KHTML Developer Allan Sandfeld announced
that KDE’s browser Konqueror now passes the ACID2 test from webstandards.org.
While some patches were taken from the Safari fixes, a good part of the fixes have been written from scratch. Konqueror is the second browser to actually tackle the hard test.
hey mozilla-foundation, where are you with it? i’m waiting!
Konqueror is really improving at an amazing speed.
Hope we’ll see it in KDE 4 then! Chances are good I think..
But why was the “Don’t just stand there” post pulled? Didn’t see anything particularly wrong with it…
Great, and again: Use the force – opensource!
Really great news. Kudos to the khtml (well, all kde) hackers!
Open KHTML Info Page Launched: http://dot.kde.org/1117898224/
It’ll be in 3.5, which is due in september I think.
The acid2 test tests obscure CSS that isn’t likely to be used, passing it, while good is also quite pointless. why? because due to IE we can’t even use the non obscure CSS that would make development so much easier + allow for better designs.
It would be great if webcore could take off as fully cross platform (mac,win,gtk, qt as khtml) as it would provide a nice alternative to gecko.
I love Linux, KDE, and Konqueror a lot, I’ve been using them for years but I really miss the adblock, flashblock, etc… plugins when I’m using Konqueror, and people have asked for these features for a long time without getting them. If the KHTML developers would make it easy to write plugins with something like Javascript similarly to the way XUL plugins for Gecko are made then surely equivalent plugins would soon follow and I would be very happy.
it’s not pointless if it tests against standards, and I certaintly hope you mean that you hope that khtml, not webcore, would go cross plattform?
but I really miss the adblock
It’s already implemented in SVN, thanks to the work by Ivor Hewitt. It’ll be in kde 3.5!
Apple’s Safari was first to pass it, but it isn’t released to the public as far as I know. It’s still in their development system. Or did it get released with Tiger? I seem to recall that it broke some things, and therefore was going to be a while before it made it out to the public.
Assuming this is the case, the Konqueror could actually come in and be the first browser publicly available that passes the ACID2 test. If they do, then I hope they make a lot more noise about it.
[I don’t use KDE or Konqueror, but when anyone does a good job on something, they deserve the proper respect. And kudos.]
This may also make its way into the next bugfix release, KDE 3.4.2, so there might not be a need to wait till KDE 3.5 to get this fixes.
Wasn’t iCab the second to pass the ACID2 test? Or it just uses the WebCore too? (if so, the CVS WebCore… as the Tiger don’t have these fixes yet…).
And talking about KHTML… that’s good news!
Does anyone knows if Firefox will (or have plans to…) pass the test for 1.1 final?
Do you really think that KDE is copletely under GPL? You are either uninformed or a troll.
I think you are a troll who tries to say “If KDE was under BSD license, Apple had not given any modifications back.”
Some parts of KDE are under GPL, others under BSD license. KHTML is under neither of these licenses. It’s under LGPL.
BTW: Apples OSS projects are under a free software license that’s simmilar to the LGPL.
http://www.gnu.org/philosophy/apsl.html
Actually, this test’s about the future… what webdevs will be able to use in the future… Because no new tecnology can be used mainstream if it’s not well supported… IF IE7 pass the test, there’s good chances that we’ll be able to use then in some years… since IE4, it supports DHTML, but it’s just begin used for webapps now… we can expect about the same timeframe, probably not so long (as internet is more commom now and software updates are too) to use these features… If not, or you can say that’s everyone’s fault not updating their browsers…………
We are very lucky that KDE is under the GPL and not the BSD license.
That’s you point of view… IMHO it’s too sad they’re using this license instead of BSD ou MIT… But hey… everyone has it’s own opinion… (talk to all BSD guys… and talk with almost everyone how ships costumer eletronic apps where all this work could be used by lot’s of people around the world…)
I’m still waiting for the port of khtml outside of apple and KDE . Where’s ma’ GTK2 khtml browser? That’d be quite cool, especially on this old box that’s far too slow with gecko (I’m using Dillo).
