TheBlackCat wrote:

StopTheFail wrote:There is an obvious streak of "don't look, don't see" in some people when the subject of KDE's weaknesses are concerned. I personally would like to see a more realistic view of KDE's current state of usability, and hopefully more reliability would follow.

Many in the Linux community like to point fingers at Microsoft Windows inadequacies, yet refuse to look at Linux with the same standards. I find this frustrating for several reasons.

Instead of continually shooting the messenger, it would be fantastic to take seriously the statements made by end users with a view of making the software appropriate for their use.

I haven't seen this at all. Quite the opposite, we have an entire forum, the brainstorm forum, designed specifically to point out inadequacies with KDE. The people on this forum, particularly the staff, are very involved and strongly encourage people to participate. Further, in the thread where people actually offer constructive complaints, instead of just "KDE 4 sucks", generally speaking several forum members go through point by point and explain the details of the issues, asking the person to submit bug reports for valid issues. The KDE developers, likewise, seem very open to the opinion of the community. They have implemented a lot of things they were not originally inclined to do because the community wanted it. This notably includes the classic launcher and per-desktop activities.

That statement I made was in response to einar, who said;
"StopTheFail, please remember the KDE Code of Conduct. This negative behavior is very detrimental to the discussion."
in response to my comment;
"Lets hope that isn't the standard that the Konqueror devs are happy to have met, and accordingly stop developing its capabilities any further."
I haven't said "KDE sucks" myself, and you will never find me say that. I will however draw comparisons between what people say with respect to various software projects and may highlight apologist statements from time to time.

TheBlackCat wrote:This idea that people on this forum or KDE or Linux developers ignore complaints and inadequacies of their software does not match with what I have seen at all.

While idea that KDE devs will sometimes belittle or minimise comments made by end users may seem alien to your experience, I can assure you it does happen, but thankfully not all that often.

TheBlackCat wrote:However, not every complaint is going to be fixed immediately. There are several issues at work. One is that distributions often add their own bugs, both in KDE and elsewhere. Users often come here blaming things on KDE that are clearly not KDE's fault, or things that were broken by distributions (either through modifications of the original KDE or through using unstable software where it shouldn't be used). KDE can't fix those.

Distributions are indeed responsible for some of the breakage end users see, and if I see a user blaming KDE for a distros failure, I'll likely point that out as well.
End users may note breakage in general conversation though without attributing blame, and I think these discussions can be productive to have on occasion.

TheBlackCat wrote: Second, developing software takes a lot of work, not every bug is going to be fixed instantly, not every feature is going to make it into the next release. But every user wants their pet bug fixed right now. That's simply not possible, there is a finite amount of time available. This is even more so for feature requests, which cannot be implemented until at least the next release, 6 months away. People have to learn to be a little bit patient.

While it would be nice to have every bug fixed instantly, this is obviously imposable. Users who bump up against these bugs may not be aware of previous discussions about said bugs, and may wish to open up dialogue about them in ignorance of previous discussion. It would be unfair to expect everyone to be up to speed on every aspect of the development background of the software they're using in oder to be allowed to comment on various aspects of said software, which may include discussions of undesirable behaviour. By the same token, I do acknowlege that respect for developers should generally be maintained. I don't think I've been unfair to the developers however as my statements were directed to comments made by end users.

TheBlackCat wrote: Finally, there are features that simply won't be implemented, either because they are infeasible or because developers have decided for one reason or another they are not a good idea. A lot of people will not accept this, they insist that KDE developers must do exactly what that particular user wants without question. The problem is that a single user is just that, a single user. KDE developers have to look out for the interests of the entire KDE community, and can't bend to the whim of every single user or else KDE would be a horrible mess (as many parts of KDE 3 ended up being). But certain people keep repeating the same thing over and over no matter how many times we point out that it isn't going to happen. These sorts of people feel entitled, for some reason, to have everyone roll over backwards to do exactly what they say when they say, and they get upset when that doesn't happen. That isn't how the system works.

Some users are unfairly demanding of the free software ecosystem. I'm not one of those. I've contributed code to both KDE, and the kernel, but by the same token I don't see this as a free pass to be unfairly critical of any dev team.

TheBlackCat wrote:There is a group of people who take this even further, with basically the mindset "KDE lacks feature xyz that I want, or hasn't fixed bug xyz, or made design decision xyz that I don't like, therefore KDE sucks and no one should use it." You could say it is the exact opposite of the mindset you are complaining about, but similarly not at all grounded in reality. People seem to not be able to understand that not everyone has the same experience that they have had, that not everyone wants the exact same things or uses the software in the exact same way, and therefore if they have problems the software is unusable by anyone.

I am on the record as saying that they should indeed not unnecessarily extend the feature set in an uncontrolled fashion without addressing stability issues in the core feature set. A diverse desktop with breakage everywhere is worse than a constrained desktop that is rock solid stable.

