Hacker News new | past | comments | ask | show | jobs | submit login
IE10 platform preview available - with native HTML 5 (msdn.com)
115 points by intranation on April 12, 2011 | hide | past | favorite | 105 comments



This reads like a political attack ad. So many half-truths in here.

Just an example: they claim that Chrome "dropped support on Windows XP for functionality that [IE Team] think is fundamental to performance." Linking to this blog post: http://googlechromereleases.blogspot.com/2011/02/dev-channel...

The reality is that Chrome disabled it temporarily on the dev channel due to crashing, and in fact brought it back for v11: http://code.google.com/p/chromium/issues/detail?id=72975

Microsoft doesn't even acknowledge the fact that they don't even have a version of IE10, or even 9, for Windows XP!

Is this really what the browser wars are going to come to? Lies and marketing? I thought we were over that.


They actually came out and said "building a new browser for the ten-year old version of Windows that came with IE6 didn’t make sense to us because of the limitations of its graphics and security architectures." Considering IE 9 dropped support for XP, I don't think they consider it "news" that IE 10 won't support XP.


>Microsoft doesn't even acknowledge the fact that they don't even have a version of IE10, or even 9, for Windows XP!

Whaa?? I'd cut them slack on not supporting XP.


You're missing the point; they are attacking Google for "disabling" features for XP while they dont support XP at all (with their latest browsers).


Why? Everyone else is supporting XP with no major problems. Meanwhile, MS has the source code for XP, and for some reason can't replicate that success?


It's over 10 year old operating system already. Software has its shelf life. I can't blame them for wanting to not spread their support matrix too wide.


I don't blame them for trying to kill it off. Truly. It's been long enough, especially now that 7 is actually a solid upgrade for nearly every use.

But I do blame them for all the under-handed, customer-spiteful tactics they've used. DirectX 10 on Vista only, though it's almost 100% compatible with DX9 (having 9 report version 10 allows many DX10 games to run). IE9+ on Vista/7 only. It's complete bullshit, through and through, and I see no reason to defend their methods.


> (having 9 report version 10 allows many DX10 games to run

The only games with which this will work is games that support DX9 explicitly and, for some reason, disable it. The D3D9 and 10 APIs are completely different. Having worked on an implementation of DX10 for XP, I can say for sure that this is 100% incorrect.


/me hunts around

Aaah, I see where I got that from. Early speculation on Crysis. Apologies!

Know how well the Alky project managed to get DX10 on XP working? I just found it now.


That was my project, and not very well haha. It worked, if you consider a lack of shaders, lighting (IIRC), and other critical things to be "working". Shaders were a PITA for a few reasons, not the least of which being that the D3D10 shader bytecode was completely undocumented. Spent a couple months doing nothing but reversing the bytecode format, and things sort of fell apart after that. All the code's out there, though, as is the complete story of the project and the company around it: http://daeken.com/alky-postmortem


Hah, what a find :D I'll be happy to read it.

I'm not entirely convinced mimicking anything Microsoft puts out is a good idea, even under the best circumstances. I've lately been getting pretty far into .NET, and the more I see, the more it terrifies me.


I'm on the fence with them supporting XP. One comparison to take into account is Safari's releases have dropped support for an older OS, much less than 10 years, they did change processor architecture though and have less business users to support. Also Microsoft offer free support.


I don't think you understand. MS is criticizing Chrome for temporarily dropping support for XP (re-introduced in v11). IE9+ doesn't support XP at all.

That's what you call doublespeak.


I don't think they are criticising, I think they are making the point that they aren't the only ones who think you can't do hardware acceleration in XP. Whether that statement is true or not is irrelevant.


The only native experience of the Web and HTML5 today is on Windows 7 with IE9.

I'm leery of the word "native" in this context. It strikes me as a marketing phrase with little or no actual meaning. They're trying to sell me on the idea that my HTML5 experience will somehow be better because it uses code provided by Windows itself rather than by some intermediary library. But what is that windows code, if not a library of code?


I think it is a brilliant marketing term. yes they hijack the term "native" but if in fact they can deliver, it will be a clear differentiator in users' mind, like this:

IE10 = native = fast

Chrome = not native = slow

Simple message, and it works IMHO.

I'm watching MIX'11 live and so far the demos are great (like the fishbowl benchmark, completely blowing Chrome's fish out of the water ;D) but they are still just demos.


