Slashdot Mirror
Firefox and Opera Fail the Acid2 Test
naylor83 writes "Four weeks ago, Opera's CTO Håkan Lie put forward the Acid2 challenge to the IE developers at Microsoft. The Web Standards Project has now silently published the promised browser test. Somewhat surprisingly, both Opera and Firefox fail to correctly render the test page. Obviously though, they're no where near as lousy as Internet Explorer. More screenshots are available at my blog, as well as at other people's."
Obviously though, they're no where near as lousy as Internet Explorer.
/.er but that line of FUD is unnecessary.
I dislike MS as much as the next
Trolling is a art,
So Firefox is the only one without an "Error?"
Right. So none of the browsers tested can display the test page correctly? And they're the best, most compliant browsers available?
And they've had how long to get it right?
In that case, it would seem to me that it is the standard that is broken, if it's really that difficult to render a page with a cascading style sheet.
(Spudley Strikes Again!)
What is the page supposed to look like when rendered???
"Victory means exit strategy, and it's important for the President to explain to us what the exit strategy is." G.W.Bush
is an analysis of what failed with each browser (especially Firefox.) None of the links told us why the browser failed to render the smiley face or what the WSP did to obfuscate the code. Any takers?
It was known before the test was published that no browser would get it right. That's the whole point!
The reason for having this is to expose bugs in current implementations. Internet Explorer is the obvious retard, implementing about 50% of CSS 2.1, but that doesn't mean that the other browsers can just slack off at 95%. That's not what the W3C is about, it's not what WASP is about, and it's not what this acid test is about.
also fails
If none of the available browsers work on this stupid rendering, who the hell cares? Either modify the spec, or pick a better test.
The whole point of the test was to show things not working in each browser! If you only pick tests that your favourite browser passes, you'll never know what it can't do.
http://home.student.uu.se/dana3949/temp/acid2/ie.p ng is what I look like after using IE too. ;)
like
safari on tiger anyone ?
please post a screenshot of that I would really be intrested
stats on web browsers market share
w3 numbers
The way the browser are rendering this its more of an On-Acid-Test
Right now I am looking at a handy CSS reference chart saying which browser supports what, and the fact is, one third to half of the standard is entirely un-implemented by Mozilla, Opera and IE.
If a CSS standard falls on browser designers to implement, and no one implements it, was it really "the standard?"
Internet Explorer's rendering engine was last updated in 2001. It's well-known that its support for the W3C specifications trails behind other browsers by years. This isn't an anti-Microsoft bias, the same criticism was levelled at Netscape 4 when it was the special ed browser.
IE6 implement far less of the CSS2 spec than any other browser - that's a fact. Hence the "obviously".
Or maybe you clicked on the reference instead of the test.
Umm take the actual test this time, rather than looking at the reference image.
I tried it on the pretty much the same machine as you (just plain Server vs Adv Server though), and it was the same hideous red mess shown in the IE screenshot.
Which browser did render it correctly??? If a standards compliant page can't be rendered properly by any browser what makes it a standard?
The fact that all browsers render it equally bad?
How do they know if they even wrote the CSS properly if they can't get any browser to render properly?
How about we just make a page that firefox renders properly and then make IE and Opera change, then call it a standard, oh wait, that's what MS tried with IE.
The W3C CSS validator fails to validate that page. So is it really a failure when the page does not validate?
Just out of curiosity...what browser did they use to get the successful reference rendering? I'm presuming there's one that successfully renders, otherwise, how do they know their test code is valid? I've clicked around but don't know what they used to generate that png.
"Hardly used" will not fetch you a better price for your brain.
Let's count the unsupported assertions in this post, shall we?
- When dealing with stuff like this, broken is broken.
- There's no degree of broken.
- It's not like Opera and FireFox can claim "well, when we break, there's only 25% degradation."
- In other words, just because the face was a little more legible in Opera or FireFox than IE, doesn't mean that in the real world, they'll at least be readable where IE absolutely will not.
- Parent poster was right, that was anti-MS FUD.
- If FireFox was more disappointing, it wouldn't ahve been mentioned at all.
