IE9 is the IE6 of CSS3 | Hacker News Reader

IE9 is the IE6 of CSS3

css3wizardry.com

96 pointstoni56 comments

While I can't say I really disbelieve the spirit of the demonstration (past performance is not necessarily indicative of future results, but it often ain't a bad approximation), it'd sure be nice to see more information, especially the source code used.

Anybody who's done some web development would have to look at those screenshots and think "all those differences could be due to just one tag/directive that's being interpreted differently by IE9"... and that different interpretation could conceivably even be more in-line with the spec than what the other browsers do (whether that's a "good" thing or not is a different discussion). So, if I had to bet, I wouldn't put my money on Microsoft being the ones in the right here... but I don't like comparing browsers' standards compliance solely based on screenshots.

I was thinking this same thing. Glad I found it posted already.

I know this is hogwash because I've made that exact IPhone interface in HTML/CSS3 for IE7, IE8, Webkit Browsers, and FF. Sure, the IE versions didn't have rounded corners or even shadows for that matter but it didn't look retarded.

I would guess that the author hasn't been developing cross-browser web applications professionally for any significant length of time.

IE does some dumb things with hasLayout and floating, so you generally avoid them. It also wasn't clear that the author knew that Safari and Chrome use the same rendering engine.

Even with IE9 and their dumb syntax, its not impossible to learn how to make something look freaking sweet in modern browsers, and passable in IE. This position of course costs businesses thousands of dollars a year running VMs for compatibility. It makes it worse that IE9 doesn't run on XP.

The entire post is a little link baity, but I certainly understand the frustration.

  I've made that exact IPhone interface in HTML/CSS3 for IE7,
  IE8, Webkit Browsers, and FF. Sure, the IE versions didn't
  have rounded corners or even shadows for that matter but it
  didn't look retarded.

So it did not have rounded corners and shadows but it was somehow exact? Interesting. Could you elaborate more, what CSS3 have you got working on IE7, IE8?

Well, it was the same for FF, Chrome, Safari, and Opera. IE 7/8 wasn't exactly the same, just didn't have the shadows or the rounded corners. Still looked good though.

Use a little javascript animation between views and getting an app to behave like an iPhone app is extremely easy in HTML/Javascript/CSS.

  IE 7/8 wasn't exactly the same, just didn't have
  the shadows or the rounded corners.

That's exactly the point.

This is a poorly written article for two main reasons:

1) No source code.

2) Incomplete test results. Why showcase two webkit engine based web browsers (Safari and Chrome) but not include a screenshot from Firefox?

3) Factual inconsistencies.

FTA:"The IE team have been chanting a mantra ad nauseum : Same Markup. That means, write once and have it render the same on all browsers. Funny, but the other browsers have been doing that for years. Seems, from the images above that it’s IE that’s unable to render the same markup."

Last I checked, implementing transitions/background gradients (etc.) in CSS required a separate line for webkit, mozilla, etc.

Yeah I was going to comment to that effect. While the syntax is the same for rounded corners for gradients even though they use -moz and -webkit prefixes the actual syntax for the gradient is different.

No. That's stupid. IE6 was IE6 because for more than 5 years it was the last browser of its line.

The "IE8 is the new IE6" claims at least have some meat to them: IE8 is the last (edit: MS) browser running on Windows XP which means as long as Windows XP exists in significant numbers so does IE8, but this is nonsense: Firefox, Safari and Chrome are not going back into their hole, which means IE9 will not be an undisputed champion of anything which means IE9 will not be the last IE ever to run on Windows Vista and Windows 7 (unless microsoft decides to get out of the browser market entirely, but I don't see that happening).

Thus IE9 can not be the new IE6. It might not be the best browser ever, but IE10 will come next and improve it, and IE11 after it.

If you're grasping for parallels, IE9 is the IE5 of CSS3, not the IE6.

What I wish is that Microsoft would just give up on MSIE. For its time, IE6 was the shit. But they let it linger for what, 6 years? Meanwhile Firefox started being awesome, and now we also have Chrome which seems to be setting the standard these days.

As a web developer, MSIE has caused me nothing but hell. Versions 7 and 8 improved a lot, but they still have very annoying headaches that I am constantly dealing with.

One of my biggest fantasies is Microsoft one day realizing how much they HURT rather than HINDER the web, and just throwing in the towel. "We give up."

Developing their own browser does nothing for Microsoft anyways. It's not like using MSIE makes you special or makes your brand awareness of Microsoft any stronger. If they had any forethought for the future, they would just give up now. One can fantasize, anyways... :P

IE is worth a lot to Microsoft because of Bing. Google pays Mozilla for Google search being on the toolbar. Chrome is Google. IE is the only browser that would have Bing be the default search engine. That's worth a good amount of money to them - I can't seem them throwing up the white flag.

