Every now and then, you get these news item that make you feel like something’s wrong. The item doesn’t make sense, shouldn’t be possible, and yet it is. Despite Microsoft’s newfound commitment to web standards, it’s still incredibly unnerving to see things like this – the W3C’s first HTML5 compliance test, in which Internet Explorer 9 Platform Preview 6 outdoes all other browsers.
While the specification is not yet complete, and while the tests only cover a portion of the HTML5 specification, it’s still incredibly strange to see Microsoft’s Internet Explorer come out on top. Sure, it’s a test release, but then, so are the other contestants in the race (except for the Safari version used in these tests).
As you can see, Internet Explorer 9 is outperforming other browsers when it comes to HTML5 standards compliance. Great news for the web, people.
Has anyone else noticed that IE9 gains a 28.57% lead over Firefox 4 for “video” and a 57.14% lead over Chrome for “xhtml5”.
The first makes me wonder if someone slipped H.264 support into that test suite and the latter makes me think “Chrome devs will probably spend the next week or two fixing XML-specific things like namespace support and then shoot up to near 100%”.
As for IE9’s lead on “canvas”, it’s probably just that Chrome and Firefox are still catching up to IE9 on the GPU acceleration side of things and, as a result, they’ve got more bugs to squash.
I notice both Chrome and Firefox betas beat IE9 pretty soundly on getElementsByClassName.
Sorry Thom, results are wrong: http://lists.w3.org/Archives/Public/public-html-testsuite/2010Nov/0… (via https://twitter.com/akahn/status/29484858294 )
I’m no expert obviously, but that email really doesn’t explain what’s wrong with the test. It just says “Opera’s results are bogus” – but doesn’t say why, what, or how. I’m not saying he’s wrong – just that I’d like to know why.
http://www.readwriteweb.com/archives/ie9_outperforms_other_browsers…
It’s only just over 200 tests and is practically worthless at the moment.
Opera chime in: http://my.opera.com/haavard/blog/2010/11/02/html5
I’d like to see how Opera on Linux and FreeBSD does. It’s a totally different product with a lessened release quality. At least with Firefox, what you get in Windows is near what you get in Linux.
I don’t see Opera on Linux and FreeBSD being very different than the Windows version. The main problems for a while were font quality — otherwise they’re pretty much identical….
It’s also based on IE9 preview – which unless things have changed, MS has said will contain features that don’t make it into the final shipping IE9 product. They get a lot of positive buzz this way, but in the end, we won’t get a compliant browser.
I’ll stick with Flash thanks. 😛
Wired has a little writeup on this subject, as well as some results from another test suite. Let’s just say it’s not looking quite as good there.
http://www.wired.com/epicenter/2010/11/ie9-leads-pack-in-html5-supp…
Anyway, it’s a step forward compared to the old IE versions and that can’t be bad.
That article is a bit disingenuous. They complain about CSS3 not being included, but… CSS is not HTML.
http://weblogs.mozillazine.org/asa/archives/2010/10/how_to_add_even…
Yeah, I understand CSS != HTML; but we really lack an all encompassing name for all of the stuff you can render in a browser. The result is people just say HTML when they mean HTML + CSS + javascript + SVG, ect.
Any suggestions for to describe the collection of open standards that ( should) make up the modern web?
I suggest “Corn”. Corn being short of Unicorn, and a staple food of millions.
That article gives the impression that CSS 3 is part of HTML5.
The author could have pointed out CSS 3 limitations without being misleading.
It really seems as if some people are upset by the result. Why am I not surprised that a Wired author put a negative slant on this. The dig at Silverlight was real classy as well.
Wired sucks, they seem stuck in 1999.
Hey now! 1999 was a great year!
The problem is, if IE9 is the next IE6 then it will just hold back development.
Opera 10.60 is not dev version. In fact is an old stable version, since the current is 10.63.
Opera 10.70 beta is out, why haven’t they used this version to be par with preview 6 ?
Kochise
So IE9 is the best implementation of HTML 5 out there?
Huh. Guess HTML 5 just looks terrible by design.
I’ve never seen anyone be so content with mediocrity as people are with HTML5.
Obviously not a web developer then are you?
I was constantly saying with XHTML 1.0 & 1.1
“Nice if there was a semantic element for NAV”,
“Nice if there was a semantic element for a Header and a Footer”
etc. etc. you get the picture.
Really nice extra semantic markup which I wish was there to describe what I normally included in a page and wanted it described in the markup. I want a nice clean separation of concerns between markup and styling.
HTML 5 and CSS 3.0 will deliver what I been wanting to do for a long time and will let me create cleaner markup and a better separation between the semantic elements and its styling.
I only wish it came sooner.