--MarkusQAccording to whom? And on what basis?
How about "number of features correctly implemented" / "total number of features in spec"?
Why not? I think it's very likely that they'll claim exactly that.
How's this for a reason: if all of the features that fail in Opera and Firefox also fail in IE, no one will use them; this means that pages written to the subset that works in at least some of the browsers may well work 100% correctly in Opera or Firefox and still fail horribly in IE. This is at least a plausible counter example to unsupported claim #4.
The fact that you share his oppinion does not make hiim right. FUD is a fairly specific charge to lay against someone, and in this case seems hard to justify. Editorial bias perhaps?
And you know this how?
I'm confused, is this supposed to be valid CSS2? The W3C CSS validator finds 8 errors in the page.
.parser-container div .parser .parser .parser .parser .parser
-molo
CSS validator results
* Line: 46
Parse Error - second two]
* Line: 91 Context :
Invalid number : color orange is not a color value : orange
* Line: 97 Context :
Property error doesn't exist : }
* Line: 100 Context :
Property m rgin doesn't exist : 2em
* Line: 100
Parse error - Unrecognized : };
* Line: 102 Context :
Invalid number : width only 0 can be a length. You must put an unit after your number : 200
* Line: 103 Context :
Parse Error - ! error;
* Line: 103 Context :
Parse error - Unrecognized : }
Using your sig line to advertise for friends is lame.
Saying that IE's renderer is obviously inferior to Firefox/Opera is like saying that Linux is obviously more difficult to use on a desktop than Mac OSX. It's just plain true. If you had ever designed a complex web site making extensive use of CSS, you'd know this. The only reason you never see IE rendering pages wrong is because every web developer out there has taken extreme pains to work around all of IE's problems.
The real lessons to be learnt from this seem to be getting lost here. If we put aside the MS vs Moz, FUD vs non-FUD and not-as-broken vs either broken or not debates we can see that web designers should have something to look forward to in the (near?) future.
Finally, here is something that could actually give the browser developers something to aim for and help to pull together the standardisation of modern CSS rendering. From how that smiley face is supposed to look I'm already quite excited about what we'll be able to do once all of the browsers are up to scratch.
Now all we need is for the browser developers to take note of this, use it as a learning tool and a target to aim for and give the web design/development community a hell of a lot less stuff to debate about.
It could happen...
But of course, in addition to this they shouldn't let the acid2 test be a final goal and then just sit back and let themselves get rusty. Personally i'd like to see a publicly available acid test for all the new versions/revisions of CSS standards so that Joe Home User can more easily choose which browser to use. An acidN test once every 8 years?
This is the fast moving world of technology, don't you know.
So? That's a good thing. I prefer to know up front where someone's bias lies. Everyone has one and when you try to hide it you have no way of knowing where it might creep it. Just be honest about it and accept it. I can go read an MSDN journal for an IE-biased opinion.
If a standards compliant page can't be rendered properly by any browser what makes it a standard?
Actually, it's a common misconception that CSS is a standard. It isn't. It's an open specification, published by a vendor consortium (the W3C), that invites input from the community and develops the specifications with help from browser developers and web developers from such companies as Microsoft, Netscape, etc.
How do they know if they even wrote the CSS properly if they can't get any browser to render properly?
In the event that a mistake was made, this would be uncovered when browser developers attempt to debug their "mistake" only to find out that they haven't made one.
yep I think its wrong in a couple of places for sure
Line: 91 Context : Invalid number : color orange is not a color value : orange
as any good web dev should know the list of keyword color names is: aqua, black, blue, fuchsia, gray, green, lime, maroon, navy, olive, purple, red, silver, teal, white, and yellow.
if you want a rgb colour (I use english from england ) then color: rgb(255,0,0)
also the error
Line: 102 Context : Invalid number : width only 0 can be a length. You must put an unit after your number
length : Specifies a fixed width.
percentage : Specifies a percentage width. The percentage is calculated with respect to the width of the generated box's containing block.
auto : The width depends on the values of other properties.
Negative values for 'width' are illegal.
For example width: 200px
regards
John Jones
please correct me if I am wrong
What publicly availbe browser can, if any? Because I would love to try it.
