OSNews

That's all well and good Jos, but neither dawg, nor I said the release ought to have been delayed.

We just said it should have been named a Developer Release or something to that effect. That's got nothing to do with delaying anything - it has to do with naming something according to what it actually is, and this could have prevented a whole lot of negative responses. "

renaming it to 'developer release' or something wouldn't have made much of a difference. Distro's would not have picked it up, or they would have - in the first case, the release would simply be another beta, and in the second case, ppl would complain anyway. And it wouldn't do much about the point about the parts of KDE which were ready, nor about the community dynamics.

We've communicated the state of KDE clearly, and the fact it's a .0.0 release should be signaling it's status anyway.


Re-read the article. It wasn't negative at all - try to tally up the negative versus the positive bits, and you'll see. The fact that everyone focused on the negative bits probably says more about the state of KDE 4.0.0 than it does about me. "
It wasn't all negative, for sure - maybe it was also the many comments and other articles out there that made me feel that way. But I did miss things like Parley and the many other great things in 4.0.0, really.

That's why I called it a "quick few impressions" and not a proper review - I am saving the proper review for 4.1.

hehe, ok, looking forward to that one, then.

"renaming it to 'developer release' or something wouldn't have made much of a difference. Distro's would not have picked it up, or they would have - in the first case, the release would simply be another beta, and in the second case, ppl would complain anyway. "

If it is DR lease, few people are going to use/test it, that make the KDE 4.0 get to mature slower.

"Many users will start using KDE 4.0.0 and start reporting bugs, so many corner issues the developers themselves would've NEVER found will be fixed in 4.1"

Ah, come on... All the visual issues have been reported many times to the KDE mailing lists, to blogs, and even here in the OSN comments.

"No good would have come of delaying the release any longer"

I read the whole article. Seriously, I'm not convinced. The normal user has nothing to do with inner development. How do commercial projects such as Leopard handle those development issues?

To answer your last question: they don't. They also release... Seen the many complaints about the latest Mac OS X? And don't even start looking for ppl who have trouble with Vista...

That's a rather peculiar argument... Vista/Leopard sucks balls, so KDE 4.0.0 may suck balls too?

I'm sorry, but that is the weakest argument I have heard ever since this debate started. We should not compare KDE to Vista when it makes KDE look bad, but we may do it when it makes KDE look good?

And about the reported visual issues, many of those have been fixed, many have not. As I said, there is a lot which can be fixed by KDE itself. But there also is a lot which we can't fix. And that stuff also has been reported. But not fixed. Now, more ppl will complain, or fee the urge to actually send in patches - and it will get fixed faster. That's one of the things we wanted to happen, for sure.

You are right about the naming thing. The KDE guys simply don't understand the richness of modern software versioning. "Alpha 1" and "Alpha 2" and "Beta 1" and "x.0 is final" are things the KDE guys have apparently never heard of.

Any other team would have called it KDE 4 Beta 1 or Developer Preview 1, and they would have avoided all this confusion and bickering.

To most people the name "KDE 4.0" simply misrepresents what it actually is. The KDE guys who think it is named well are probably in a minority.

Any other team would have called it KDE 4 Beta 1 or Developer Preview 1, and they would have avoided all this confusion and bickering.


The thing is that it's neither of those things... You could have made that argument with RC 1, but 4.0.0 is stable, and not terribly buggy. Is the desktop / taskbar underfeatured? 'yes'. But it's stable, and works. Do kwin effects expose hardware issues? 'yes'. But accumulating hacks for driver bugs for the next 5 plus years sounds terrible. The 'fix your drivers and we'll fix our bugs' approach is the only way to go. KDE 4.0.0 is what it is...

Being stable but underfeatured in this way is not suitable for a consumer product. I think you don't understand part of the KDE audience See: http://osnews.com/permalink?295864

Edited 2008-01-14 15:58 UTC

KDE 4.0 ist not a consumer product. Distributions are consumer products. Sane distributors will wait until it's mainstream-ready or handpick a few selected components (like openSUSE 10.3 already did).