TheBlackCat wrote: I notice you were very quick to criticize someone who generalized their good experiences to the software being good for everyone (or at least that is how you interpreted, I interpret the statement very differently), yet you don't seem at all inclined to criticize the person that comment was made in response to, namely that because konqueror does not work for his or her use case therefore konqueror sucks for everyone. These two perspective are just as flawed, yet you only seem to have a problem with one of them.

I was pointing out the folly of minimising the assertion that "Konqueror sucks" (not my words) (but indicative of what people are saying out there)
It's sometimes important to focus not on the words used, but on what is being said. And what was being said was that Konqueror is not usable for a general use case web browser.

TheBlackCat wrote:

StopTheFail wrote:We have had the situation for a few years now where the Linux desktop looks achingly close to being usable by the general masses, but then never gets pushed the final mile needed to make this a reality. This "by developers, for developers" attitude has produced a desktop that is usable by developers and those more technical end users, but still cant be pushed out for everyone else to use.

While Linux has quite a decent userbase now, and doesn't necessarily need to become 10% or more of the landscape out there, it would still be nice to see it gain good market share.

I don't see this mentality much at all. It is a common complaint, but I have seen little evidence for it myself. On the contary, a lot of thought, time, effort, and work has gone into making linux in general and KDE in particular suitable for everyday users, and in increasing its market share. To ignore all this work, claiming it doesn't even exist, is pretty insulting to all the people who have spent the effort to try to figure this out. You must recall we are dealing with a competitor who has pretty much complete control over almost all software and hardware distribution mechanisms. Just making Linux better isn't going to be enough.

In general, the devs are doing a good job, and then you come up against problems such as the https://bugs.kde.org/show_bug.cgi?id=165044#c12 issue.
This isn't the first, and may not be the last "wontfix" that renders the KDE desktop unusable for some users. Luckily not a deal breaker for everyone however. The proposed solution is drop to a console, and fix with commands. Not an option for some out there.

TheBlackCat wrote:

StopTheFail wrote:Lets hope that with 4.4 that comes even closer to reality, but for it to, with the limited resources available, a sensible attitude needs to be taken to where and what to spend time on. Developing Konqueror to be a Firefox competitor doesn't seem to make sense to me, and seems to make more sense to integrate Firefox into KDE so that the developer time can be spent of more important things. My comments in this thread are motivated by that thought process. How that can be considered harmful to the debate, well I just don't get it.

The debate has been done, nothing you have said has been remotely news to anyone who has been watching KDE for more than a few months, certainly not for the konqueror developers. They are not suddenly going to see the light (from your perspective) now that the same arguments have been repeated for the thousandth time. You aren't going to change anyone's mind, so why keep going on and on and on and on about it? That isn't productive, it is just a waste of everyone's time. If you had something new to say, we would be happy to listen, but you don't.

I don't expect the Konqueror devs to take anything I've said about that with a grain of salt. They will do what they will do. I saying from my perspective that duplication of effort isn't the best way to go, but that's the one of the beauties of the free desktop. You can spend your time where you see fit.

TheBlackCat wrote:

StopTheFail wrote:But the context that his comment was made in suggests that he feels wtbennington's comment is not valid. If john_hudson felt that Konqueror was good for light-weight sites, but not for general use sites, then I doubt he would've responded in the way he did.

Once again, I disagree. I think the point of the comment was simply to point out that konqueror does not suck for everyone. This is a valid point, as I said earlier. Saying that because konqueror sucks for the web pages a certain user likes, that it must therefore suck for everyone, is just as flawed as what you are criticizing yet you seem to have no problem with that statement.

Well, I guess we disagree on what john_hudson was getting at then.

TheBlackCat wrote:

StopTheFail wrote:I do like the general direction of KDE and will shift back to using it as soon as it stablises, but I am surprised by the amount of people giving its current level of stability a "free pass" that they probably wouldn't give to other operating systems.

I am not sure it is necessarily a "free pass". Yes, KDE has problems that other operating systems and desktop environments do not have. It also has a ton of benefits compared to other operating systems. People can weigh the benefits and limitations of a set of software and see which comes out ahead for them. For software that lacks strong benefits for a certain user, that user will be less tolerant of flaws. If there are major benefits for certain users, they will be more likely to tolerate flaws. This is not some sort of blindness to flaws or ignoring flaws, it is simple economics. Everything has a cost and a benefit, you go with whatever has the biggest benefit-to-cost ratio for you. That doesn't mean you don't want to see the bugs fixed, don't report, them don't ask developers about them. It just means you report the bugs and get on with your life.

Indeed, and is the philosophy I follow, and is also not incompatible with my previous statements.