Webstandards.org may have good intentions, but I'm not sure I trust their test or their organization. For one thing, someone has pointed out that the W3C's CSS validator rejects it. (This link currently slashdotted.)
Gecko renders part of the second row of the face at the right edge of the viewport. I believe that that is actually the correct behavior according to the specifications. Row 2 is generated by
The blockquote is a block-level element with unspecified width, so except for the 60px left margin, it should occupy the entire width of the containing block (CSS 2.1, Sec 10.3.3). Therefore, the address element, which is floated to the right within the blockquote, is correctly placed at the right edge of the viewport. I don't see why the "shrink-to-fit" rule for computing the width of the blockquote should apply as they claim it does. There may be other errors, but I haven't investigated further.
When the W3C publishes a test, one can be pretty sure that it's authoritative. But who is behind webstandards.org? They claim to be a grassroots coalition, yet I don't see where on their website they invite site visitors to join or contribute to their cause.
Their explanation of the test lacks a contact address at the end to report errors. It seems rather careless of them to publish the test in that state, and it doesn't inspire confidence in the rest of the test.
It's good to see someone making an effort. I do hope that this test will help browsers improve, but I wouldn't trust that the test itself is 100% accurate, either. I would love to see the W3C devise more compliance tests of this sort.
Don't know if I'm just missing something, admitedly I've only skimmed over the tour, but how did they actually render it?
If firefox (and thus I'd assume mozilla), ie, opera, and konqy (as tested by me in kde-3.4) all fail, what browser is so great it does it properly?
Is it open-source? Can it's code be used as a reference for the other open-source browser?
When every single major organisation and company in the business of implementing a standard fails I can't help wondering if the problem lies with the standard rather than the implementors. Just how hard is this?
Doesn't it make you feel good to know that our freedoms are protected by politicans, lawyers and journalists.
Fair and balanced as always!
Where did they get the reference image?
I dont know what you do for a living, but I make web pages, and god damn if IE isn't the biggest piece of non-standards-compliant horseshit browser I've ever had the displease of dealing with. As someone else who replied to you said, you don't see too many problems when browsing the web with IE because us lowly developers have lost countless hours and hair follicles figuring why the FUCK internet exploder is rendering our beautifully coded pure-CSS site like it was a big pile of steaming crap. So yes, it is obvious. IE renders CSS like your mom gives head - HORRIBLY.
Joseph?
Firefox and Opera Fail the Acid2 Test
In a perfect world it should read "every browser fails the acid2 test". Instead somebody chooses to single out firefox and opera.
/^([Ss]ame [Bb]at (time, |channel.)){2}$/
Couldn't this end up just like the war between video card manufacturers with 3Dmark? Where they simply fix thier software enough to pass this test & still not be completely compliant?
Now if only Gmail would work properly in Konqueror, I'd be all set...
I blame it on sleep deprivation, I mixed up the test and reference browser windows. Ignore me.
No, it didn't.
I just can't help but notice how non-objective this site is.
On slashdot, the users submit (and thus, author) stories. They aren't generally schooled in the intricacies of journalism. It's not fair to expect 'professional' journalistic practices from them.
As unfair as it is to expect standards never stated nor implied, the comparison to Fox News is especially bad. Fox outright lies about their objectivity. Most people don't hate Fox News because of their conservative bias, but because they try to pass it off as fair and balanced.
This is "News for Nerds, Stuff that Matters". Nerds don't tend to suffer poor software gladly.
Last, it was obvious that IE would be the worst of the bunch. Some other article could just as easily have said, "Obviously, the Perl version was slower," or "But obviously, the RIAA continues to cripple their music with DRM." Which are roughly equivalent in their subjective way of communicating objective truth.
[I] am no way an MS advocate.
I'm not all that convinced.
Lynx fails also...
I honestly hoped for something without the "ERROR" parts at least.
No, I wasn't expecting a picture...
It comes out something like this (mangled slightly thanks to the lovely filter):
The Second Acid Test Standards compliant?
Take The Acid2 Test and compare it to the reference rendering.
Hello World!