Being stable but underfeatured in this way is not suitable for a consumer product. I think you don't understand part of the KDE audience See: http://osnews.com/permalink?295864


In short, stable and underfeatured, but usable for daily work is fine. Heck, I could argue that all of GNOME remains stable but underfeatured If you're going to nag, I'd nag about the need to use the big 3rd party KDE apps and KDE Pim from KDE 3... Hauling all of those libraries and icons and services into memory, with redundant technologies eating my CPU. Oh, woe is me that can't run a 'pure' qt / KDE 4 stack :p (Then again, apt-get remove 'KDE 3' does give a person something to look forward to.)

OK, so what about blackbox, fluxbox etc? Stable, but severely under featured. I don't hear people bagging them...

Dave

> We just said it should have been
> named a Developer Release

call me cynical, but releasing it as "4.0 - Developer Release" would've resulted in just a different sort of annoying feedback.

fact of the matter is that we're here to create a good set of technologies, we're on the right path and that's what matters. if people don't like, or even agree with, the necessary steps to get where we're going .. well .. so be it.

when i read your original article i though, "typical Thom" but i read a lot of positive things in it and a certain "i want to like this, but it's not there yet for me" tone. i'm cool with that ..

perhaps what is rankling is people for whom this release is not for getting pissy because it isn't. as they say, not everything is about you

I totally agree and that is exactly why I think it should be called a Developer Release. The technology is there and stable enough for developers to start learning, testing and experimenting with. The general path they want to take is laid out giving developers an idea of directions they can take KDE4 in. There are plenty of areas screaming out for some love from a developer.

Basically this is a release that has a whole lot offer to developers and not a lot to offer to general users.

Aaron,

As an OSS developer myself, I know how the avalanche of negative comments after every release can grind you down, you rarely get a pat on the back, and that when you do it makes all the difference.

So, for the record, I think you guys did absolutely the right thing. The 4.0 release is exactly what you said it would be, and in my opinion is completely consistent with modern versioning schemes.

Well done

No-one seems to quite get what he was saying. The entire idea was to NOT call it a developer release so that more people would try it, and find bugs. This way, all the bugs people find in 4.0 will NOT be in 4.1. If they called it developer release, it would not have gotten a quarter of the attention it has, you would not have done your review, and many of these bugs would not have been caught. So, these bugs would just end up in the 4.1 release (which is the 4.0 release had they called this a developer preview), and we would be in the exact same boat with people ranting about using bold fonts for, god forbid, HEADINGS.

As for it being negative, I think thats alot more useful for developers than positive is. If everyone just said "hey, its great" because they were afraid of offending developers, what would kde be then? I suppose it would be rather similar to the giant state of denial that skyos is in

Edited 2008-01-14 20:50 UTC

The reason KDE 4.0 and pretty much every other x.0.0 release of any software out there is a little rough around the edges or incomplete is because it requires user attention to be polished and users just don't want to use something named "Developer Preview" or "Release Candidate".

That's what OSNews does. We publish opinions, and if someone responds, even if it is "against" an editor, we post it. You know, open discussion, fair and balanced (that one never gets old ), that sort of stuff.

Edited 2008-01-14 15:14 UTC

Thom and the others want KDE to become like Enlightenment. Enlightenment 0.17 is in beta since ages. The final release won't get "ready", even though Yellow Dog Linux ships with it as default desktop.

"You won't drop or use it becouse of a number."

Do you really believe that the version of a software does not factor into the decision to use it?

I know some people who disagree with you. They don't use alpha and beta software. Those words "alpha" and "beta" mean something. They mean that the software isn't finished.

Back in the good old days you could actually use the version number as some sort of indication of how complete and stable the software was. A 0.x version or Alpha release was not feature complete, a Beta release was complete but probably had bugs, an RC was actually a condidate for release and an x.0 release was the first feature complete release considered to stable.

This system is good and it had developed and is used for a reason. As a developer, and a user, I don't expect anything less than an stable release to actually be stable. I know that if I pick a library which is still in Alpha, the API may change and cause me work in the future. I know that if I use that financial software that is in Beta it may crash and I risk the chance of losing data.

It seems the current trend is for version number inflation. Alphas have become "Betas", Betas are "Release Candidates" and Release Candidates are x.0 releases. Apparently even that is not enough now, and projects are skipping right ahead to "Release" without mucking about with all that stable APIs and bug testing nonsense. That's boring!