Like I've said. I like the ideas of KDE4, and will shift to it as soon as it stabilises. I keep a Kubuntu VM on my system and keep testing over time. I suspect 4.4 will be the release that I shift to. As soon as it's released, I'll be testing it.

Kryten2X4B wrote:

StopTheFail wrote:He wouldn't be using Konqueror for sites like the ones I mentioned, as it doesn't work with them.

I know there were just examples but still:

facebook - works for me, although it's possible that there a specific fb-apps that don't. I'm on facebook right now using Konqueror as a matter of fact.

Sorry, I ment myspace. Just tried it again with 4.3.2 and it locked up Konqueror.

Kryten2X4B wrote:youtube - Works. That fb-page mentioned is playing an embedded youtube-clip. It works as a "standalone" page too.
twitter . No problems as far as I can see.
slashdot - Seems to work.

With respect to slashdot, it certainly has performance issues, especially if you're viewing a thread with a few hundred comments. Will re-check its performance again to see if it's changed.

Kryten2X4B wrote:hulu - can't check since I'm not in the U.S.

Now, it's of course possible that there are certain places of those sites that do not work but I haven't run into any. Could you give a specific situation where Konqueror can't handle say Slashdot?

Otherwise I have to conclude that my Konqueror (version 4.3.2) is delusional and run those sites even though it can't...

StopTheFail wrote:If he meant sites that are very light-weight or Web 1.0, than yes, he certainly could use it for them, but the way his comment is framed, I don't think that's what he was getting that. In fact he seemed to be rebuffing wtbennington's comment that Konqueror wasn't much chop.

I thought he was saying it doesn't suck for everyone, and that it's unfair to claim it's worthless.

Well, I don't think there is software that sucks for absolutely everyone, but for your average Windows user for example, that you might like to migrate to Linux, well Konqueror doesn't seem to foot the bill just at the moment.

Kryten2X4B wrote:

StopTheFail wrote:John Hudson has just as much right as anyone else to comment on the state of KDE, and this right I myself will always uphold, but it doesn't necessarily follow that his comments are as valid or correct as anyone else's.

True, but even if they're factually wrong it doesn't follow that it's warranted to call it ridiculous. Instead, say _why_ you think he is wrong and under what circumstances.

And had the criticisms been directed at that comment in that way, I would've addressed that criticism. But fault was attributed to other statements I made.

To address that point directly, I would qualify what I said, so as to be clearer to exactly what I meant.

If he was saying something along the lines of "it's useful for somethings, but not for the general user", well then that goes without saying for just any piece of software and therefore doesn't bear comment.

If his comment meant something more along the lines of "what are you talking about, I've found Konqueror to be great for every website I visit" implying that the sites he visits are indicative of the general user, then I consider his statement folly.

Kryten2X4B wrote:

StopTheFail wrote:Are you suggesting that embedding Firefox or Chrome into kmail, etc. would be imposable, or just much harder than correcting Konquerors current abilities?

It's not impossible. It just hasn't been done. Well, with the exception of a webkit "plugin" that in my experience is rather buggy. Still, Chrome uses webkit so if the plugin was less buggy it would essentially mean that the Chrome-engine could be embedded in kmail if the user chooses to use it. And I guess you know that webkit is based upon code written for KDE?

Indeed I do.

anda_skoa wrote:

StopTheFail wrote:There is no point in restricting the internet to a tiny sub-set, and then boldly proclaiming, "Yay, Konqueror is a great, useful and functional web browser."

We don't know if john_hudson was artifically limiting the subset of websites. In fact his comment suggests that the subset he was commenting on includes all his web related activities.
If, by his experience, Konqueror is a great webbrowser, then there is nothing ridiculous in it.

But his comment, in context was in response to someone complaining of Konquerors inability to work with websites, and therefore, in presenting a counter argument, must disagree with the original premise somewhat, eh?
In isolation, his comment might make sense. But it wasn't made in isolation.

anda_skoa wrote:I currently can't come up with a site that does not work for me. Some do not work perfectly but I still think that Konqueror is a great browser because of the experience on all those sites where it does.

In my experience, the typical range of websites for your atypical Linux user is different that for your general population user. The very user that will need to be comfortable with the Linux desktop in oder to see its market share increase.

I will now switch from Firefox, back to Konqueror for my web browsing, and keep a log of my findings.

anda_skoa wrote:

StopTheFail wrote:I'm suggesting that Konqueror needs a lot of work to be considered viable to the general population, and that the effort would be better spent integrating an already great browser into the KDE desktop, and then spending the time saved on other more important things.

You obviously don't feel this way.

I don't because these are orthogonal concept, independent options.
As I wrote, it would be nice to have popular applications better integrated assuming the application's vendor supports that (Mozilla for example tries to avoid that to have the same "experience" everywhere).