Is Bing really a big source of revenue for them though? I thought something like 90% of their profit was Windows and Office.

Either way, good point in terms of promoting their own search engine, but unless it's a significant source of profit then again I say, what is the point?

Bing will never be a significant source of revenue if it has no users. Like it or not, browser defaults drive market share.

Besides, what's wrong with a large company trying to diversify streams of income?

Nothing's wrong with that in itself.

However, it can have undesired negative side-effects.

They could easily release a WebKit-based browser with Bing search built in and call it MSIE, solving both your/Microsoft’s and schammy’s problems.

(Edit: This is obvious to industry observers and has been “requested” by frustrated devs for years. But Microsoft won’t do it. Pride, probably, is the biggest reason.)

And it would instantly break every single intranet built on MS technologies.

They could bake in Trident (IE’s rendering engine) and allow intranet administrators /IT to declare intranets render in IE7/8/9 mode. They already do something similar with their browsers, and IETab for Firefox essentially does this.

It's time to break them.

To me, “IE9 is the IE6 of CSS3” means exactly what the article illustrates: “You will be unable to deploy cross-browser sites using exclusively CSS techniques like CSS gradients and text-shadows without fallbacks specially crafted for one browser.” IE9 will be holding CSS3 back from meaningful adoption and impact.

Others here are griping the author could have done more to approximate the Webkit rendering in IE9. This is true but is counter-productive to the point of the article: Using only new CSS3 tools is impossible thanks to, well, IE9 being the IE6 of CSS3!

> To me, “IE9 is the IE6 of CSS3” means exactly what the article illustrates: “You will be unable to deploy cross-browser sites using exclusively CSS techniques like CSS gradients and text-shadows without fallbacks specially crafted for one browser.”

It is already impossible right now, even ignoring IE9.

So - it works in Webkit and Webkit, and not IE9 Beta. How's it look in Firefox, Opera, etc?

We need to start promoting ChromeFrame to every single IE user on as many sites as possible.

http://code.google.com/chrome/chromeframe/

It will solve the IE6/7/8/9 issue.

The only problem is machines without the privileges to install it.

Google needs to see if they can allow unprivileged users to install it (it would presumably only work on the current user, but that's better than nothing.) I'd love to use it at school, but it requires an administrator.

I can load it up on my school account, but DeepFreeze keeps it off regardless. I expect between security features like DeepFreeze and user policies it would not be possible.

> IE9 is the IE6 of CSS3

No. Two points: 1) IE9 is still beta 2) CSS3 is still beta

From what MS has said, #2 is crucial. They've stated that precisely because of the staying power of IE6, they need to ensure that the next version adopted doesn't get stuck with proposed/intended standards that wind up not materializing. I don't necessarily like this, but I can see some wisdom in it.

The right solution would be to have more frequent feature updates.

Upvoted, but there is a benefit to fewer releases, of course: less diversity. Corporate IT reasonably finds this attractive even in the face of its downsides.

So the lesson if you're gonna be Microsoft and you're gonna have few releases seems to be: pick the standards (de jure or de facto) that seem very likely to have a good chance of being widely-adopted, keeping in mind that you have huge influence over that variable since you control the most widely-used browser in the world, especially if you make your moves wisely and with high quality.

I understand why Microsoft is doing it, but I still know this is against my best interest.

but there is a benefit to fewer releases, of course: less diversity

In theory. yes. In practice, it doesn't seem to work like that. Firefox and Chrome get automatically updated quite rapidly, so the current version is the only very important one.

You're talking about benefits to the progression of the browser. I'm talking about [perceived, at least] benefits to the user (large corporations) of the browser.

Who else other than the user would the benefits go to? Web developer maybe, but their interests are aligned, unless you like being stuck using a IE6-only site during the day, and weeping when you compare it to the modern internet that you use in the evenings.

Corporate IT managers think (rightly or wrongly; probably some of both) that they benefit from fewer browser updates. And to some degree, other folks in corporate IT... developers also prefer targeting fewer browser variations, etc. Perhaps they don't prefer it more highly than their preference for being able to write to modern standards, but when you're into user testing and there's two weeks left on your deadline, you hate to have to spend a lot of time dealing with the fact that it turns out that you have more versions of browsers out there than you thought and they don't all deal with your shiny new web app in the same way. It also sucks to have to deal with calls that the z in Firefox x.y.z broke your app that you weren't expecting to spend any money on maintaining this month.

you hate to have to spend a lot of time dealing with the fact that it turns out that you have more versions of browsers out there than you thought

But there are effectively only 2 versions of firefox in play on this day today:

3.6.10 : The current release

4.0b : The current beta

And you can tell the people on 4.0b that it's their problem for using a beta.