Edited 2010-11-02 23:27 UTC
Yea but thanks to XP/IE8 holdouts it makes sense for commercial sites to stick with HTML 4. We’re just now getting to the point where mainstream websites can drop IE6 support.
HTML5 is going to be mostly a toy technology for the next few years.
I hardly call them Windows XP hold outs given that there are netbooks still being sold today through retail channels that still have Windows XP pre-installed. Until Microsoft/OEM”s do something about this massive backlog of Windows XP machines being sold you’re going to see the scourge of Windows XP and Internet Explorer 8 hang around for quite some time.
I would since the majority bought their machines before Vista.
MS should offer a flat $50 upgrade fee.
Well, according to recent statistics, all versions of IE fell below 50%.
http://www.readwriteweb.com/archives/internet_explorer_drops_below_…
http://blogs.computerworld.com/17106/take_that_microsoft_ie_drops_b…
XP has 60% of the market, and if a person is using IE then it is most probable that the machine is running Windows. So probably only 50% of the users on XP also are using IE now … so 30% altogether are using IE on XP right now.
When IE9 comes in, it is such an improvement that a good percentage of Vista and Win 7 users will upgrade to it. This leaves the previously-identified 30% of users as the only group who might still be staying with IE6/IE7/IE8. 70% of users will be able to render HTML5.
OK, so that is a small enough market that some websites might begin to ignore it. Certainly there were websites which ignored compliant browsers (Firefox+Chrome) when they only had 30% of the market.
So if the web starts to shift to HTML5, those 30% IE6/IE7/IE8 users left out will be told that they can still see HTML5 pages if they: (a) move to Firefox, or (b) move to Google Chrome, or (c) install a plugin for Gogle Chrome Frame.
Option (c) might even be desirable for users who want to stay with IE6.
So it might be possible to change the rich web content away from HTML4 over to HTML5 despite the older-Windows holdouts.
It’s not even close for commercial sites to ignore. Jimsjerky.com is not going to turn away 30% of visitors just to have a flashy website.
Most websites will just use HTML4 + Flash for multimedia.
Then how come Jimsjerky.com used to be able to afford to throw away 30% of users (Firefox+Chrome) when in years past they were an “IE only” site?
Most websites will just use HTML4 + Flash for multimedia. [/q]
Perhaps, but then they will lose the large market of users using iPad or iPhone.
The only way to win will be to go to HTML5 and trust that a good percentage of WindowsXP users will download Firefox or Chrome (which many have anyway).
It’s a bit like saying something is a daunting task only to have a third party quote you as saying it was a nightmare.
Different people will interpret the same results or comment in a manner that best suits their own stance or preferences. As is pointed out in the WIRED article, having better raw support for certain HTML5 functions – specifically those that are targeted by the official test – doesn’t necessarily mean it’s better at doing certain things from a user’s perspective.
That being said, it’s excellent to see Microsoft moving in the right direction – will certainly keep Google, Apple and Firefox devs on their toes…
Right you are. For the most part, as long as pages look ok when they load, only geeks and web devs are going to care about HTML5, benchmarks, and all that shit. As for me, if it doesn’t have adblock, it’s irrelevant. That goes for any browser.
However it does matter when you come to improving the website.
I work for a large charity that currently has a very crappy CMS, and a website that has been largely done so it works in IE6 & 7.
Making changes are a nightmare and costs the company more down the road than on the initial implementation, because instead of making nice clean changes, I have to constantly hack around what is already there.
Not an ideal solution, I would gladly fix it … If I was allowed the time to, but I don’t.
If the devs followed web standards from the start, I would not be having to hack upon hack to make things work, this cost me time, and cost the charity money.
Which means I am not working cost effective as I could have been.
It all nice looking alright on the page at the time, but maintenence costs will spiral if the implementation is not correct.
The old saying “A stitch in time, saves nine” is very pertinent, when it comes to any software development.
Foolishly, when creating a prototype web app for the same organisation , I decided to cut some corners so I would have something “that worked but was dirty” to save me some time … I am now paying for that when creating the actual web app.
Edited 2010-11-02 23:38 UTC
Like I said, only webdevs will care about that kind of thing Most of the rest of the world isn’t going to choose a browser based on how well it scores on some compliance test or benchmark, unless it happens to be a LOT faster than its closest competitor. Even then, it still might not get the nod from many users, if it’s lacking in features and/or has a crappy UI.
No not only webdevs care, not directly anyhow.
My manager cares when my time is being wasted because I have to work with someone elses crap code and it take many times as long as it should do to get a job done … his boss will care when the project is late and whoever above him will care when he sees time and money being wasted.
When I get well written clean HTML and CSS to work with I can do my job quicker, which means the project gets done on time, which means the bosses up the top are pleased and the business saves money.
This things have a ripple effect, I see it everyday through the organisation I work in.