The beauty of this is that if IE10 is actually faster than Chrome, then Chrome has fulfilled its mission perfectly.


"Faster" is not the holy grail metric. Standard compliance and well-thought-out interoperability are high on my list.

Don't get me wrong, I don't want to use a dog of a browser, either, but in modern times with dual-core this and 8GB that, one would probably have to actually work at it in order to have a genuinely slow browser.


It's easier than you'd think.


I think the Google folks will disagree with you :D


I think they will agree, the availability of good browsers means their web apps work great everywhere.


yeah, that was the actual intent of Chrome as far as I understand it and it has worked.

Look at Firefox 4, you can't say it wasn't influenced by Chrome. And the fact that Microsoft is now trying to move faster with IE10 already being announced is probably also largely influenced by Chrome's fast paced development.

We now have a better faster web. We just need it to be standardized a little more (Come on Microsoft, please add WebGL support)


Except for when they deliberately cripple them for Opera users :\


True, but MS will always try to steer people toward their own services, so it's not a complete win. But definitely better than having people on IE6 trying to use Gmail, or even search for that matter.


How is that any different from what Google does? or Apple for that matter?


From Google's perspective?


It's a "brilliant marketing term" in the same way that any falsehood is. Other IEs have also been "native" in that way, yet they are dog slow, so "native = fast" is clearly wrong. "IE10 = fast" would skip the lie and seems like just as good a differentiator.


True, but "fast" is a crowded concept in people's minds. If you look around the web you'll read that Firefox 4 is fast, Chrome is fast, Opera is fast. IE is "native" which is then connected with performance claims. It's a new name for an old feature that has a kernel of truth to back it up. I'm reminded of an early scene in Mad Men where a salesman convinces a cigarette company to advertise their tobacco as "toasted". Sure, so is everyone else's, but that doesn't mean that one company can't own that idea in the minds of the public.

A similar thing thing with the IE blog's new favorite phrase "same markup" which seems to be "standards compliant" in a new blue dress. Search the web for 'standards compliant browser' and you'll generally find folks heaping scorn on IE and praise on all the others. Search for 'same markup browser' and the opposite is true.


Sometimes "native" is a bad thing in browsers. Like ActiveX.


Or NativeClient. Ironically it contains several runtimes although "Native" in this case means non-web programming languages for the client-side. Remember the HTML component is a runtime.


It may be a brilliant marketing term but for us technical guys it sucks - from now on we have educate everybody we meet about what the term "native" really mean.

Not to talk about all the wrong decisions that will be made based on this.

Marketeers, if we could only get rid of them.


Well, don't we already have to do that for "open", "closed", "evil", "chrome" (multi-definition), "app" (web, native), "social", and so on?

and BTW, Microsoft didn't start this. Remember Native Client?


I remember native client, but was actually native in the tech sense since it ran x86 code.


Here is another funny one, from http://www.intel.com/products/processor/index.htm : "Desktops powered by the 2nd generation Intel® Core™ processor family let you see and do more with visibly smart performance." From http://www.intel.com/products/processors/previousgeneration/...: "Previous generation Intel® processors in the Intel® Core™ processor family delivers desktop performance, reliability, and energy efficiency."


Its so native to win vista/7 that it does not support windows xp which Opera/Firefox/Chrome ALL support WITH their performance improvements AND they've done so while staying ahead of IE9.

I guess if that's native, yes IE9 wins. I think in user's minds: if its built for MY operating system specifically its better than the other. However thats not true as even the tech unsavvy are flocking away from IE.

HOWEVER. I am glad to hear this news. It only means one thing. PROGRESS.


I suspect their messaging is driven by internal politics as much as it is by external marketing imperatives. "We are a vital, integral part of Windows and should therefore be heavily funded".


Yeah, it sounds like they're trying to hijack the term to give the impression that other browsers are somehow not "native."

Did IE9 not have "native" HTML5, or is it just that marketing only came up with the idea now?


Please read the post. Direct quote:

> IE9 delivers native support for HTML5 on Windows.

Another direct quote:

> Native HTML5 support in Windows with IE9 makes a huge difference in what sites can do.

And another:

> The only native experience of the Web and HTML5 today is on Windows 7 with IE9.


The first two quotes give me more of a sense of "IE9 brings HTML5 support to your native apps, by installing a new version of the MSHTML web control."