But as I wrote as well, it is important to not lose focus and forget that the browser is just one application displaying web content.
People working "on Konqueror" from a layman's perspective are actually almost exclusively working on things that are not specific to Konqueror but benefit a whole range of current and future applications.

I can imagine that it is difficult to understand probably due to Firefox being such a stand-alone product, but most browser are actually just an application housing a web component used in all kinds of other applications and thus work on these components is necessary independent of other applications existing for a few of the use cases these components are deployed for.

As a programmer I do understand the issues at hand.

anda_skoa wrote:

StopTheFail wrote:"You mean ridiculing other posters with different experiences than yours, stating that they don't have any credibility anymore because their use cases don't match yours?"

I think it would've been more acurate to say something along the lines of
"You mean ridiculing other posters with different experiences than the general use case for a web browser."

I don't think there is a difference. The general use case for a web browser is to display and allow interaction with content served by a web server.
Which is obviously what the other user is successfully using Konqueror for. Might not match your experience but doe not invalidate his. Especially since he specifically mentioned that he has been using it successfully.

I think the general use case for a browser is to successfully render websites that your general garden variety user will visit, not just your typical technical user.

anda_skoa wrote:

StopTheFail wrote:Konqueror can already render the help pages, etc with enough success not to warrant much attention.

Help is just one of the use cases of a web render engine. People come up with new uses cases all the time, e.g. when Amarok started to offer embedded Wikipedia information in its context view.

StopTheFail wrote:From your comment I take it that the Konqueror devs are in fact not trying to expand its functionality to be able to be used as a general purpose browser.

Right. Because it already is being used as a general purpose browser.

Given the sites that I've found Konqueror to not sucessfully render, I wouldn't consider it a replacement for something like Firefox, IE, or Safari (even though Safari is Webkit based)

anda_skoa wrote:

StopTheFail wrote:This is good news as I think their time could be better spent elsewhere.

Exactly!
Which is why a lot of engineers from a lot of projects and companies are working together on software that can not only be used as a web browser.

Cheers,
_

Konqueror works for me. As a perfect filemanager, a not so perfect webbrowser (as I never got flash working reliable) but it's ok. For business environment the flaws can be handled by policy. I do not see it as a problem for the end user, as on windows use firefox desite IE.

But the root of all evil - and why the comments are harsh - is the dropped support of KDE3. KDE forces users & distos to use KDE4. Ok, that would work if KDE4 were mature. Anyway it isn't. (e.g. UTF-8-only is a nogo for business use.)

I know it's beating a dead horse, but if KDE had supported KDE3.5 for minimum one year longer then the situation would be much more relaxed and nobody would press / be pressed that way.

In my case I had to stop rollouts using KDE in july, since then it's just XFCE/Gnome (I absolutly dislike Gnome, but what should I do?). Sorry to say, but I don't know if there's a way back.

Anyway, I think KDE4 will be mature some day, just not in the near future.

StopTheFail wrote:Sorry, I ment myspace. Just tried it again with 4.3.2 and it locked up Konqueror.

Immediately? Or when you did something specific? It shouldn't lock up in either case of course, but just curious. I'm not a myspace user usually, so I was just browsing around a bit. I didn't manage to get it to lock up, but I did see one odd thing...it didn't affect the usage but looked odd. I for some reason got a really long horizontal scroll-bar even when there was no need to scroll...

StopTheFail wrote:With respect to slashdot, it certainly has performance issues, especially if you're viewing a thread with a few hundred comments. Will re-check its performance again to see if it's changed.

Okay, that does ring a bell. It seems to have improved in that regard though. Still, your "test" whether Konqueror is usable by a windows-convert (or non-technological user) is...well, a bit off when it comes to slashdot...somehow I don't see them caring one way or the other for that site. If it's a general problem it can of course manifest itself elsewhere as well.

StopTheFail wrote:Well, I don't think there is software that sucks for absolutely everyone, but for your average Windows user for example, that you might like to migrate to Linux, well Konqueror doesn't seem to foot the bill just at the moment.

It may, or it may not. The windows-users that has used Konquereor at my place didn't seem to have any problems in accessing the sites they wanted to. Still, it would be an interesting experiment to conduct with more participants.

StopTheFail wrote:If his comment meant something more along the lines of "what are you talking about, I've found Konqueror to be great for every website I visit" implying that the sites he visits are indicative of the general user, then I consider his statement folly.

But that goes for every browser. Even firefox can behave differently depending on a variety of factors, such as what OS is running it, what plugins are installed, or what fonts are available to name but a few. For example, just you try to use a site that uses shockwave extensively. No Linux browser can do it. Or the site looks slightly different on Linux than it does in Windows (usually not a deal-breaker but can be very annoying if the user is allergic to change) because the web-designer assumed font whatever would always be there.

Yes, there are sites that firefox runs better than Konqueror. On the other hand, I've found that there are more sites that work in Konqueror than sites that do not. They don't always work perfectly but they work.