My point is that unless IE supports it, it is pointless. IE is not going to lose market share in a hurry, if at all, so the use case if very limited. Standards support is great, but until developers can use them it remains something that we can only ever long to be able to use.
There is no way in hell IE7 will pass the test from what MS have been saying there will not be any real improvment in support for what they call a flawed standard.
As for being the future. Acid2 is testing standards from 1997 in the case of object support, and 1998 for CSS 2. This is 7-8 years later and we still can’t use everything in the spec, even in great browsers like konqueror, opera, firefox, safari etc.
As for webcore / khtml going cross platform, I mean webcore as it abstracts away the QT reliance, hence there is gtk-webcore which is done by a couple of nokia employees.
Nice stuff, there will be a KDE 3.5 before KDE 4?
Dave Hyatt says they just released 2.0 and Opera, Opera 8.0. Asfor Firefox, they are on the 1.1 and the President of Moz-Europe said Acid arrived too late.
http://webstandards.org/buzz/archive/2005_04.html#a000519
For French speakers
http://standblog.org/blog/2005/04/29/93114143-en-vrac
I am using Mac OS X Tiger 10.4.1 and Safari does NOT render the page correctly.
So how come they say it is the first browser to pass the test?
Yes, iCab was the second and iCab uses its own rendering engine, not WebCore or something.
Konqueror is either the third (if you count the not-yet-public releases of Safari and iCab) or first browser (if you only count browsers the public can actually download). It certainly wasn’t the second browser 😉
(1+3)/2=2
See? =)
It would be great if webcore could take off as fully cross platform (mac,win,gtk, qt as khtml) as it would provide a nice alternative to gecko.
That was never very likely, and it certainly doesn’t help that Apple have made it dependent on the Cocoa API in addition to QT.
Nokia has done a GTK2 port of WebCore and this has been used to make a Gnome KHTML browser. I think it was done by OGALAXY of GonMe(sp) fame. Do a search for it, it has been mentioned several times on this site.
Actually, so far no browsers pass the test. Sure, Apple’s unreleased Safari changes pass it, according to David Hyatt, but it’s not publically available, so this does not count.
Similarly for iCab, their last official release does not pass it.. all we have is a screenshot – same deal as Safari. (And don’t get me started that iCab just can’t render regular web pages properly.. it screws up very visibly on http://www.apple.com – a website in its freaking bookmarks bar by default!).
And Konqueror with these changes is not out yet either, so it doesn’t count for first browser to pass the test.
Any one of these three organisations (or one of the others as well) can make a release that passes the Acid2 test, and they will be _the_first_ one to do it – what’s under wraps doesn’t count, folks.
They are the best in terms of standard compliance.
how long will it take…?
I’d like to get involved into an internet suite like Opera or Mozilla with the KHTML rendering engine. I searched on Sourceforge, but there’s nothing.
stop being pedantic.
the future stable branch of Webcore and KHTML pass the acid 2 test.
Well i tested firefox latest version and it failed it too, but at least its much better then IE 6.x which is just horrible, nothing worked under it, it said some error and the hello smilely was like, it was murdered. lol
Great, that khtml has passed the test.
But i will still use firefox. Why? Konqueror/khtml has some strange font handling. Depending on the font and font-size i or the website used some “special” symbols/letters weren’t shown correctly. Sadly only KDE (konqueror/khtml) has this problem. I’m not sure if this is a Qt, KDE, or khtml problem because it also happens on other applications like knewsticker. With firefox i had never such problems.
“We are very lucky that KDE is under the GPL and not the BSD license.