In the sense of using the platforms' libraries and APIs, then Chrome is not native (especially some of the UI as Javascript is not native in any sense) because it comes bundled with its own runtime as well as the WebKit runtime. But in terms of systems languages they are using yes that is native.


They used the same messaging with IE9.



And, if that's their criteria for a web browser being "good", how would Safari on Mac OS X also not qualify?


Webkit?


WebKit ships with the OS, just like IE's MSHTML/Trident engine. Does being open source make it less "native" somehow?


The argument that webkit is not native because it is open source is a plausible reason to exclude Safari - hence "webkit?" rather than "webkit!"

In my opinion, claims about nativeness are on a par with those about magicalness - but marketing flacks don't get paid based upon sound ontology.


I think "native" means they can leave supporting new features everyone else has had for years until a new version of the OS comes out, rather than being bound by this silly "upgradeable" thing.


Yea, I know MS is the most non PR2.0 compliant browser vendor for a while now.


HTML5 has become a marketing term, it's not so surprising that Microsoft needs to differentiate themselves with their late jump on that bandwagon.


In discussing why IE9+ will not be available for XP, they write:

Others have dropped support on Windows XP for functionality that we think is fundamental to performance.

This is completely disingenuous. What they are actually referring to (and link to) is Google disabling GPU acceleration and WebGL on XP starting in Chrome 10.0.648.114 due to stability issues. Importantly, Google intends Chrome 11 to re-enable these features on XP for known-good drivers. Here is the relevant ticket:

http://code.google.com/p/chromium/issues/detail?id=72975

The thing is, I actually agree with Microsoft's decision to not support XP. XP, like IE6, is a fundamentally flawed platform and the sooner people move off of it, the better. I just wish they'd make the argument honestly.


I doubt I can ever agree with Microsoft.

They have this nice campaign running about how people should upgrade from IE6 to the lastest version:

http://ie6countdown.com/

What does this mean ? IE8, because there is no IE6 on Vista or 7, so they are talking about Windows XP. And their is no IE9 for Windows XP.

Windows XP extended support is till 2014. Last year (2010) Windows XP was still being sold.

The market share of Windows XP has over 50% worldwide, Microsoft should not ignore those users and just release a IE9 for Windows XP already.

But who am I kidding, IE9 was a rush job, it doesn't even have a proper JIT javascript engine when you run IE9 64-bit.

On Windows 2000 IE6 is actually the lastest version, but I won't comment on that further. ;-)


1. All this talk about 'Native' is basically propaganda. There is no such thing. All modern browsers generate native code when running JavaScript, and most use GPU acceleration to render content (for example, Firefox uses Direct2D, exactly the same as IE9). IE9 and 10 are not more 'native' than other browsers.

2. I am impressed by the work on implementing new standards - kudos to Microsoft. But I did not see anything about WebGL, which is a very important standard that is already implemented in Firefox, Chrome, Safari and (soon, or already) Opera.


Mozilla is now going full-steam ahead with a plan to support Native HTML5 in Firefox. :)

https://bugzilla.mozilla.org/show_bug.cgi?id=nativehtml5

http://arewenativeyet.com/


WebGL won't happen until it makes its way into the HTML spec. Historically MS hasn't been supportive of Khronos standards. And I don't really know of any customers important to MS that are demanding WebGL.


Yep the story about OpenGL and DirectX is well-known now.


Glad to see that Microsoft isn't going to release IE9 and just sit on it. That is good for Windows users and good for the web.


Criticizing Firefox for not rendering something correctly seems like throwing bricks in a glass house.

Also, I didn't see a Javascript benchmark in there.


Mozilla should just thank Microsoft for finding the issue then fix it. We want one standard. Not a big deal. It'll be fixed before IE10 ships.


Warning - Shameless plug - I elaborated on the issue here: http://news.ycombinator.com/item?id=2437945


IE9 renders nice and quickly but in contrast to their claims of speed, I've found their DOM manipulation to significantly underperform in contrast to the other browsers. Firefox 4 and Chrome 10+ are able to create and insert large numbers of new elements at the same time significantly faster than IE9, while IE9 stutters and lags under the same conditions.


It's good that they will keep working on IE, but since they will only release these "previews" for a year, that means new features will always be way earlier in Chrome and Firefox (soon) with their fast release cycles. IE10 will launch with features that Chrome and Firefox will have 10 months before.


I really hope that Microsoft starts to shorten their release cycles. Releasing once every few years made sense in the 1980s and 1990s. Not so much today.