ERROR
ERROR
(four asterisks vertically)
Makes me wonder how these guys can possibly test the performance of the system without a full implementation.
I tried on both Linux and Windows, works fine.
No it doesn't. You're looking at the reference rendering, not the test itself. Just like the 50 other dumbasses who've made fools of themselves by claiming it works when it doesn't.
No. There's a Gecko 1.8 image in one of the blogs linked by the article. It's one of the better ones there but it *is* wrong. So unless it's regressed since 1.7 you're probably wrong.
There are myriad free tools to read PDF. There are myriad free tools to write to PDF (or especially "print to PDF")
There is a serious, design driven lack of any way to edit PDFs. As in, I create a PDF in an application foo, send it to my friend who also has foo, they make changes and send it back. Word has been doing this for more than a decade, and PDF, as far as I can tell, doesn't do this ever. If one of those pieces of software does this, please point it out.
If you want a more detailed text case, try editing text in a paragraph without having to manually reflow the paragraph.
The next-best option available is to use Quark or InDesign, LaTeX or some other decent page layout program, and send back and forth the original files, exporting to PDF when you need to. But you're not editing PDFs if you do that.
Looking for freelance Actionscript (Flash/Flex) or ColdFusion work and/or freelance developers. Email me, put Slashdot
I've got some of the non-IE-supported stuff going on like PNG files with alpha-channels, etc. Right now the pages I'm developing look like crud in IE as colours are missing from transparent windows, etc...
.CSS file that loads for the IE users. They won't look as spiffy as firefox/etc but the pages will at least not be hard on the eyes in IE.
Once all the pages are together though, I'll throw in a
Since CSS is an open specification and not a standard, that is why tests like these don't mean jack to Microsoft. Yeah, I'm sure they would like to be able to render all styles in CSS properly, but they don't care. It is not a standard that they "have to meet," they have the majority of the market, almost everybody standardizes on their product (with regards to company wide products), and if they say they are working on it, they will keep being the leader.
A Firefox (only works on Firefox) plugin that I'm certain everyone's going to find interesting (works on Slashdot too).
x pi [8.7Mb]
http://simile.mit.edu/dist/piggy-bank/piggy-bank.
I wonder how IE6 would do if IE7 was used: http://dean.edwards.name/IE7/
Most people don't hate Fox News because of their conservative bias, but because they try to pass it off as fair and balanced.
It's not FOX News alone that's fair and balanced. It's an hour of CNN plus an hour of Fox News that's fair and balanced. Rupert Murdoch saw CNN's liberal bias as un-fair to the more conservative segments of the populace, so he founded FOX News to balance it.
Thousands of websites use JavaScript and CSS for just one silly reason: To create menubars (with submenus etc.).
So why has the standards committee in all their wisdom never added a MENU tag? It would have been so easy, just allow nested <MENU> tags and voila. On any platform, menus are so common that implementing such a tag is as easy as killing babies.
Rather than giving the website makers what they want (a MENU tag, positioning on font units instead of pixels, proper vector image support) they concentrate on standardizing things no developer ever asked for.
I know why. It would put them out of a job. Try to standardize something like JavaScript and CSS and you're ensured of employment for a lifetime.
Heck, if you want the page to look exactly as you intended, just make a bitmap image of it and send it. It'll probably use less bandwidth than the equivalent in CSS/HTML.
Musicians don't die. They just decompose.
Dave Hyatt Is making progress with Safari's handling of the test, and blogging as he goes. very cool.
Lynx gives this:
ERROR
ERROR
*
*
*
*
Has anyone tested Links?
Organization: alphabetical, sometimes numerical or messy
I get to metamoderate somebody's rating of this article as "Flamebait". I'm going to mark it unfair, but with mixed feelings. It's a great over-the-top flame, concise and well-deserved, but it's also an attractive nuisance that *ought* to attract flaming responses (and hasn't :-)
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
if this is the usual response by a developer at Apple...
5 _04.html
http://weblogs.mozillazine.org/hyatt/archives/200
he hasn't gotten it all fixed yet, but what an effort.
-- it's ridiculous how many people misspell ridiculous... (damn, damn, damn...)