That being said, I think concentrating efforts on integrating another browser into KDE is short-sighted. Why? Because it only solves the problem in one place: the browser. Updating khtml is likely to take longer but the results improves other aspects of KDE as well.

And in the meantime, there's nothing stopping the distros to KDE-ify firefox. openSUSE does that for example. The toolbar looks like Oxygen, the button-order is the KDE-way rather than the Gtk-way, and it uses the KDE file-selector. For more info, you can check http://en.opensuse.org/KDE/FirefoxIntegration . Yes, it's for suse only at this time but other distros can build upon it if they want to.

Kryten2X4B wrote:

StopTheFail wrote:Sorry, I ment myspace. Just tried it again with 4.3.2 and it locked up Konqueror.

Immediately? Or when you did something specific? It shouldn't lock up in either case of course, but just curious. I'm not a myspace user usually, so I was just browsing around a bit. I didn't manage to get it to lock up, but I did see one odd thing...it didn't affect the usage but looked odd. I for some reason got a really long horizontal scroll-bar even when there was no need to scroll...

I logged in, and tried to update by status, and it hard locked. No segfault or such, just a lockup.

Kryten2X4B wrote:

StopTheFail wrote:With respect to slashdot, it certainly has performance issues, especially if you're viewing a thread with a few hundred comments. Will re-check its performance again to see if it's changed.

Okay, that does ring a bell. It seems to have improved in that regard though. Still, your "test" whether Konqueror is usable by a windows-convert (or non-technological user) is...well, a bit off when it comes to slashdot...somehow I don't see them caring one way or the other for that site. If it's a general problem it can of course manifest itself elsewhere as well.

I agree that Slashdot isn't a good candidate as a test case for a garden variety Windows user, but more points to performance problems with dealing with a heavy "ajaxy" web site, of which there are more than a few.

Kryten2X4B wrote:

StopTheFail wrote:Well, I don't think there is software that sucks for absolutely everyone, but for your average Windows user for example, that you might like to migrate to Linux, well Konqueror doesn't seem to foot the bill just at the moment.

It may, or it may not. The windows-users that has used Konquereor at my place didn't seem to have any problems in accessing the sites they wanted to. Still, it would be an interesting experiment to conduct with more participants.

Your experience of this doesn't match mine, but I'm now keeping a log of misbehaviours that may be of some use to the devs.

Kryten2X4B wrote:

StopTheFail wrote:If his comment meant something more along the lines of "what are you talking about, I've found Konqueror to be great for every website I visit" implying that the sites he visits are indicative of the general user, then I consider his statement folly.

But that goes for every browser. Even firefox can behave differently depending on a variety of factors, such as what OS is running it, what plugins are installed, or what fonts are available to name but a few. For example, just you try to use a site that uses shockwave extensively. No Linux browser can do it. Or the site looks slightly different on Linux than it does in Windows (usually not a deal-breaker but can be very annoying if the user is allergic to change) because the web-designer assumed font whatever would always be there.

Yes, there are sites that firefox runs better than Konqueror. On the other hand, I've found that there are more sites that work in Konqueror than sites that do not. They don't always work perfectly but they work.

The main thrust behind my comments are that as a general rule, Konqueror is not appropriate for use by a garden variety user, who will be easily confused by misbehaviour of the browser, and unable to work around some things without resorting to another browser. In migrating a user across to Linux, things are difficult enough, to complicate things in one of the most used pieces of software infrastructure is counter-productive I find.
While it is definitely true there are sites Konqueror will render correctly that Firefox wont, it is most usually the other way around. And then there are examples such as nVidia.com, where due to the was Konqueror embeds flash, you loose the top line of menus. In Firefox, you do get the menus, but on the main page they still aren't rendered properly, but they are at least rendered. I beleive this relates to the wmode support for flash. This quote from the Adobe Linux blog may relate to the root cause of the problem.
Oh yeah, this beta also introduces some other features that may be of interest to some people. There's this thing called windowless mode, a.k.a. wmode, a.k.a. transparency, a.k.a. proper stacking order, a.k.a. DHTML/JS menus unroll over a SWF vs. under, such as on the main Adobe website:
Adobe Flash Player - Windowless mode
A.k.a. fullscreen Flash overlay ads! Yes! They're on Linux now. Let me tell you, I have never been so happy to see one as when I first got one to work in Linux. I suspect the novelty will wear off rather soon.
To be clear, there are 3 major modes for SWFs embedded in webpages: 1) Windowed mode, which is the only one supported by the Linux Flash Player up to this point; 2) windowless/transparent mode; 3) windowless/opaque mode. CommunityMX has this great, simple page that illustrates the differences between the 3 modes. (There are also 2 new wmodes introduced in Astro -- direct and gpu -- that are already supported in Linux where allowed by hardware.)
Thanks to both the Firefox and Opera teams for their support in implementing wmode. That's right-- this feature will work on both Firefox and Opera on Linux. For Firefox, you must have version 3. For Opera (as of this writing), you will need the latest beta of 9.50. Thanks also to Swfdec for implementing this feature concurrently, helping to attack wmode from many angles and bring Linux web browsing up to ~2002 standards.