You could always donate your time to fix their website and get it as a tax write off
I don’t work in the USA.
How about updating your profile darling – no need to be some sort of paranoid individual believing people are out to get you lol.
Have I not? Thanks for the heads up.
On my Windows machine Chrome is version 8.0.5** dev, and it also appears the version of Safari used for the test is the latest stable one, and not the latest dev build.
So the test isn’t using the newest alpha/beta/dev versions of all the browsers.
Same with Opera. This is the fifth month that version 10.60 has been out.
IE’s shaping up alot lately. Good stuff.
If it doesn’t support <input type=”date”> I’m not interested.
so you are only interested in opera?
Yes, im so dissapointed that just opera has implemented it.
“The HTML5 Test suite is still being developed. The number of tests and the results on these tests will change.
The results in this document may be updated, replaced or obsoleted by others documents at any time.
It is inappropriate to cite those results as other than work in progress and unstable.”
All the tests at w3 are provided by the browser makers themselfs, I wonder how many came from Microsoft.
The whole test looks like it has been tailored to be misleading. I came to that conclusion when I read the next message on the thread (that Kroc linked to in the 2nd comment to the story):
It’s unbelievable that he has the gall to write such a thing. I can only guess but I don’t see other browser vendors collaborating in the production/publication of those conformance tests: Opera’s Anne van Kesteren called it “extremely silly”, Apple’s Stachowiak throws in a tepid “We should probably be cautious about the chance of creating PR events based on incorrect information”.
On second reading, the most weird thing is W3C’s Le Hegaret’s own words (“Thank you all for sending test results.”) imply: other vendors ran the tests and returned results to the w3c… I instantly thought “How stupid of them. In these conditions, I don’t see how to blame Microsoft for reaping PR profit”. But I kept on reading the email thread and a few messages later came across this from Apple’s Stachowiak:
and this:
wtf is going on with the w3c?
Last message in the thread, apparently by the person who collected the results:
Maybe one of the websites that started and fueled the fuss are watching the w3c public list?
Yeah, it’s pretty clear that this was just some extremely preliminary test results that wasn’t intended to make it to the media.
Thom, perhaps an update to your article is required, stating just how unofficial and unrepresentative these test results are – as indicated by those responsible for them.
Apple’s permission is not required.
And how does this meaningless comment of yours tell us what YOU think of the story? what your contribution to the discussion is? Have you read the email thread and put the different quotes in context?
Seen this phrase before, it’s embrace. We leave them, but we forget and they get us back with how they have changed. Then when we’re in the house and not watching anymore, they lock the doors. Then they start hitting us again. With chairs, and everything goes bad. Then we can’t use a browser of our choice for the company internal website, or for our wives to do SATS marking data entry, or anything else. As we have to use IE, we have to use Windows. No no no no. The moment that door closes, because “it keeps the warmth in”, or what ever, we don’t wait for the click of the lock and run regardless of what “good thing” the extend is. We hit first and run out the house screaming because we know the extinguish and bad things are coming.
I’ll be sure to wear my tinfoil hat.
Sure, never happened before or anything, or been a stated policy…..
everyone doubting the results should take a look at
http://test.w3.org/html/tests/reporting/
don’t be fooled by the filenames
the actual browsers are:
————edit———–
just noticed that they updated the results today
Edited 2010-11-03 18:25 UTC
lemur2 p0wn3d by IE9.
How does that work?
My only comments on IE9 on this site to date have been to applaud the efforts of Microsoft to bring its standards-complaince into line with or ahead of other browsers. Well done.
I have had a go over other Microsoft web initiatives such as Silverlight, which is precisely the opposite of a standard, but here again Microsoft seem of late to be doing the right thing by diminshing the role of Silverlight. So another welcome initiative right there.
Microsoft even have said the IE9’s HTML5 will support WebM if the user install a codec (which I’m sure Google will provide). It would have been better had Microsoft built in support for WebM by deafult, as they have done for H.264, but I suppose one can’t have everything, and even this position is reasonable.
Now Windows XP still has 60% of the desktop market, and IE9 apparently won’t be offered to XP users:
http://itmanagement.earthweb.com/entdev/article.php/3904131/Interne…
… so that is a little disappointing (as far as adoption of web standards goes) I suppose. However, all is not lost even from that persepective … that 60% of the market can keep an older version of IE if they wish/need to and still keep up with web standards by either installing Firefox (and IEtab extension), or installing Google Chrome Farme.
So it is all good for users in the end.
This outcome is, after all, what we should always support, no matter where it comes from.
They’re almost current. Now they just need to get WebKit Nightly 71204 Trunk tested.
Microsoft’s software sticks with standards ?
oh no, it’s gonna rain frogs.
They have updated the test with Opera 11, but it’s still the first Alpha (1029) while a second alpha has been released (1045).