They wont. It looks like IE10 is being bundled with the release of Windows 8. In late 2012. Merde.


That is a fast release cycle for Microsoft! It's a step in the right direction. Microsoft is a giant cruise ship trying to compete with speed boats. It won't happen over night, but hopefully it either does happen or they give up. I still feel this is great news overall, it shows Microsoft is aware of the situation. If they ever do truly remedy this (and by remedy I mean a very fast IE release cycle, ideally with automatic updates like Chrome does) it will take them a while to get there. I'm just glad they appear to be slowly turning this cruise ship around.


Going to http://ie.microsoft.com/testdrive/ on Chrome (dev-channel) shows a little banner with the following message,

"Cool, you're using a Chrome 12 nightly build! Don't forget to enable your partial hardware acceleration in the about:flags thingy..."

Seems like they're actually worried about Chrome, since Firefox and Safari don't have any similar messages.


I think they also want any impromptu benchmarks people run to be as fair as possible. To the best of my knowledge, the latest FF and Safari both have all their supported hardware acceleration turned on by default.

I've met some members of the IE tech team (who don't control the marketing). The people I know really want to win without cheating and support as much as possible, but are very wary of releasing anything too early. They're very careful to wait until the standards are pretty precise and stable before releasing an implementation, lest they be accused of trying to "sabotage" something with an unintentionally different (but still within the vague spec) implementation.


Firefox already has what they call "full" hardware acceleration, and Safari has none I think, so there's nothing to show there.


IMHO Safari actually has the best hardware acceleration. Chrome-dev channel is catching up quickly.

By hardware acceleration I'm talking about things like -webkit-transform and the support for various 3D transforms. I'm unsure of how it's implemented under the hood but I know Safari kills on 3D transforms and the best part is it works on mobile.


Mere support for CSS transforms is actually really limited. What IE9 and Firefox 4 do on Windows 7 is accelerate all drawing and layer compositing. What it means is that as the browsers gain support for CSS animations and such, they get accelerated for free. Stuff like Canvas and SVG also get accelerated. See http://hacks.mozilla.org/2010/09/hardware-acceleration/.


It's good that Microsoft is viewing integration into it's OS stack as a way to squeeze out performance but if that also doesn't translate into an accelerated ship schedule then who cares.

One year is an eternity.


FTA: "... CSS3 Multi-column Layout (link), CSS3 Grid Layout (link) and CSS3 Flexible Box Layout (link), CSS3 Gradients (link), and ES5 Strict Mode in action. We also demonstrated additional standards support (like CSS3 Transitions (link) and CSS3 3D Transforms (link)) that will be available in subsequent platform previews of IE10, which we will update every 8-12 weeks."

Oh, hell yes! CSS gradients and animations/transforms? Christmas came early this year. This version can't come soon enough.


Where in that list do you see CSS animations? I'm old and have difficulty with my eyes sometimes.


Transitions and 3D transforms are animations.


When will I be able to use the latest CSS and HTML features without vendor tags in any browser? I'm tired of -wekbit -moz -o and -ms.

I won't count any of them as supporting stuff until it works without vendor extensions, and I can finally stop saying things in quadruplicate.


You will be able to use them without those prefixes as soon as the W3C agrees on the final implementation of a standard and as soon as the respective browser implements said standard perfectly.

Browser prefixes are actually a great idea that save us from the madness of the old days (like browser hacks) and there is nothing bad about them.


Isn't the reason for the vendor tags that the features aren't part of the official spec? I'd guess the answer would be once they become official or you start generating your CSS. Doesn't even have to be that fancy, just something to run over a css file and duplicate a declaration a few times in all the different formats.

Probably something you should look into, I don't expect there to come a point where vendors stop adding new proprietary properties, at least not this decade.


Those properties are not proprietary (the vast majority at least). There is a draft spec for most of them but agreement on the final implementation doesn't yet exist. Browser prefixes give vendors the ability to test said draft spec in the wild and correct course if necessary. They also make sure that differing implementations don't make the lives of web developers hard for decades. They are a great idea, actually.


I guess proprietary was the wrong word for what I was saying about the properties. It wasn't my aim to imply that they were proprietary features, just a lack of knowing a better term to refer to the property names.


You'd think the 'best native support' would have WebGL.

Oh well, I just hope the History API makes it soon so we can stop abusing #anchors in a few years ..