Releasing KDE as 4.0.0 and then claiming it is "a developer release" is just odd. There used to be a name for "developer releases". It was "Alpha".

Edited 2008-01-14 16:40 UTC

What do you mean with back in the good old days?
Kernel 2.6.0 days? KDE 2.0 days or even more back, like Windows 3.0 days?

I have the feeling that a lot of people have a concept on what an alpha, a beta or a final release should be. If a release does not fit into that concept they argue that the devs made a mistake.

It would be nice if more people would read and understand KDE's release concept here. This would solve a lot of the questions and fears posted here as well as the reports of lack of features et al.

So please let the KDE devs sound their own release concept and accept their concept for their product.

So, you will just suck up whatever the KDE developers say, without thinking for yourself? Without reflection, without critique, just because they are the developers? If they said KDE 4.0.0 was the best ever release ever and ever, and it would never be toppled by anything else in the whole wide world, would you accept it too, and "attack" [you're not literally attacking, but I don't know a better word a.t.m.] anyone not willing to accept that same position, like you are "attacking" people now who are not willing to accept the current situation as-is?

If the KDE devs want to promote an alpha release as a stable release, that's fine by me, go ahead, it's their responsibility, not mine. But don't go crying in the corner when people start complaining afterwards that your so-called stable product is, in fact, "nothing more" than a developer release.

[cliche alert] If you can't handle the heat, don't stand so close to the fire.

Edited 2008-01-14 17:29 UTC

If a commercial software editor (say an OS editor for example ), puts a big sticker saying "Warning ! This product is not a finished product, is possibly buggy, and may eat your babies" along with the rest of the software package in the shop, I know what I'm buying, even if the product is advertised as the new version of my favorite OS.

Edited 2008-01-14 18:29 UTC

> So, you will just suck up whatever the KDE
> developers say, without thinking for yourself?

i'm all for applying critical thought to things. i'm also for listening to what people have to say and when they say, "this is what i mean" not turning around and saying, "no that isn't what you meant" because that is, in a word, an *absurd* position to take.

i may not have understood what you meant, but once you explain what you mean i'd have to be an idiot to think i understood what you meant better than you did.

the problem is that there are people in this world who take advantage of that and when they get caught out say, "well, that isn't what i meant." due the abuse of this norm of communications by too many politicians, business people and outright crooks many people have become so jaded and cynical that they have lost the ability to ever recognize credibility even where it exists.

this is a bit off topic for this discussion, i know, but i think it's a sad reflection on our times and societies.

> If they said KDE 4.0.0 was the best ever release
> ever and ever, and it would never be toppled by
> anything else in the whole wide world, would you
> accept it too

c'mon, i expect more from you than this, and i think the fellow you replied to deserves a little more respect.

KDE didn't say it was the best ever release, we didn't even say this is what should be put into production installations or spread to the general consumer. in fact, *because* we didn't say that and instead stuck to what it is people can accept what we said. when someone is speaking plainly, it's alright to accept it as such. i admit one has to have the ability to tell the difference between straight speak and weasel words ...

No, in fact. But neither would I ignore what they say.
If someone tells me their product is not ready for everydays use and I try it nevertheless and encounter problems, well then I've been warned.

Not counting the problems that are probably connected to the distribution rather than KDE 4.0 itself.


You see that is the difference. They don't say it. Practically they say use at your own risk. Apparantly you are neither a distributor nor a developer of software for KDE you are rather a curious person like there are many.

But don't be disappointed if you find your parents preparing the presents for christmas instead of christ child/Father Christmas/...


None is crying but the ones that are not able to read:
It's not a release for Joe Public. Go ahead and write about 4.1 what you want as that is supposed to be the release for the general masses.


Then why did you install KDE 4.0 if you can't handle the heat?

Ah Thom, get off your bike before you fall off and hurt your behind!

Sure, KDE 4 has some issues, there are certain things that I've seen in screendumps that I don't like the look of. The new start menu doesn't make me too happy to try it at all. But - I don't go on bitching about it because of that. I'm smart enough to know that this was a technology release, a foundation, nothing more and nothing less. Why don't you just drop the bash KDE excuse and live with it? If you don't like KDE, don't use it. But don't go on endlessly bashing it because you don't like it, or the KDE developers fundamental design decisions.