Possibly Konqueror doesn't support these modes? I'll be looking into that aspect a bit closer tonight.

Kryten2X4B wrote:That being said, I think concentrating efforts on integrating another browser into KDE is short-sighted. Why? Because it only solves the problem in one place: the browser. Updating khtml is likely to take longer but the results improves other aspects of KDE as well.

And in the meantime, there's nothing stopping the distros to KDE-ify firefox. openSUSE does that for example. The toolbar looks like Oxygen, the button-order is the KDE-way rather than the Gtk-way, and it uses the KDE file-selector. For more info, you can check http://en.opensuse.org/KDE/FirefoxIntegration . Yes, it's for suse only at this time but other distros can build upon it if they want to.

I do mention that exploring the technicalities with embeding Firefox or such in a way as to provide it as the internal web renderer would be a worthwile persuit in order to address the many places HTML rendering needs to take place in a desktop.

I was made aware of Suse's KDEafication of Firefox by Zayed, and I must say it was good news. I'll be integrating those changes to any KDE desktop I run in the near future.

I've just checked Konquerors behaviour with websites that utilise the various wmodes for embedded content and it would seem that they're not supported as they are in Firefox.

See http://www.communitymx.com/content/sour ... opaque.htm

Visit the site with both Konqueror and Firefox to see the difference.

samhain wrote:But the root of all evil - and why the comments are harsh - is the dropped support of KDE3. KDE forces users & distos to use KDE4. Ok, that would work if KDE4 were mature. Anyway it isn't. (e.g. UTF-8-only is a nogo for business use.)

No, as we keep explaining this was not KDE's decision. It was the decision of the distributions, and only their decision, to stop supporting KDE 3.5. The KDE leadership cannot force anyone to stop using KDE 3.5, least of all distributions. It is simply impossible, even in principle, for anyone to stop anyone from using an open-source software set. And the KDE developers did not even encourage distributions to abandon KDE 3.5, in fact they actively discouraged it but distributions didn't listen. And a bunch of distributions, like openSUSE, still support KDE 3.5 in one way or another. If you really want to use it it is still easy if you pick the right distribution.

> No, as we keep explaining this was not KDE's decision

It was not a KDE decision to stop work on KDE 3.5? How is it then that the packages on ftp://ftp.kde.org/pub/kde/stable/3.5.10/src date to oct. 8th 2008? debian testing used to support KDE3.5 til may 2009, stable is still kind-of supporting it (fixing), but as there's no upstream any more, what should distributions do? even on fvwm ther's still an upstream ..

> And the KDE developers did not even encourage distributions to abandon KDE 3.5

And who is running around and telling everybody KDE 4.3.2 is secure, stable and bugfree? I cite from http://kde.org/info/4.3.2.php:

BUGS: [...] None known currently

That's merly a joke, isn't it?

samhain wrote:I know it's beating a dead horse, but if KDE had supported KDE3.5 for minimum one year longer then the situation would be much more relaxed and nobody would press / be pressed that way.

I know it kind of feels like support for KDE3 was discontinued early, but if you have a look at the announce timeline http://www.kde.org/announcements/ you can see that KDE3 has been actively developed for six years, its 3.5 branch for three.

Thle latter is about as long as Canonical gives you with its LTS releases (on desktop).

Speaking of LTS: several distributors and even more service providers continue to support KDE3 because of their long term policies.
As of this day the most recent commit to the 3.5 branch is 8 days old.

Cheers,
_

StopTheFail wrote:I logged in, and tried to update by status, and it hard locked. No segfault or such, just a lockup.

I went to the trouble of setting up a myspace account (tied to my spam-account since I have no intention to keep on using it), and I could do everything from creating the account to verify the e-mail-addy even though , hotmail as usual thinks I should upgrade the browser...but it worked, as well as updating my status. No problems. In short, I couldn't duplicate the hard lock. I'm not saying the problem isn't there, just that I couldn't trigger it.

StopTheFail wrote:I agree that Slashdot isn't a good candidate as a test case for a garden variety Windows user, but more points to performance problems with dealing with a heavy "ajaxy" web site, of which there are more than a few.

Absolutely. However, since slashdot now works better than it used to (at least in my experience) hopefully whatever was made to make it run faster will boost the performance on other "ajaxy" sites as well.

StopTheFail wrote:Your experience of this doesn't match mine, but I'm now keeping a log of misbehaviours that may be of some use to the devs.

