Opera Lays Down Acid2 Challenge

sebFlyte writes "The CTO of Opera has proposed a new version of the acid test for browser compatibility, and has challenged Microsoft to make IE7 a browser worth having that will do the Web good. He's asked to help from Web designers the world over to build a new page for Microsoft to test IE7 with to make sure it does everything Web designers want it to. "

Comments

Implementing full standards would help

[PornMaster]

It'd be nice if there were reference pages made by the standards committees, so a browser could be simply deemed compliant or not.

Re:Like, render Slashdot the same way every time?

[porcupine8]

Am I the only person who has never had any issues with slashdot and firefox? Or if there are, they're minimal enough that I've never noticed them, so I doubt they're really worth bitching about.

Test suites

[cortana]

Aren't there test suites that test the conformance of an implementation to all aspects of CSS2 standard already? And if not, why not?

Re:Like, render Slashdot the same way every time?

[ScentCone]

Actually, I use Firefox all day long, even though I frequently turn to MSIE for some testing or specific corporate sites. I build pages for a living, and I've a long, long list of gripes about every browser, in every version. Opera in particular seems to give some table nesting some trouble that IE doesn't exhibit, but all things considered, I'd rather use FF. There's a lot of pot-calling-kettle-black about the Opera challenge, that's all. All of the kettles and pots are black. That being said, I think some sort of Ultimate Browser Agony Test is a good idea. But to suggest that it's somehow Microsoft's fault that we need one is, well, just dumb.

I hope not!

[Anonymous Coward]

I don't want my browser doing what many web "designers" want it to do. I want my browser to do what I want it to do. The challenge is a bit misleading, or its restatement is. The browser should be capable of doing things in a standards compliant manner. Too many web designers (or designer wannabes) want to do stupid, irritating things. Nobody has to let them.

Re:Like, render Slashdot the same way every time?

[Jerk City Troll]

View the source on any Slashdot web page and observe the gigantic, sloppy cluster-fuck that is their output. Do not blame Gecko for this nightmare of deeply nested tables, font tags, missing close tags, and other crap. (What's funny is that Slashdot gives an HTTP 403 to validator.w3.org.) When Slashdot makes their code sane, then we can blame the browser. It's amazing to me that any user agent can parse this and make sense of it.

Slashdot != xhtml

[Arbin]

Slashdot produces some of the worst html code available, and it doesn't produce the same exact code EVERY time. Some days it offers spans, some days tables missing end tags, it's just random garbarge. How do you expect ANY browser to render code the same, EVERYTIME? God. I fed a troll... *shaking head*

Wrong target

[whitehatlurker]

While it would be nice if MS IE 7 were a more compliant browser, the real target is the lazy web authors who do not try to comply with standards.

Or the malicious ones who miscode their site to intentionally over-support a browser.

I support Hakon, but I think he's aiming at the wrong spot.

Caveat: I have used (and liked) Opera since version 3 or so. I am have used (and hated) IE since version 2 or so. I am hardly unbiased.

Sometimes, IE renders bad HTML well

[benhocking]

This is not meant to be an attack on IE. If reasonable assumptions can be made about what the code should do, even when coded incorrectly, then it's great that IE does this. I'm not sure of any specific examples, but when I first started writing web pages (years ago), I remember that Netscape would cough on some pages that IE rendered well. Invariably, the problem was that I had left off some terminating tags, and IE correctly figured out my intentions.

Three caveats:

First, having Netscape scold me allowed me to fix my code. IMHO, a better way to do this, however, would be to have an option called "pedantic" that would insist on matching tags (where appropriate). This might exist now, and if so, that's great.

Second, trying to "guess" what was intended is rife with problems. Anyone who has used MS Word for long enough knows what I mean.

Finally, I currently use FireFox the vast majority of the time. I do not know if any of what I said is still true.

Just a thought...

[Saeed al-Sahaf]

Just a thought (many dead bodies spinning in graves), what if IE7 is the Tit, the Jones, the Cake, the next best thing to drugs, and secure? Will it be a sign of the apocalypse?

Opera? Compatibility?

[leshert]

I'll be more impressed by something Opera says about compatibility when they fix their own issues--particularly their shoddy XMLHTTPRequest implementation.

Eric Meyers ComplexSpiral example

[JanJoost]

I'd like to submit someone elses example page of the horrific way IE6 handles CSS2:

http://www.meyerweb.com/eric/css/edge/complexspira l/demo.html

It not only describes what goes wrong, but why, how and where.

Oh: Eric: if you're reading this: Thanks! :)

Some real CSS examples

[mathmatt]

Look at this detailed analysis of how IE, Opera and Mizilla render CSS. Note that Opera and IE were both wrong at first, but Opera has adopted Mozilla's convention.