That’s you point of view… IMHO it’s too sad they’re using this license instead of BSD ou MIT… But hey… everyone has it’s own opinion… (talk to all BSD guys… and talk with almost everyone how ships costumer eletronic apps where all this work could be used by lot’s of people around the world…)”
BULLSHIT. if KTHML was BSD licensed, apple would not have contributed back. they barely do, now – they comply with the GPL and give A LITTLE support (it IS getting better, yes).
so the (L)GPL has done what it was designed for: ensure what IS free STAYS free. no code writen by any of the KDE dev’s is going to be incorporated by a big company, in its proprietary operating system. THEY MUST CONTRIBUTE IT BACK. GPL WILL KONQUER THE WORLD etc etc etc
(of course you can say this is bad. well, THAT indeed depends on what you think is good or bad. but this has proven, for me, that GPL is superior to BSD: it enforces company’s to contribute back, instead of letting them free-ride on the hard work of others.)
(of course you can say this is bad. well, THAT indeed depends on what you think is good or bad. but this has proven, for me, that GPL is superior to BSD: it enforces company’s to contribute back, instead of letting them free-ride on the hard work of others.)
Some people don’t mind this. They give and don’t expect to have something back… In my opinion, it’s futile to have a debate on the superiority of a licence since not everybody see the world the same way.
From SVN, granted – but everyone’s free to check out KDE and compile it. That’s not possible for Safari and iCab AFAIK, therefore KHTML/ Konqueror is the first browser _available_ to the public that passes Acid2… 🙂
Who cares if a browser passes the ACID2 test? Do you realize what it even is?
It’s a test to make sure your browser properly handles MALFORMED CSS! The real question is whether or not your favorite browser can handle PROPER CSS. ACID2 is supposed to prove excellence with CSS, but actually, failure of the test doesn’t mean you can’t handle valid CSS properly at all.
You are completely wrong.
Error handling is only a part of the Acid2 test suite.
Most of it is about standards. And to comply with CSS standards, they have to take in consideration proper error handling too.
Read here: http://webstandards.org/act/acid2/guide.html
11 set of features listed, and error handling is only one of them (CSS parsing)…
These features are part of existing standards but haven’t been interoperably supported by major browsers. Acid2 tries to change this by challenging browsers to render Acid2 correctly before shipping.
Everything that Acid2 tests is specified in a Web standard […] After careful consideration, we have selected and are testing the features we consider most important for the future of the web.
who cares about an ugly smiley?
No one can use any new technologie for a website as long as MS doesen’t support it.
Because of the huge marketshare of the IE it is much more important that every new browser is 100% compatible to IE.
Almost every Website ist optimized for IE.
It is a mess for any website creator to take care of different render methods of different browsers.
It is nice that some browsers have additional features and better rendering, but they are still not 100% compatible to IE. This should be fixed first to see and use the webpages with every browser like they meant to be.
It’s great that it passes this test, but Konqueror still frustrates me.
I just designed a new website for the company I work at. I’m a standards freak, and my code completely validates. No visible differences between Firefox and IE, except for some very minor font spacing that isn’t really a problem, just a difference. Also looks good going from Windows to a Gnome desktop – no appreciable differences.
Konqueror and Opera, however, both manage to get the width of the Masthead DIV off by over 100 pixels. It’s a very simple DIV that contains nothing but two other divs – and in these are just two links and an image. The positioning isn’t fancy – it can’t be, since IE is a requirement. And yet, Konqueror gets it wrong. I’ve gone over the code several times, and I don’t see anything that would produce this result.
It doesn’t matter for the job, since my boss doesn’t feel that a non-useability issue on two low-marketshare browsers is a big deal, but it’s frustrating, none the less.
I actually use Konqueror fairly often, and I see CSS issues like this come up more often than they should.
Come on, the situation won’t improve if the devs have no clue about it! File a bug report at bugs.kde.org!
Come on, the situation won’t improve if the devs have no clue about it! File a bug report at bugs.kde.org!
If the problem is in KDE, not in his code…
>>Come on, the situation won’t improve if the devs have no clue about it! File a bug report at bugs.kde.org!
>If the problem is in KDE, not in his code…
Doesn’t matter, if he bothers to file a bugreport with a small testcase. He get the bug fixed if it’s in Konqueror or he get told he has a bug in his code, with explanation. A classic Win-win situation.