Problem is, the "threat" of uncontrolled update rollouts significantly hampers adoption in many large organizations (which is where the majority of IE6 laggards are located).

I think they should base it on what the other browsers have implemented themselves too. If everyone else has implemented it already, the likelyhood of that feature being removed is pretty close to nil. Have you heard of a css feature (or any feature?) created in webkit or mozilla and then later removed?

I'm not sure about "removed", but I believe CSS2 specified an algorithm for generating numbers (think automatically numbered headings, paragraphs, etc) that was so poorly specified that only Opera managed to get a coherent, working algorithm out of it. CSS2.1 changes that algorithm to be more sensible, and of course it no longer matches what Opera did.

WebKit came up with a proposal for specifying gradients in CSS, implemented and shipped it. Later, Mozilla came up with a simpler syntax, and the CSS working group adopted Mozilla's proposal.

Lets not forget that the main problem with IE6 was that it was the primary vector that infected millions of computers with viruses and created the gigantic botnets we have today.

Sure IE6's CSS compatibility and other rendering quirks are issues, but it was far overshadowed by the huge security hole it created.

Source, please.

Its non-inclusion throws this whole demonstration into question, imo.

That seems a bit sensational considering IE9 is in Beta . . .

This isn't even the beta, it's one of the earlier preview versions.

The web site I'm working on at the moment will reject IE altogether. I'm doing this for the first time ever in my career. A bit risky, I know. Just looked at some of my pages in IE8/9 and realized how much work I saved by not supporting this nonsensical monster that is Internet Explorer.

What is the website's audience and is it private or for work?

Whenever I've done side projects, I just build it for modern browsers and then maybe make it just a bit better than not working on the most recent IE.

That seems sort of user-hostile, but hey, it's your call.

I don't consider even IE 8 to be a burden any more. It doesn't get some CSS effects right, but stuff generally renders OK without modification. I can sort of understand blocking IE 8 if you really don't want people to see your site without border-radius or whatever, and if you don't feel like taking the time to patch it up with conditional code, and if all those lost visitors don't matter to you...

But so far I've found IE 9 to be spot-on. Are you experiencing a rendering glitch with the latest beta? If not for rendering issues, then what's to be gained by blocking potential visitors or users just because you don't like their choice of software?

> The web site I'm working on at the moment will reject IE altogether.

That is... dumb.

Just don't to anything, it'll look broken in IE but no more. If you want to do something specifically about IE, add some kind of warning bar with a conditional comment. You have no idea what the future holds, and locking out a whole line of browsers out of a site is stupid.

And please, if you add a warning, make it to the effect of "Hey, we know there's an issue with your browser, sorry.", instead of some line about IE being evil and standards and so on.

Well, on the other hand saying "we know there is an issue with your browser, sorry" may mark you as an incompetent fool to your readers.

Have you considered redirecting IE to the mobile version of your site? Surprisingly, when using IE6 at school I have found that more and more websites are doing that for at least older versions..

Web sites (potentially) have different levels of dictatorship over the client depending on the value of information or service they provide. For example, only after opening a bank account I found out that their online banking requires Windows and IE (and that, in fact, their online banking is incredibly crappy). I bought a Windows laptop just for accessing my bank account online. For whatever reasons changing the bank would be a big hassle for me. In a way, the bank has a monopoly for accessing my account and related services. This is why their dictatorship works most of the time.

On the other hand, if you are a startup trying to gain larger market share, the situation would be of course exact opposite: you will make your service work on as many platforms as possible.

In my case it's a monopoly. People will access the web site because it would be the only official source. Unless my client revolts of course, I'm going to experimentally drop IE support and see what happens.

> The web site I'm working on at the moment will reject IE altogether.

Don't. What you're doing will be no different from what some of the sites do when they say "You need Internet Explorer" installed.

I'm not a web developer here, so I don't know how much effort will be required but the end result is the same frustration that I encounter when I visit a web page which says "You need Internet Explorer" installed.

The problem I constantly find with all IE versions is that they render fonts much bigger than other browsers, so they constantly make things crowded and cause items (e.g. the About button on that screen shot) fall to the next time.

It's immensely frustrating.

I wonder if he was in quirks mode by not having an appropriate doctype tag?

Not surprised. Btw, how does other browsers like Opera and Safari perform in the same test?

What about Opera?

THIS IS DATED AUGUST 14, 2010. THE IE9 BETA WASN'T EVEN OUT.

Stop blindly upvoting.

> Don’t believe me? OK, I’ll prove it. I have the latest Beta 4 of IE9, in Microsoft speak that means the version before Release Candidate

The author is definitely a troll here.

Reply on news.ycombinator.com