This clearly demonstrates that the "browser war" is really a one(IE)-on-one(Firefox) battle with Opera and others simply choosing which side to mimic.

Re:Opera is already dead.

[bhtooefr]

First off, Opera's still innovating. Who has voice? Not Firefox, that's for sure... Who has small screen rendering? Not MiniMo (yet, anyway)...

And, I've heard Opera's biggest moneymaker is NOT the desktop, where one can pay $0 and get text ads instead. It's mobile, where they're the only good game in town on some OSes (and they're working on Windows Smartphone - Pocket PC and Palm will be the only major OSes lacking Opera).

Re:Hrrrmpph!

[ArsonSmith]

Which begs the question "What is a standard?"

The problem...as I see it

[msimm]

Is the web developers. I've had some down-right nasty exchanges by developers who believe that because IE represents 95% of "their target market" that 'standards' don't matter.

A good example would be something along the lines of this (a response from an actual discussion I took part in, the funny thing was I wasn't trying to tell anyone anything about the W3C or the importance of standards, I just asked a question about a script that was acting strange in Firefox, my current platform of choice):

A group of nameless and faceless aHoles got together and declared to the world, this is the "standard." You have to "validate" your code or it's no good, because we say so.

And people like you, fall all over yourself, worshiping them.

In my opinion, only a damned fool let's someone else manipulate him, whether he can see him or not.

Those aHoles are meaningless to you, they are meaningless to your web site, but you are so weak and gullible, that you can't and will never understand that.

Admittedly this is an extreme example, but I believe it is representative of a broader belief that might makes right. Firefox/Mozilla/Opera/Safari are still a relatively small ripple and there are some stodgy people out there who at best, simply don't care if their code works on a minority browser.

Until it hits their pocket-books thats not going to change. The pressure needs to be put on businesses so that when say Bank AAA gets a site built that can't/doesn't support your browser (because of non-standard code created by people either too stubborn or too lazy to spend the extra 3 seconds to create/read about browser-friendly code) they hear about it. Maybe even lose some customers.

Then our friendly web-developer can come back and learn how to fix his/her code. If that happens enough they'll get tired of doing it the old way and maybe play nice from the get-go.

FTR, the code we were discussing in the the above quoted passaged did get fixed, by me and I have about 2 weeks of javascript programming under my belt (and if your wondering about the preceding conversation, no, I wasn't impolite or anything like that, I'm too old to pull that kind of crap).

Re:Like, render Slashdot the same way every time?

[bemenaker]

occasionally, I will get a huge empty black space in the middle of my screen and the text is over it and unreadable, (being black). The white space is far shifted to the right. Or the text is shifted left over the navigation column. Refreshing once or twice clears it up. I get this probably once a week. I use Firefox exclusively.

Re:Like, render Slashdot the same way every time?

[Curtman]

Have you never seen it do this?

That one isn't anywhere near as bad as the times when the entire page content is shifted one screen width to the right, but it still annoys me very much, and happens almost every time, whereas the page shifting bug only happens I would say 1/20 times as a rough estimate.

It's not hard to create valid pages that render differently in different browsers. I just want to know why.

:)

Re:IE not worth caring about

[Doctor O]

This is not +4, Insightful, because the OP is correct. Amongst others, I work for BBDO, TBWA, Publicis and RSCG, and none of those care much for IE, because if you have designers who understand the medium, it's easy to build nice (==compliant) XHTML/CSS which renders just fine everywhere. Most problems derive from graphics-laden sites sketched up in Photoshop (or even worse, XPress or Freehand) by aging print designers which require nonsensical pixel-perfect rendering and are demanded to be built "exactly like the layout!". Broken design, broken result.

I might add that of course there are many huge players in the field who still travel the 'optimized for IE' road and build shitty stuff which just renders in one or two versions of IE (rendering of CSS changed severely in some areas between IE 5.0, 5.5 and 6.0). But those who understand what they do, produce good quality. The IE only pages are already disappearing, and actually my perspective on this is not that it happens because every new IE release introduces more incompatibilities with old versions and people are fed up with kludging up their HTML when they can just rebuild it and have it work everywhere in half the time which is required for debugging all those stupid side effects. Look at Google Maps as an example. Doesn't look very 'optimized for IE' to me.

(I only talk about professional web design. I refuse to discuss the HTML practices of hobbyists, they'll build broken pages until browsers refuse to display them completely (won't happen), and frankly shouldn't be bothered with such discussions. Let them play with it, maybe some of them get enough out of it to learn how to do it the right way later.)

Useful?

[tepples]

However, the CSS and XHTML specs on the W3C site are difficult to understand and therefore not all that useful when you're trying to debug a rendering problem. In addition, which browsers support searching the specs for, say, all rules that contribute to this border?