OS X 10.0.0 was a shocker. Vista was, and is a shocker, and SP1 isn't expected to fix it. Vista was long overdue, and there was a lot of pressure on Microsoft to release something, anything, which is what they did. KDE got the underlying technologies pretty stable, and then did a release - so that distributions will take up kde 4.0.0 and it will get tested further by a wider audience.

When you really look at it, what is a quicker development method? 5% of your target audience using it and testing it and reporting issues for a 5 year development period until it's deemed stable, or 65% of your target audience using it and helping to make it stable in 18 months? I know what I'd prefer...sure, things will be not so nice for 18 months or so, but it's a start, and a solid start. You have to walk before you can run as the old saying goes.

Dave

Before this trend for release inflation began.



Well yes, that's because the terms "Alpha", "Beta", "Release Candidate" & "Final Release" have defined terms. Pretending that they don't doesn't help.



I have every respect for the KDE developers, but that doesn't mean I have to agree with their arguments on this matter. If they had released 4.0.0 as "Stable" and stuck to that, fair enough. What they have done is released it and then tried to tell people that it isn't stable. This to me is just back to front and is clearly confusing to a lot of people.

Edited 2008-01-14 18:02 UTC

> If they had released 4.0.0 as "Stable" and stuck to
> that, fair enough. What they have done is released
> it and then tried to tell people that it isn't
> stable.

we released 4.0.0 as stable and added advisory information so people would know what to and what not to expect from it. it's that simple and not particularly amazing.

i'm sure you'll have a lot of time to consider how that all came to change as you ride into town in your horse drawn buggy so that you can send a telegram and put a mail order in to Sears-Roebuck by post.

life is not static.

You're kidding me, right? We can just redefine accepted terms with however we like so that they fit how we've done things? Why did no one tell me this before?

I declare Syllable "stable" and that all "bugs" are imaginary. We have ten "million" users, although my definition of "million" may differ from yours. That's O.K, you'll catch up eventually to avoid confusion.

Sorry, you'll have to excuse the snarky post but this is the sort of thing that really gets my goat. We're being told that the version number isn't important but if it isn't important, why couldn't the release simply have been called an Alpha or Beta?

As you already said, alpha means still under development, meaning API changes in KDE, which is mostly a platform more than an application. KDE API is frozen, no alpha state applies anymore.

Beta is testing, and KDE 4.0 is stable, as it runs well. It may contain bugs, like any other x.0 release.

Seriously, take a look at all software projects out there, both open source and closed source, gratis and commercial. You'll see a lot of 1.0 releases not featurefull nor completely stable. Still nobody talks about them or cares.

The reality is KDE 4.0 as a framework is stable. It's not a complete desktop, as it's not intended to be yet. How could it be if most important applications are still under heavy porting? A desktop is nothing without applications, thus KDE 4.0 is merely a framework as of now.

And speaking of it, I'm running Kubuntu with KDE 4.0 right now and works. I've tried the soon to be KOffice and I like it, I use KTorrent 3 beta and works fairly well, KGet is amazing with its new interface, plasma *will rock*, Dolphin is what I wanted for simple and easy file management, and Gwenview sports a very interesting interface (best I've seen, actually). So it's not a full release ready to replace KDE 3.5, but it was needed for users to test it and report bugs, and for developers to catch up with the much needed ports (as already mentioned).

You look too much into numbers, as many people here, but fail to see what's important: KDE 4.0 had to be released when it achieved stability, not when all ports where completed, all features developed, etc. 4.0 is just a number, not really important, and if you can only see a number in such a big step in desktop environments, then shame on you, because you're missing the whole picture.

'Stable' does not mean the same as 'bug free'.

'Stable' generally refers to the stable nature of the API, and that bugs considered to be 'showstoppers' by the release team have been fixed.

It is doubtful that any complex application has ever, or will ever, be bug free.

There's a rule describing the amount of effort you should spend caring about the criticisms of random nameless people on the Internet...

Yes, but...