Sounds like a good idea. I have no idea of why Konqueror works better for me than it does for you, but hopefully its flaws can be remedied.

StopTheFail wrote:Possibly Konqueror doesn't support these modes? I'll be looking into that aspect a bit closer tonight.

I could easily replicate the problem on the site you mentioned in the other post. I tried it in Arora as well, and the problem is evident there as well. Of the three browsers I have installed, firefox is the only one that managed to do it.

StopTheFail wrote:I do mention that exploring the technicalities with embeding Firefox or such in a way as to provide it as the internal web renderer would be a worthwile persuit in order to address the many places HTML rendering needs to take place in a desktop.

I'm not sure that's the best course of action. If I'm not mistaken that would entail ripping out gecko from firefox, and re-implement it in a kpart. Possible, yes, but may be more trouble than it's worth. Making sure to either improve khtml or making the webkit kpart more solid seems like it could yield results sooner. And webkit should be more than good enough (personally, I think webkit is a lot better than gecko).

From what I've gathered, Konquerors problems mainly fall into two categories (with exceptions, such as the flash-problem you highhlighted):

1. Deficiencies in KHTML. When that is the problem, two things can be done. Either fix khtml, or using the webkit kpart. Even though the latter is buggy right now, it can provide better compatibility. It does on gmail.com for example. On other sites, it doesn't - probably due to bugs. livejournal.com falls into this category. It looks awful in konqueror no matter what (can be used though), but rendered correctly in Arora.
2. Sites that look at the browser-agent-string and don't know what to do when they come across Konqueror. It doesn't always work, but sometimes spoofing as another browser helps at least somewhat. Shouldn't be necessary, but stupid web-developers is not something that Konqueror can do anything about. Still, I've noticed that when webkit is used this problem is less of a problem. Probably because the site thinks it's Safari that's coming for a visit.

That's why I think the best approach to fix Konqueror's problem is to make sure the webkit kpart is up to par and available. Or the changes made in webkit are merged back into khtml but I have no idea of how feasible that is.

Kryten2X4B wrote:

StopTheFail wrote:I logged in, and tried to update by status, and it hard locked. No segfault or such, just a lockup.

I went to the trouble of setting up a myspace account (tied to my spam-account since I have no intention to keep on using it), and I could do everything from creating the account to verify the e-mail-addy even though , hotmail as usual thinks I should upgrade the browser...but it worked, as well as updating my status. No problems. In short, I couldn't duplicate the hard lock. I'm not saying the problem isn't there, just that I couldn't trigger it.

When updating status and mood, it didn't locking up. Must of been "one of those things", however the behaviour of the text entry field is wrong, but can be worked around.

Kryten2X4B wrote:

StopTheFail wrote:I agree that Slashdot isn't a good candidate as a test case for a garden variety Windows user, but more points to performance problems with dealing with a heavy "ajaxy" web site, of which there are more than a few.

Absolutely. However, since slashdot now works better than it used to (at least in my experience) hopefully whatever was made to make it run faster will boost the performance on other "ajaxy" sites as well.

Well, it may "work" with slashdot, and it might be faster than it was, but on modest hardware I'm sure users will be glad if when Webkit comes along , a faster javascript engine comes along for the ride.

On a related ajax note, Google Wave doesn't work, and haven't been able to get google docs to work as well. though.

Kryten2X4B wrote:

StopTheFail wrote:Your experience of this doesn't match mine, but I'm now keeping a log of misbehaviours that may be of some use to the devs.

Sounds like a good idea. I have no idea of why Konqueror works better for me than it does for you, but hopefully its flaws can be remedied.

Are you using a 32bit installation? That might explain some of the differences in user experience.

When using a 64bit install of Kubuntu running KDE 4.3.2, and I visit the KDE Brainstorm, Konqueror consistantly segfaults after about 20 seconds. Doesn't do this with a 32bit install. Haven't tries opensuse yet but will in the next few days. Rebuilt a clean 64bit VM and it's still doing it.

Kryten2X4B wrote:

StopTheFail wrote:Possibly Konqueror doesn't support these modes? I'll be looking into that aspect a bit closer tonight.

I could easily replicate the problem on the site you mentioned in the other post. I tried it in Arora as well, and the problem is evident there as well. Of the three browsers I have installed, firefox is the only one that managed to do it.

StopTheFail wrote:I do mention that exploring the technicalities with embeding Firefox or such in a way as to provide it as the internal web renderer would be a worthwile persuit in order to address the many places HTML rendering needs to take place in a desktop.

I'm not sure that's the best course of action. If I'm not mistaken that would entail ripping out gecko from firefox, and re-implement it in a kpart. Possible, yes, but may be more trouble than it's worth. Making sure to either improve khtml or making the webkit kpart more solid seems like it could yield results sooner. And webkit should be more than good enough (personally, I think webkit is a lot better than gecko).

