary API. It targets enthusiasts and developers, and until functional parity with KDE 3 is achieved the 3.5 branch is maintained, what more can you reasonably ask?
I see arrogance from some users demanding attention and their every single desire to be satisfied immediately, I don’t see it from the developers who have to make decisions based on technical and practical grounds. And yes, they sometimes have to say no to some.
Yes, sure, it’s always best to wait for others to do the work.
I’m talking more about adding features nobody really needs or wants (like plasmoids) before things that people actually DO need or want (like everything else in KDE 3.5)
That was the whole point of my origional comment, .0 does not mean “we have our APIs in place”. On a pre 1.0 app that would be a .5 thing, on a post 1.0 app that would be a milestone in the development branch. Or at least that is what it means for every other project except KDE. Retrofitting that reasoning is a plain old cop out.
Again, implementing things people do not need and/or want, and waiting on things they do need, and/or want.
“Write it yourself” is not an answer if you want to have any credibility at all telling someone else to use something. I do not make my living off of linux, I enjoy playing with it just like half a dozen other operating systems, because I am a geek and I enjoy learning about different operating systems. This is a hobby, one of several, and on the grand scheme of things not high on the list.
The comment I made was not a demand, it was a desire to see what I believe to be absolutely essential at this point.
That’s a bit contradictory since Plasma is one of the foundations of KDE4. If you want a functionally equivalent Kicker you must have Plasma first, you just can’t have it the other way around.
I’m sorry but for a library that’s precisely what it means. Now, okay KDE is a desktop and users expect it to be more than a set of libraries but that’s still how developers see it.
If all that users are asking is feature parity with 3.5 then there you go, use 3.5, where is the problem?
It’s not as if 3.5 wasn’t sold anymore.
It entirely depends on the critics you’re dealing with.
When the critic looks like an ultimatum, I say it’s a valid answer.
I still don’t understand the critic. If it all boils down to “KDE4 in its current state is immature, 4.0 was released to early.” Then so be it, just wait for a release that answers your expectations and in the mean time keep using 3.5.
I see this kind of comment quite often and I do not really understand it. The whole point of constructive criticism is to effect change. If everyone who has a problem with KDE 4.x or feels that the KDE team missed a feature that is critical for use stays quiet and uses KDE 3.5.x then how can we hope to effect the course of KDE4 positively?
Statements like “KDE4 blows, goodbye forever, I am going to use Gnome” really serve no purpose at best and is harmful at worst. This is not what the vast majority of criticism towards KDE4.x has been like though. The internet is a nasty caustic place so much of the criticism uses harsh wording but I believe the intent in most cases is in the right place.
Most are saying that feature X, Y, and Z are critical to their use and thus important to add to KDE4. A significant group of others are saying that new feature K is interfering with their ability to set up a comfortable desktop environment. This form of criticism gives the users a means to let the developers know what users feel they need. The developers do not have to listen, but I think it wise to at least give the trends some attention.
And in the KDE team’s defense they are listening and making changes they feel will address the comments in a planned and orderly manner. It very probably will not please everyone, but it is progress in the right direction.
I have issues with the way in which 4.0 was handled and choices made but I also have no illusions that I could have done better. In turn though KDE developers, especially the vocally defensive among them, might accept that the comments and criticisms are not meant as an insult personally (most of the time) but as a comment directed to hope for improvement.
So now it is basically “if you don’t like it, be quiet”? What is that going to help?
Can we agree that there is no other project under the sun where n+1 means LESS functionality? .0s usually suck, but they usually suck because of unintentional bugs, not because the goal was to suck.
Putting something out half done is symptomatic of a problem, especially when the only reason to do so is safe face. Pretending that everyone else is going crazy when they call you on that indicates that the problem is bigger.
I am not a KDE dev, but I have enough of this.
Now ? Yes. There has been enough of that.
No.
Each time a full refactoring is done, both FOSS and proprietary software follows that trend. In FOSS, I remember the cries when Gnome 2.0 was released, it was as stupid as all that vitriol is actually.
The problem is bigger, yes. After something like 50 years, human people still doesn’t know how to handle software projects. There is no answer.
I’ve stayed out of the KDE4 foodfight. But I can’t believe you’re trying to get away with this spin about the devs (well, one dev) listening to users.
Desktop icons back? You gotta be kidding. Only in small folder windows (vs. whole desktop), can’t persistently arrange them, etc. etc. Maybe by 4.2 enough people will have complained that sanity will overcome the ideology that users need to be forcibly prevented from “splattering icons” on the desktop.
An even better example is the hated toolbox cashew — users asking for the option to hide it is by far the 2nd top vote getting bug reported against plasma. Closed as WONTFIX because the developer thinks if he forces it down people’s throats long enough they will learn to accept it.
