So, I seem to keep coming back to the same question with KDE (and, for that matter, many other FOSS projects). Which is... Why do so many of these projects often seem only ever to ALMOST reach their potential? In the KDE context, Konqueror stands out in my mind as a huge missed potential in this regard.

Before the flames ignite and the accusations and insults begin to fly on this one, let me point out that I've been a dedicated and loyal KDE/Konqueror user since KDE2. Steadfastly using Konqueror as my default browser, even still. Falling back to Firefox ONLY when absolutely necessary. Which is why I feel, not only qualified, but compelled to voice my Konqueror concerns.

Anybody who regularly uses Konqueror as their default browser would be less than fair in claiming that Konqueror even begins to fully meet the Web Browser needs of the end user. Something that I personally consider a huge and bitter disappointment. Konqueror has long been a centerpiece of KDE technology, serving as Web Browser, File Manager and universal viewing application. While it continues to serve admirably in the latter two roles, it suffers greatly as a Web Browser and, unfortunately, always has. Debates will likely continue over this indefinitely. Meanwhile, we find ourselves in similar situation as our Redmond user friends who, until only recently, were forced to abandon Internet Explorer for lack of features. Konqueror certainly doesn't lack features. Rather, what Konqueror seriously lacks is basic Web Browser compatibility and functionality. In it's current form, Konqueror simply can't begin to compete with Firefox, or even IE, on the Web Browser front. And if the years of second/third class citizenship here are any indicator, its designers/developers seem to have had no intention of correcting this. Standards compliance vs. usability may be the hindering debate here. Ironically, a clear recognition from Konqueror developers of this tremendous flaw is that they've even provided the (much needed) Konqueror context menu option to allow users to open any URL in Firefox instead of Konqueror, which I find myself using much more frequently than I feel I ever should. Never would be nice.

I frequently have to direct family and friends (ashamedly) away from Konqueror and towards Firefox, not because I want to or because I'm necessarily a great fan of Firefox, but because they're frustrated by the Web browsing experience provided through Konqueror. Put simply, too many sites just don't render properly with Konqueror. Users don't understand why, nor do they care. They just want it to work. Blaming the web sites for standards non-compliance is pointless and non-productive. Somehow Firefox and IE manage to get beyond this debate and I think it's high time for Konqueror do the same. As I pointed out earlier, I've been using Konqueror myself for years and still tend to prefer it to Firefox. But then, I'm willing to suffer its weaknesses and switch over to Firefox on the (increasingly frequent) occasion where it fails me. General users should never have to do that. I tend to like the fact that Konqueror is feature rich, is tightly integrated with the KDE Desktop and, as such, provides a highly cohesive user experience. I tend to hate that at least 10% (conservatively) of the sites I visit with Konqueror fail in one way or another. Hence my disappointment. Konqueror's developers should set aside the standards debate and, if necessary, re-engineer Konqueror's Web Browser to meet the browse-ability needs of today's Web. Konqueror remains a centrally featured component of KDE (Dolphin aside), much as IE is to Windows. This one usability issue alone stands out in my mind as a substantial road block to the broader acceptance of KDE on the desktop.

While I'm well embraced for the flames that may rage over my criticisms, I can tell you that I'll personally ignore anything other than productive dialogue on the issue.

And there you have it. Consider this less a critique than a call to action for the KDE/Konqueror development community. A simple plea for Browser equality, usability and dignity.

But when KHTML tries to respect every mistake it will be slower.
Maybe there should be a XML-mode like in Firefox: Normal ugly HTML-sites are parsed in an ugly way, proper XHTML is parsed using XML-technologies.

But there are two more important things:
-missing scripting-support
-slow KJS

I like Konqueror and I always make tries to use it (I'm using it right now) because of the KDE feel, but I work a lot with Google websites and Gmail or Google Docs not working properly with KHTML is a killer to me...
Even if I don't use Gmail there are a lot of "key" websites (e.g. Facebook) that make extensive use of AJAX and I think that has a lot of impact on users.

Arora and Rekonq also don't work well with Gmail, btw (infinite loop on login) so it's not a KHTML .vs. Webkit issue for me. I just think that support is lacking and it should be fixed. Whether it's with KHTML or Webkit doesn't matter, the real problem is the lack of support for "popular" sites that use heavy scripting.

Just my $0.02 anyway

Cheers

Would you please stop with the 'Develop this for me. kthxbai.' attitude? It does not contribute at all and there is enough awareness of these problems already.

I propose a short term 'solution' for all of the Firefox users to report bugs in your distributions asking for a much better integration when the KDE desktop is installed. (very much welcome would be program associations, at least)

Well, I followed today Lubos advice and installed (from sources) kdewebkit.

After few hours of testing:

- many smaller and bigger bugs including:
a) "Malformed URL" preventing entering to many sites (eg. GMail, BTW khtml 4.3svn handles GMail & Co very good)
b) Flash is lottery
- pure and low-to-mid JS HTML+CSS rendering in KHTML is just nicer, something about text layout, borders, gives it better look, also in many of such sites KHTML is faster
- QtWebKit has working transparencies
- QtWebKit has waaaay faster JS

So, dropping of KHTML for WebKit will not magically solve all problems. Just replace some of them.

Simple case in point - Konqueror doesn't even render the current KDE Home page (http://www.kde.org) properly. This is simply sad. Not to mention seriously disappointing. I'm sincerely hoping that someone from the KDE development community with influence and concern is willing to champion this cause and help restore some degree of vision and direction to the ailing web browser.

I shall add to this conversation because, whilst I'm a little too late to join the Konq fan club (having started with KDE at the 4.0 release) I feel I should make my own thoughts known;

I'm an avid Firefox user. Yes, you might have Adblock Plus, but do you have Save Session and Exit? Or a YouTube Video downloader?