Some people don’t mind this. They give and don’t expect to have something back…
i use the GPL for all of my software. And it’s not because i expect to have something back. Also GPL software can be used in a company, or by private users and you will never see the software again and their changes, nobody have to release the software to the public and/or to the original author.
I use the GPL, because i give everyone who gets a copy the freedom to use the software for every task but i don’t want that someone takes the freedom i’m giving to him and use it to refuse the freedom to his “customers”. Therefor i use copyleft licenses, it’s not that i want to get something in return, it’s about all the other people who gets software from the people who gets my software.
There is an firefox bounty at http://www.bounty-hacker.org/
Why not make Konqueror as individual product for Windows, Linux, Solaris, FreeBSD, OpenBSD.
Konqueror and Opera, however, both manage to get the width of the Masthead DIV off by over 100 pixels. It’s a very simple DIV that contains nothing but two other divs – and in these are just two links and an image. The positioning isn’t fancy – it can’t be, since IE is a requirement. And yet, Konqueror gets it wrong. I’ve gone over the code several times, and I don’t see anything that would produce this result.
Try to mess with padding and margin, like adding padding:0;margin:0; even where it isn’t necessary. I am currently developing a website that is going to be compatible with these four browsers while not drifting from the standards (save for some MSIE-specific hacks) and I have encountered this kind of issue fairly often.
I use the GPL, because i give everyone who gets a copy the freedom to use the software for every task but i don’t want that someone takes the freedom i’m giving to him and use it to refuse the freedom to his “customers”. Therefor i use copyleft licenses, it’s not that i want to get something in return, it’s about all the other people who gets software from the people who gets my software.
Since _any_ customer have access to your open code, where is the problem? For programs, I do agree the GPL has its value but for system libraries, I just feel the BSDL avoid everyone of us from reinventing the wheel. But my point is that KDE would still fare well whichever the licence they had chosen.
Why not make Konqueror as individual product for Windows, Linux, Solaris, FreeBSD, OpenBSD.
For the various Unixes is up to the packagers to make it available separately (but of course it will still require kdelibs).
On Windows the problem so far has been that KDE’s underlying QT library has not been available in a free and native version. But that’s gonna change soon with QT4, so perhaps we can look forward to Konqueror on Windows too.
Since _any_ customer have access to your open code, where is the problem?
There are more problems:
1. Who says, that i have put all my code on the internet and have released it to the public?
2. If 1 would be true, how the “customer” know that this is the same?
3. Who says that this one who provided a proprietary version of my program has used it 1:1?
…
No, there are many question marks.
The rules are simple. I give my costumers all freedom, but they have give it to their costumers too. Nobody can abuse the freedom he gets from me to cut of the freedom of other users. If someone want to do this, he have to do it by his self and not with my help.
Apple haven’t made it dependant on cocoa though.
http://gtk-webcore.sourceforge.net/
sure there are parts that are will be using it, but apple changes are all #ifdef’ed IIRC.
BULLSHIT. if KTHML was BSD licensed, apple would not have contributed back. they barely do, now – they comply with the GPL and give A LITTLE support (it IS getting better, yes).
First of all, on what do you base your assertion that Apple wouldn’t have contributed back had KHTML been under the BSD license? I’m hardly an Apple fan, but reality seems to contradict you. For instance, I don’t think that the code which Darwin / Mach were derived from were originally GPL-licensed, yet Apple does release the source for Darwin.
Secondly – there’s a difference between what a license requires, and things “would be nice if” Apple did them. Just because some companies who use OSS go above and beyond what is required of them by the particular license, it doesn’t mean that everyone else is required to be as generous. If you dislike this, you could always create a modified version of the GPL that contains a clause a la “No portions of this code can be used in any software that is sold/licensed commercially” (and then see how widely it’s adopted).
It’s unfortunate that many OSS advocates have an attitude that’s best summed up by a saying from Soviet Russia: “Anything that isn’t forbidden is compulsory.”