I use Qt to develop apps and KDE because it is/was a great environment. I’m happy that developers are exploring new UI concepts. But to have a whole year of a KDE release with non-configurable misfeatures and missing functionality depended on by most of the installed base is a failure of leadership and common sense.
From what I have heard, the default behavior in 4.2 will be for a single folder window to contain the entire desktop, effectively acting exactly like other desktops do. They simply didn’t get the functionality done in time for 4.1.
Ok, you might have a point there.
Still, you shouldn’t take 1 developer on one project to be the entire open source community – there are people of all types developing out there, and without sales and PR employees between them and the users, you get their opinions directly instead of filtered.
Yes. This complaint is common enough that I fully expect “cashew” to find use as a verb. For example:
“As a developer, be careful not to cashew your users.”
“I reported the problem to the mailing list but got cashewed.”
If this does occur, the KDE devs will almost have to take some sort of action regarding the problem, likely changing the icon to an almond, pistachio, or Brazil nut.
Next up on Groklaw: How I Learned to Stop Worrying and Love the Cashew.
Edited 2008-07-15 02:48 UTC
“The reason they released 4.0 when they did is because they were horribly late and looking worse every month that rolled by (kinda like Vista and leopard). 4.0 was beta quality at best, not release quality code (kinda like Vista and Leopard).”
Read and understand the article which was written to dispel such as the cr*p you have just written. He explains what 4.0 was all about. Seems like there are loads of people in this world that cannot read and comprehend.
I am a huge kde fanboy, and even I think releasing 4 was an mistake (not that it isn’t going to eventually become something more than it is). I think that anytime you have a team of developers telling users that they are wrong, over and over, or “explaining” why things were done (like release alpha software as a .0) you have a fundamental disconnect between the writers of the codebase and the users. This is not Apple OS9 versus OSX “I like the old way better” stuff, this is one codebase works (though may be a PITA to maintain) while the other, quite literally, does not kind of stuff – at least for most people I discuss this with. These people, BTW, are all linux and KDE enthusiasts.
I am personally amazed than anyone at all can look at the 4.0 “release” and say it was a good idea – regardless of the purpose, it has caused a lot of turmoil, and more than likely a lot of anguish within the KDE dev team and community. People who are using it now would have used it back then – developers and people who enjoy using forming technology. People who do not use is now obviously didn’t back then – people who like working in a stable environment, etc. No amount of spin doctoring can change the fact that the 4.0 release is, by and large, seen as a disaster, and this could have easily been changed by giving the code another year to mature. Refusing to “hear” this doesn’t change anything at all.
I’m crossing my fingers and really, really hoping that things change and we can all look back a few years from now and say “Hey, remember that whole shitaree with the 4.0 release? Sure glad it was all just a lot of smack talk over nothing” rather than “Hey, remember KDE?”
Releasing the stable libraries of KDE4 as 4.0 was OK, especially as they repeatedly told users that feature-parity is not achieved yet, and the user experience would be bad.
I am myself author of a small KDE application (kRaidMonitor), and I have not yet ported the app to KDE4. But think about it: Would ANYBODY port their smallish little helper apps as long as a release is not available? No. Definitely not. For me to port my app, it is absolutely necessary to have 4.x on my own desktop, how else should I port it?
I also do not think that you will find anyone telling you with a strait face that he believed 4.0 would be an excellent desktop ready for daily use, and he replaced 3.5 with it and then 4.0 crashed on him. In that regard nobody was mislead.
You always have to remember: KDE’s prime goal is not to have as many users as possible, the prime goal is to attract as many developers as possible. The reason for this is simple: KDE does not get payed with money, they get payed with code, hence coders are what they want to attract, because only they can make them rich.
You run the dev branch after they make the announcement that they have achieved a level of API stability.
This is why apple has the developer connection, and MS has MSDN, and every other open source project (other then KDE aparently) has a clear development plan with solid milestones that are properly communicated to the developers who use the platform.
Edited 2008-07-14 21:39 UTC
Ideally this would be the way to go, realistically application developers prefer to install devel packages from their distributor instead of building the platform themselves.
Since you are also referring to the Apple developer connection and MSDN, I am puzzled to learn that they have their application developers build their respective platform from source instead of providing readily installable or VM runnable images.
It’s always interesting to see how much one can learn in websites’ comments sections. Up until your info I believed that at least Microsoft shipped binaries to their MSDN subscribers and actually didn’t allow access to the sources at all.
Quite a revelation I’d say.
I never said that, or implied that. A consumer of a well documented API doesn’t need the source code to be able to use it. It is nice to have when documentation is bad, or the API is buggy and you need to work around or fix something, but in general any time it is required is a pretty big fail.
Agreed, but how is this related to what I wrote?