I do.

I personally consider Konq's current interface to be a bad copy of Firefox's. I have seen little use for it due to its complex nature as a file manager, it's inabilty to work with 'Web 2.0' sites properly, and it's complete adhesision to a simple maxim I was taught in school; "A Jack of all trades is a Master of none."

Thus I feel there are three possible courses of action.

1. Konq. is completely depreicated and removed, replaced with a QT-Firefox, or something similar.
2. Konq. is partly removed but continues to serve as an inbuilt browser for Akregator and other software.
3. Konq. is overhauled and brought to Web 2.0 compleience with technologies.


The single reason I keep my parents computer off KDE is because of this mess. The inbuilt web browser cannot fulfil tasks properly (It's unable to handle flash-based bingo sites without crashing, nor can it fully display eBay pages properly) nor can Firefox sit natively in the system. (IE, the look; my parents are easily confused and given a shiny glossy arrow button, then a normal arrow, they will assume they do two diffrent things)

I, as an end user, wish to see 2 happen. Konq. had it's day, now it's time to retire and allow a more vast, more powerful tool take over. Until there if a QT-Firefox, I dont think that'll be any time soon, though.

Mind you, if 3. happens AND it comes out with many more useful addons, I'll switch to it.

Konqueror's web rendering capabilities are tied to the KPart it uses to render HTML. Work is underway on a Webkit based one in playground, however it is not known if or when it will be integrated.

Btw, even if you removed Konq, the inbuilt web browser portion would remain since that is KHTML and part of KDELibs.

fcwells59 wrote:So, I seem to keep coming back to the same question with KDE (and, for that matter, many other FOSS projects). Which is... Why do so many of these projects often seem only ever to ALMOST reach their potential? In the KDE context, Konqueror stands out in my mind as a huge missed potential in this regard.

I think the best way to go is to acknowledge the difficulty in rendering current web content, and to stop wasting time trying to re-invent the wheel with Konqueror, and instead invest the development time in integrating Firefox into the KDE desktop in a seamless way.

There are many more important things to be fixed in KDE than having Konqueror act as the web browser as well. Considering the time that would be needed to correct Konquerors failings in this respect, even if 5 new additional and extremely competent developers were to magically appear and start working full time on the problem, we would still be waiting well into next year, if not beyond for a properly functional browser.

I just cant see how the current team is going to be able to accomplish what the firefox team is able to, given the much larger landscape that the KDE team has.

StopTheFail wrote:

fcwells59 wrote:So, I seem to keep coming back to the same question with KDE (and, for that matter, many other FOSS projects). Which is... Why do so many of these projects often seem only ever to ALMOST reach their potential? In the KDE context, Konqueror stands out in my mind as a huge missed potential in this regard.

I think the best way to go is to acknowledge the difficulty in rendering current web content, and to stop wasting time trying to re-invent the wheel with Konqueror, and instead invest the development time in integrating Firefox into the KDE desktop in a seamless way.

There are many more important things to be fixed in KDE than having Konqueror act as the web browser as well. Considering the time that would be needed to correct Konquerors failings in this respect, even if 5 new additional and extremely competent developers were to magically appear and start working full time on the problem, we would still be waiting well into next year, if not beyond for a properly functional browser.

I just cant see how the current team is going to be able to accomplish what the firefox team is able to, given the much larger landscape that the KDE team has.

have a look here, there is a good new for you.
http://en.opensuse.org/KDE/FirefoxIntegration

Zayed wrote:have a look here, there is a good new for you.
http://en.opensuse.org/KDE/FirefoxIntegration

This is very good news indeed.

Thanks for the link Zayed

If I may venture an opinion, Konqueror in KDE4.x is certainly doing better than Konqueror in previous versions of KDE, though most certainly there is plenty of room for improvement. One of the key areas I think would be the addition of a javascript compiler (a la tracemonkey for Firefox, V8 for Chrome, etc.) as more and more websites are moving to heavier javascript (of which it is probably Google who is leading the charge; see http://www.chromeexperiments.com/). A start in this direction would be the union of KHTML and its ever-popular child/fork WebKit which I believe already has its own.

Of course integrating Firefox into KDE will also bring plenty of benefits as there are so many users who are already familiar with Firefox and use it full time.

I like konqueror and I'd like to use it more, especially since it seems faster or more lightweight than firefox. Unfortunately, for instance, right now, I'm using firefox because of gmail.

But I just wanted to say, please don't drop konqueror, improve it. I don't like the google-firefox business relationship and I don't like how prevalent firefox is. I'd rather there be a good alternative. Keep up the good work.

peterius wrote:I like konqueror and I'd like to use it more, especially since it seems faster or more lightweight than firefox. Unfortunately, for instance, right now, I'm using firefox because of gmail.

But I just wanted to say, please don't drop konqueror, improve it. I don't like the google-firefox business relationship and I don't like how prevalent firefox is. I'd rather there be a good alternative. Keep up the good work.

I'd like to see it improve as well. It has come a long way and it has a lot of potential. Like several have said, the javascript needs to be improved. Google is pushing it and xml/ajax technologies. For example, there are rendering issues at Youtube when using Konqueror. I noticed that it was javascript:/

So, the two prevailing themes here seem to either be - integrate Firefox more tightly with KDE or improve Konqueror. While I don't see why both can't be accomplished, my intent for this thread was not to pit one against the other so much as it was to draw attention to the serious Konqueror lag and, hopefully, pique the interest and concerns of those who are influential in the direction of Konqueror/KDE. I'd really hate to think hat Konqueror has become a lost cause.