All fine and well exept the networking performance.Why is konqueror (kde 3.4) remarkably slower than any other browser?It takes 40 seconds before OSnews.com is loaded wheras it takes firefox 2 seconds on the same machine connected via the same adapter to the same network.
Not here. Here they both take 2-3 seconds (seems to be limited by the server side…).
A load time of 40 seconds means there is a severe problem with your installation or maybe your DNS.
Maybe this hint is of help:
http://docs.kde.org/en/HEAD/kdebase/faq/webbrowser.html#id2971134
Why is konqueror (kde 3.4) remarkably slower than any other browser?
I had the same problem with Konqueror (and Firefox). You could try this tip from Kubuntu: http://kudos.berlios.de/kf/kisimlar/tipsntrix.html#boostkonq and see if it helps.
I like the way it is set up: it tests not only for proper processing of correct CSS, but also for the proper rejection on incorrect CSS. I wish that the file were better commented, though, describing the pass/fail expectation of each test case, and what precisely is being tested.
One Anonymous writer wrote ..
My point is that unless IE supports it, it is pointless. IE is not going to lose market share in a hurry, if at all, so the use case if very limited. Standards support is great, but until developers can use them it remains something that we can only ever long to be able to use.
The solution for this is for all web developers that work in MS shops to hound Microsoft until the implment the W3C standards. Everytime an MS developers blogs about the kewl new features in IE 7 , ask about support for CSS 2 and CSS 3. Dog them at every opportunity. Microsoft managers scream, “Developers, Developers, Developers.” So let them hear from the developers that we expect support for standards. CSS can make web pages more compact and easier to maintain. Developers should be allows to ask for these features.
> It takes 40 seconds before OSnews.com is loaded wheras it takes firefox 2 seconds
Your DNS setup is broken. e.g. do you have an entry for localhost in /etc/hosts?
> Come on, the situation won’t improve if the devs have no clue about it! File a bug report at bugs.kde.org!
Been there, done that. I haven’t been able to login to Netscape webmail since they took away the option to use gecko on the backend (2.2?).
I searched and found that others had filed bug reports for the same problem. They were all closed as either “fixed” or “it’s not our fault, it’s Netscape’s fault”.
I’m sorry, but if every other browser on the planet can do it, then so should Konqueror, regardless of Netscape’s code. Until they can launch a simple webmail program, they shouldn’t be bragging about passing squat.
John
Maybe this hint is of help:
http://kudos.berlios.de/kf/kisimlar/tipsntrix.html#boostkonq and see if it helps.
Thanks for the tips,it made konqueror usable for me.
>Your DNS setup is broken. e.g. do you have an entry for >localhost in /etc/hosts?
Yes,and some ipv6 entries which i deleted.
Now shut the hell up.
Written from scratch. I love it. Hey bob check out all this source code in ObjC++.
Looks foreign to me. I couldn’t possibly figure out what they are doing and this will delay us much longer, Neal.
What are you talking about? You call yourselves developers? Read the code and rewrite it in C++. Are you one trick ponies?
No.
Good. You get to keep your respect.
> Looks foreign to me. I couldn’t possibly figure out what > they are doing and this will delay us much longer, Neal.
It’s more like:
– We can’t use this solution because it won’t work with our code base. Neil, we’ll have to come up with a solution that works for us. (This has been discussed to death. I still wonder why they don’t keep yelling: Why the $!§*& did Apple have to create an incompatible fork?! Instead they just complain about people like you who assert their work was trivial.)
Or even
– Wow, now *that’s* an ugly way to solve this problem. But I understand what they’re trying to do. Give me a minute, Neal, I’ll come up with a clean way to do it.
It might pass the acid test, but I still can’t get my mail from my netscape account with it (firefox works fine!).
Since the problem are Netscape’s code it probably lies in broken browser detection, change the browser identification to something other than standard for the webmail site. Perhaps some Netscape version or even a IE one will do the trick.
Broken browser detection are probably the case with 90% or more of all the sites people claim not working in Konqueror.