Your previous comment suggested that source code in a stable branch would be sufficient and then referred to Apple’s and Microsoft’s developer portals.
Therefore I can only assume that Apple and Microsoft also see it as sufficient to just provide access to the code, which is quite the opposite of what I thought to know about their developer products.
But maybe this is a misunderstanding and you were saying that Free Software application developers are so much more advanced that they actually prefer building the platform themselves instead of using pre-built binaries.
In which case I have to disagree, even they prefer pre-built development environments, since it does not require knowledge of all build options, dependencies, options dependecies, etc.
For example, many don’t say that releasing KDE4.0 was a mistake but that *naming* it KDE4.0 was a mistake, it should have been either KDE3.99 or KDE4.0alpha or KDE4.0Developer Edition.
Being coherent in the communication is nice, but it’s even better when your coherent with the ‘common’ usage of numbering.
And yes, this is important: this may have prevented the too fast uptake of KDE4 by distribution which was a mistake IMHO.
As for the rest, it all boils down that users prefer incremental change over ‘stop the word’ changes, something that KDE devs doesn’t want to provide..
Given that the developers are benevole, you cannot really criticize this part..
is that the KDE team was bored with 3.5.x and they couldn’t find anybody who was interested in just making things work, so they said WTF, lets just change it all. The rest is history.
You can dance around the “what the version numbering scheme means” nonsense all you want, but it all boils down to horribly managed engineering and communication. And it proves that both open and closed source projects have similar challenges when it comes to delivering overreaching projects and managing expectations.
I guess it is human nature to feel that when one does not like some aspect of a piece of software their view just has to be shared by everyone else. I see a lot of rhetoric about KDE4 in here that presume to speak for the majority. I beg to differ.
I am presently testing PC-BSD Alpha7 with KDE4.1 Beta and I must say I am starting to like it. My first impression was a bit negative but the more I use it the better I like it. I am also beginning to understand a lot of the choices that were made.
Now I will admit that I have encountered some problems, but it is, afterall, early in the release cycle. If you dislike it that much I would suggest you not use it. But I would also suggest using it for more than a few hours before you decide how you like it. Most people view change of any sort as bad. Even change for the better. Get over it and move on.
I keep hearing the complaints about KDE4. But what people sometimes fail to realize is that KDE4 is not plasma or a new k-menu. Most of the time and effort was put into rewriting things under the hood, the benefits of those are not directly tangible to the KDE user. These changes, however, matter to developers, so it should be the developers of KDE apps like amarok, ktorrent, koffice and k3b, who are voicing their opinions. If the application developers were the ones complaining about how much better KDE3.5 was I would say KDE4 was a mistake.
As a user I don’t care what has been done in kdelibs, all I want is a better word processor, or an improved web-browser. Things like phonon and solid mean nothing to me, but they probably will allow apps to do more for me than in the kde3 days, and at the end of the day, that is what I am looking for. We might complain that kde4 isn’t listening to the users, but for the most part they should be listening to the app developers, who in turn should be listening to the users. I believe that last part is happening, people complained about the kmenu and now we have over 3 different options to suit everyones needs. The speed in developing these alternatives, I belives, is a testament to how well KDE4 was setup.
So the applications are slow to appear on KDE4, which is why I still use KDE3 daily. Does that mean KDE4 was a mistake? Not really. All the applications are undergoing a major overhaul, and that will take some time. Once those apps are ready I will gladly switch.
Just my 2 cents.
I’m an Ubuntu user and I consider KDE 4 interesting enough, so I’ve installed openSUSE 11 on a second drive and I have to say that I like it a lot. It’s getting better by the day and it’s already another breakthrough in the world of the Linux desktop… Mac OS X went through a similar thing as well and look were it is now.
Agreed. Yes, it’s only been six months or so since KDE 4.0 arrived, whereas the Gnome 2.x series has been out for years. Six months isn’t long enough for everything in the new KDE to be in place yet.
Although I still mainly use Gnome at present, I do use a number of KDE 4.1 apps, and I’ve found them to be excellent.
Another iteration of KDE and the devs still don’t usable defaults and sane graphical layout.
Things like asking cookie questions and the horrible graphical quality of KDE apps are still the same.
I hear a lot about the power of KDE and I try it once in a while. I always get the feeling: this was not designed for humans.
See, this is funny. I am thinking the same about Gnome. But with one exception: I believe, that Gnome is not designed for people who think similar to ME.
It is easy for me to process a lot of Information at once, as long as the information is presented in a graphical way (diagrams, pictograms, tree views, you name it). I am slow at reading a breadcrumb view and finding out where in the filesystem tree the file will get stored.
Therefore I cannot effectively use Gnome applications, KDE is much better suited for me.
Lucky thing that we have both environments, and some more (XFCE, FVWM, …).