After 2+ years of development, the investment 30 minutes for a concise response is reasonable, IMHO. Most of the KDE developers have been 'lucid' about the state of KDE 4.0.0-- and even the KDE 'fans' have been largely the same. There's has been very little PR gloss over the 'pig' from the KDE camp. With that being said, letting 3rd parties define perception without having a response after all of the work that went into KDE 4.0.0 would be a little short sighted. Not for the people that use KDE 4.0.0, we know what it is (and isn't), but for the broader pool of potential users.

Indeed. As semi-official KDE promo dude, I feel I need to step in and make things clear, if I see ppl misrepresenting or simply not understanding things. We, as in KDE, don't do that enough. I must say aaron is really leading the way in changing that, though.

Interesting that you mention spin and Kubuntu since there seems to be a lot of spin trying to justify why the LTS release was dropped. As countless others have already mentioned: All other big players (Mandriva, Novell and RedHat) have no problem supporting KDE 3.5 for the next couple of years. The difference is that the other big companies have the resources to support KDE for many even if the upstream support lessens over time while Kubuntu only has Jonathan Riddell who is constantly occupied with cranking out another release while at the same time being the sole paid developer for supporting several releases at once.

Seriously, blaiming the premature release of KDE 4.0 for Kubuntu dropping a LTS release is probably trying to divert from the substandard financial support KDE is getting from Canonical.

Edited 2008-01-14 15:33 UTC

> probably trying to divert from

probably. there's nothing like ignoring 35% of your user base, especially when that user base includes entire school systems (canaries comes to mind; 300k people involved in that one), governments (french parliament) and businesses...

i continued to be a little shocked at the missteps of Canonical and *buntu in recent times.

For the 300th Billionth time. Just because Ubuntu is the most popular distro out here right now doesn't translate into cash. Ubuntu relies on money from One person at this time and he is under no obligation to fork over 10 million of his hard earned cash just to please everybody. They make do with what they have. Ubuntu isn't RedHat with a Time Warner deal under their belt, they aren't Suse with daddy Novell to throw them cash. I'd be surprised if, after the dell deal, they broke even.

KDE doesn't fit the criteria that Ubuntu needs, so it doesn't get the bulk of the support. Gnome does and Ubuntu has built its services and general timetable around Gnome. Like many have said (and I have reiterated countless times) KDE is not complete, why would a distro that is preparing to release an LTS release choose an incomplete unproven DE, and try to support them for 18 months. Ubuntu LTS releases stay frozen and they don't usually upgrade packages, they just focus on security updates, and minor updates. Would you really want to use KDE4.0.0 without any major updates for a year and a half or would you rather wait until KDE4.1 rolls around and is stable and mature, with all applications ported and running as they should. With everything completed, and the latest additions to the qt kit trickle their way down to KDE?

Personally i can wait a couple of months to get something usable than to have something now that isn;t really meant to be used by the average user.

This is quite true, however, by offering and marketing the LTS releases, they have basically entered the same playing field as the two distributions you mentioned, i.e. "enterprise" distributions, which guarantee their customers an actively maintained software stack over a certain extended period of time.

By dropping such an LTS release without prior warning they loose trust in this area, since the dropping of the LTS basically is a message to their customers that this part of their business model is not as important to them as others, e.g. OEM deals.


I am not sure how you measure completeness but pretty much everyone else on the planet would consider KDE 3.5 a proven software stack.
It is part of the LTS equivalent of other "enterprise" distributions and even used in custom distribution installations like Munich, Germany.

Whatever lead Canonical to leave their LTS customers in the cold rain of an uncertain future, it is pretty certainly not the quality of the KDE software stack.

It is more likely that they are leaving the traditional maintenance contract business model behind in favor of new ones and the dropping of the LTS option for part of their customers is the beginning of this phasing out process.

I was initially critical of the Kubuntu decision, assuming that they were taking an "Oooh! Shiny!" approach to their management of that branch of the distro, by going with KDE4 rather than taking the more sensible route of sticking to 3.x for one more release.

But then I heard that you guys were not committing to supporting 3.x for the duration of an LTS desktop release. Is that correct or incorrect? If it is correct, I don't see how they could reasonably, and in good faith, promise the things that an LTS release implies, regardless of KDE branch they picked. In that way, they were simply taking more responsibility regarding promises made to their user base than has the KDE project.