From what I've gathered, Konquerors problems mainly fall into two categories (with exceptions, such as the flash-problem you highhlighted):

1. Deficiencies in KHTML. When that is the problem, two things can be done. Either fix khtml, or using the webkit kpart. Even though the latter is buggy right now, it can provide better compatibility. It does on gmail.com for example. On other sites, it doesn't - probably due to bugs. livejournal.com falls into this category. It looks awful in konqueror no matter what (can be used though), but rendered correctly in Arora.
2. Sites that look at the browser-agent-string and don't know what to do when they come across Konqueror. It doesn't always work, but sometimes spoofing as another browser helps at least somewhat. Shouldn't be necessary, but stupid web-developers is not something that Konqueror can do anything about. Still, I've noticed that when webkit is used this problem is less of a problem. Probably because the site thinks it's Safari that's coming for a visit.

That's why I think the best approach to fix Konqueror's problem is to make sure the webkit kpart is up to par and available. Or the changes made in webkit are merged back into khtml but I have no idea of how feasible that is.

I'm sure the effort in maintaining KHTML would outway the initial pain of integrating gecko, but as we all know, this is not going to happen. When the webkit kpart stablises, we'll just have to be satisfied with what that brings to the table. For general browsing I still maintain that Firefox is the only viable option at the moment, but of course YMMV.

Also, hopefully the issues with Flash will be addressed, along with the problem where under certain circumstances, animated GIFs don't animate, and sometimes the UI seems to block while resolving URLs.

StopTheFail wrote:Well, it may "work" with slashdot, and it might be faster than it was, but on modest hardware I'm sure users will be glad if when Webkit comes along , a faster javascript engine comes along for the ride.

I wouldn't mind either Speedups are always welcome, after all.

StopTheFail wrote:Are you using a 32bit installation? That might explain some of the differences in user experience.

Actually, no. I'm on a 64-bit installation although with some 32-bit libs (don't remember why I needed those for any longer though...it wasn't related to KDE though. It probably was to make some proprietary app to find the libs it needed).

StopTheFail wrote:I'm sure the effort in maintaining KHTML would outway the initial pain of integrating gecko, but as we all know, this is not going to happen.

It most likely won't, no. But I do think you're underestimating the effort needed to integrate gecko. I seem to recall there are gtk-dependencies for just gecko - that is, without the UI. I could remember wrong though, but it would explain why the QFirefox or whatever you want to call it seems to have been shelved once again.

StopTheFail wrote:For general browsing I still maintain that Firefox is the only viable option at the moment, but of course YMMV.

Only viable one or not, I'm of the opinion that all browswers suck to various degrees. Firefox just sucks slightly less than the others. And it suffers from bloat.

Kryten2X4B wrote:Actually, no. I'm on a 64-bit installation although with some 32-bit libs (don't remember why I needed those for any longer though...it wasn't related to KDE though. It probably was to make some proprietary app to find the libs it needed).

Might be related to differences between openSuse and Kubuntu. Will be testing openSuse in the next few days.

Kryten2X4B wrote:

StopTheFail wrote:I'm sure the effort in maintaining KHTML would outway the initial pain of integrating gecko, but as we all know, this is not going to happen.

It most likely won't, no. But I do think you're underestimating the effort needed to integrate gecko. I seem to recall there are gtk-dependencies for just gecko - that is, without the UI. I could remember wrong though, but it would explain why the QFirefox or whatever you want to call it seems to have been shelved once again.

I'm sure over time, that gecko integration would pay huge dividends, vastly outweighing the initial pain. For the cost of integrating someone else's code, you get free access to correct web site rendering which is maintained by another team, and probably more importantly, a HTML rendering target that web site developers code to and test against.

Kryten2X4B wrote:

StopTheFail wrote:For general browsing I still maintain that Firefox is the only viable option at the moment, but of course YMMV.

Only viable one or not, I'm of the opinion that all browswers suck to various degrees. Firefox just sucks slightly less than the others. And it suffers from bloat.

And this is the problem. While most software suffers at least to some degree with suckyness, there are quality and usability standards that people have come to expect, and Konqueror as yet doesn't meet them for the vast majority. I have some opportunity coming up to slip Konqueror under the noses of some more non-technical users to gauge their reaction. There will be no prejudiced comments prior, as I want to hear untainted reaction, and will be interesting to see what they have to say.
Just in the last couple of days back with Konqueror, I find it exhibits the odd minor usability issues. Some may say that's driven by being used to other software patterns. I don't think do at this stage.
Things like the hoops you need to go through just to get it to recognise extra buttons on a mouse for click to go back and forward, the "highlight everything" in a text field when you press and hold, begin selecting text and then move the mouse out of the text field before releasing the button (more an issue with multi-line text fields) , to the weird behaviour when scrolling over tabs.