1. “KDE4 is finished”
Is anyone arguing this? I think everyone realizes that the software will improve in time the problem was that KDE released software with a label that is supposed to denote completeness and stability. A x.0 release should be feature complete and relatively stable. I think people understand that not all applications will have taken full advantage of all the new tech but that a x.0 release can be used day to day for most users except those that require the utmost stability and such users will wait for the x.1 or x.2 release before upgrading (or KDE 4.3 or 4.4). KDE is more than its libraries; it is a complete desktop environment that shipped with its GUI in a beta (possibly alpha) state and the project is taking hell and rightfully so.
Most people realize that KDE 4 will get better the problem is they released too early with a stable label.
2. “Releasing KDE 4.0 was a mistake”
People aren’t arguing that releasing KDE 4.0 was a mistake they are making the argument that it was released with the wrong label. The appropriate way to have handled the release would have been to enter full release mode and release “Release Candidates” up to the cut-off date and if the build was not ready for general use then they should have released but with a different label.
KDE 4.0 should have been released as KDE 4 Developer Release 1 and a new target should have been set for this July with a similar approach with a ramp up for a full release and if it didn’t pass muster again for general use then release it win another pre-release tag such as KDE 4 Developer Release 2.
The KDE devs could have released their DR noting which segments had achieved API & ABI stability and which hadn’t. This would have worked fine for most developers who would have had the same access to same builds as is currently available while giving the community time to finish up the basic parts of the desktop.
Is anyone outside of KDE third party apps developers trying to develop against KDE 4 as is anyway?
3. KDE needs a fork
This was just dumb and SJVN should have known better. Most of what he was requesting is going to be added to KDE 4 within the next year and he would have known that had he been a journalist instead of just another ranter.
4. “KDE needs to drop Plasma”
KDE does not need to drop Plasma they just released too early with a stable label. Plasma needed another 6 months to a year to get fully flushed out and instead they rushed it out of the door. Plasma needed to be released with a fall back mode that mimicked KDE 3.5 as closely as possible which will be possible by about KDE 4.2.
5. “Plasma lacks functionality”
This is temporary but once again this criticism is understandable because KDE released Series 4 with the wrong label.
6. “I cannot put files on my desktop”
You can put files on your desktop but it still works in an odd way. The old usage style should have been included and made the default at least for now. Yes I understand why that can’t be done right now which is why the software should not have been released with a stable label until basic functionality in the GUI was finished and working acceptably for non-developer use.
7. “The whole KDE4 desktop interface is radically new”
The underlying architecture is radically new but once it is finished it will be able to be as plain jane as anyone might want it to be or blinged the hell out if that’s what someone wanted. Once again a problem relating to the use of the wrong label.
8. “I am forced to use the kickoff menu”
When KDE 4.0 was released in January this was true. It is no longer. Once again it was released with the wrong label for software that was essentially alpha/beta quality.
9. “The KDE team does not listen to its users”
The KDE team does listen to their users. Many features have been bumped up the list implementation wise in order to please the users which is delaying the completion of Plasma which is why they should have just released as a DR in order to avoid this problem.
10. “KDE 4 vs 4.0 is confusing”
No it’s not confusing except for the fact that the KDE team released a desktop environment with an alpha/beta quality GUI. That is why tagging this release with a x.0 tag confuses people.
11. “KDE should just have ported KDE 3.5 to Qt 4 and not add all that other experimental stuff right away”
I agree with the author on this. A straight port was not a good idea. Plasma will be better but it was released with a stable release moniker too soon.
***
I think the author is misunderstanding many of us in that we are not arguing that they should not have released at all only that they should have released with a more accurate tag, such as KDE 4 Developer Release 1. Releasing a developer release that featured API/ABI stability for the libs and featured the proper build system would have accomplished everything that releasing with the inaccurate tagging of KDE 4.0 stable did but without the venomous reviews. The developers would have been left to develop in peace until they released a production ready (or near production ready) system.
Having tested KDE 4.1 Beta 2 I would say that it is almost there as far as a good stable release goes but even at that I think this too would be more accurately called a developer release albeit a mostly stable one. Plasma is the main hang-up now that KDE-PIM has been ported. I don’t think KDE should have received a stable release tag until Plasma was mostly functionally complete and stable. I don’t mean feature parity with KDE 3.5.x but just a stable interface with regards to panel resizing, desktop backgrounds, stable applets, better UI, etc.
Had the KDE devs waited to release later they could have released a mostly finished product to wide acclaim as it is people are just grumbling and waiting it out until KDE gets its act together or at least that seems to be sentiments from many in the blogosphere outside of die-hard KDE fans. From a PR standpoint this release was a mistake.
KDE will recover from this and most people will forget by the time that KDE 4.4 is released with all (or most) of the promised goodness of the KDE 4 series.