Or, in simpler language:

"I am rubber, you are glue. Whatever you say bounces off of me and sticks to you." :-)

Edited 2008-01-14 19:57 UTC

Well said - that is the real truth of it all. This is one reason why I do not like Ubuntu, and will not recommend it. If Canonical was serious about improving Linux, and the public's view of it, it would support KDE in the main release, and support it with the same money/number of developers that it throws to the Gnome desktop environment.

I see no reason why Ubuntu cannot package KDE and release a DVD instead of a CD, especially since Ubuntu will ship out CD/DVDs to you free of charge (great for us still stuck on dialup - thanks Ubuntu!).

Dave

The appeal and the reason I think Ubuntu is so popular is because they only have a one cd installer. It doesn't require 6 cd's full of crap to install anything, and they don't install everything under the sun, just what you need.

CD's are cheaper to make and ship to users, maybe not a lot cheaper, but when you are a company that doesn't have a lot of capital and no real revenue coming in other than donations from a certain billionaire, you are going to cut corner however you can.

KDE wasn't chosen as the default DE for reasons specified in every f--king KDE vs. Gnome post ever made in the last 2 years. If you think Kubuntu should be better then please by all means help out. They can use the help, but unless you can tell me how they are going to spend money they don't have on something that doesn't meet their criteria and was apssed over because of it, then you really have no say in the matter.

like I said before, Canonical doesn;t have the type of cash that RedHat or Novell does. They are not a publically traded company with investors and stock holders. They are a small company, last I heard they had only six people who actually work at their headquarters. The fact that they are able to do big things is only because of the communities involvement. Get involved, change Kubuntu's second tier standing, otherwise you have to live with the fact that resources are scarce and that they made their choice and are sticking to it.

"I don't quite understand what this article is trying to say. Most OSS projects are very careful with their reputation for quality, often staying at 0.9.x for longer than they really needed to."

The 0.9.x series generally culminates in a 1.0.0 release. KDE did exactly the same for KDE 1.0.0, IIRC. Projects generally behave the same up to 1.0.0: it usually represents the refinement of a well-tested existing code-base until is is declared "production-ready". Post-1.0.0, versioning schemes often work differently, as is the case with KDE and GNOME (see below), so I'm not sure why you bring this up in the context of a 4.0.0 release.

"Are we to understand that KDE is unique in that it is OK for them to push something out the door that they know is not release quality and call it going gold? "

No, in fact it is in the company of luminaries such as GNOME (GNOME 2.0.0) and the Linux kernel (2.6.0), both of which were first releases of *major* departures from the well-established GNOME 1.x.y and Linux 2.4.x codebase, and both of which thoroughly sucked. I'm sure that students of open source history can furnish us with other examples. Heck, we even see the same in the proprietary world: OS X 10.0 shared the same attributes as GNOME 2.0.0 and Linux 2.6.0 - big break; big breakage.

GNOME and KDE both use the same versioning scheme: up to 1.0.0 is the process of refinement up to production-readines; x.0.0 for x>1 is explicitly noted as the first step in a major *break* from the production ready code (more precisely, it is a break in the API, which generally results in significant code-base re-writes).

So personally, I'd turn your question on its head: is KDE so unique that its x.0.0 releases be held to a far higher standard than other core Free Software projects?

Edited 2008-01-14 16:03 UTC

I've been running Linux as my primary desktop and at my clients' sites for 11 years in my business. I've used both Gnome and KDE over various periods. I went through the kernel 2.0->2.2->2.4->2.6 cycle. I used KDE before it hit 1.0. Likewise with Gnome. I was using Gnome when it went 1.0. Likewise with KDE. I also used Gnome when it went 2.0. I am using KDE4 *right now*. And so it is with some grounding in history that I can say that this KDE release is seriously *substandard*.

P.S. I also went through the libc5->glibc6 transisition. And I have to admit that the current KDE breakage does not hold a *candle* to that experience!

Edited 2008-01-14 16:32 UTC

> I also went through the libc5->glibc6 transisition.

ugh. don't even remind me of those times. *shudder* and then that was then followed by the "years of weekly security issues in either bind, sendmail, apache, cron, ssh, etc or all of the above. have fun compiling from source.". it's pretty impressive that the community pulled through all that to have what we have now =)