Going by the rest of the article, I'm assuming by native HTML5 they mean hardware accelerated compositing? Good for them if so, but 'native' is definitely the wrong word to use, and the development channel of Chrome has supported hardware accelerated compositing for quite some time.

https://sites.google.com/a/chromium.org/dev/developers/desig...

Can even try it out in the current stable channel if you enable it: about:flags

IE10 and WebGL is the big question for me, which will be interesting to see pan out given their obviously conflicting position with OpenGL.


By 'native' they mean built for Windows without any cross-platform framework getting in the way.


Unfortunately, IE 10 is at least a year to two years away from any significant marketshare.


It still creates buzz. The biggest problem is getting people to upgrade. IE6/IE7 can't die soon enough. People on XP who are stuck on IE8 might start feeling abandoned and switch to Chrome or Firefox.


It's a year to two years away from release, much less marketshare.


FlexBox, transitions, gradients, and 3D transforms... glad to see progress continuing to be made -- and they were demoing it on an ARM processor. Cool.


I've seen people mention that the demo runs on ARM, but I can't see how that works.

As there is no build of Chrome for Windows ARM as far as I know. And the things isn't available from Mozilla either, which I assume would be the Firefox-button in the tray at the bottom.


It's at about 2:03:48 in the Mix stream. There is no Chrome or FF logo in the taskbar. Its the machine they use to do the last demo. It's not the machine used in the picture on the website linked to from this story.


I've seen the video on the blogpost, I didn't see it there.

I'll take a look.


Maybe we'll get text-shadow with IE10


It's not in the platform preview 1. Neither are css transitions or 3d transforms, but their messaging indicates they are coming in a subsequent PP release.


I'm not a regular Windows user, but it seems like many even more tech savvy users continually fall for this same cycle of BS from MS. The users frustrated with IE 7 were excited about how much better IE 8 would be. It turned out, IE 8 had many of the same rendering weirdness that IE 7 had. Then came IE 9 to save the day. Only it doesn't do nearly as well supporting HTML 5 features as Chrome, Safari and Firefox.

So now here comes IE 10, and what's the pitch? It's gonna be damn fast. Who gives shit? Chrome and Safari are blazing fast AND you don't have to through all kinds of hacks, and html5.js at them as they work perfectly as is, with nearly ALL of the HTML 5 specifications.


I wish that Microsoft would just use WebKit in IE10. It's just the rendering engine, they can MSN and Bing up the chrome as much as they want.


"Our legal terms have changed since your last visit. By agreeing to the legal terms and requirements, you can continue making contributions to MSDN. Your profile is linked to from all contributions that you make, so others can learn more about you. It also shows your recent social activity and your medals/points."

I have to agree to this to read an article? No thanks.


I posted this on their comments: "Please, Microsoft, don't try to confuse people by hijacking the word "native". An HTML 5 application does not run on a magical HTML 5 CPU, even in IE9 and IE10. Being single platform may allow you to develop IE faster but it is not an advantage for the user."


Is that Rob Mauceri speaking in the Video? I saw him at a round table at the Web 2.0 Expo last month and Douglas Crockford and Alex Russell tore him a new ass about Ecma Script 5 strict mode. Looks like he got out of the building alive and actually listened. Cool


Does it support the History API?


No mention of this nor text shadow.


Native is to browsers as Open is to Mobile OS. That is a great catchy phrase that will get the tech blogs lots of page views while the fans argue about whose definition is more correct.


IE in a nutshell: http://i56.tinypic.com/311s410.jpg

The damage has been done. The legacy of IE won't be fixed with a new version number.


IE10 Platform Preview 1, available for download today is the first step...

The only native experience of the Web and HTML5 today is on Windows 7 with IE9.


Comments on that page killed the browser-star. But I appreciate Microsoft for not removing them.


I am liking the pugnacious tone. I suspect that with IE9 shipped, they feel they've earned the right to make a forceful argument on how browser development should happen.

"Hey kid. Releasing a new version every few weeks ain't professional."


I thought that was called "Agile Development." What's unprofessional about it? I'd say it's a lot more professional than letting your browser stagnate for years, leaving security holes unpatched, etc.

As for "how browser development should happen," I think the people who can make the argument on that are the ones who've been leading the pack for the last 5+ years, not the ones who are still trying to catch up despite starting out with a giant lead.


I think your sarcasm filter might be malfunctioning. Everything you just said was already implied by what @innes wrote...


"Unprofessional" is not the same thing as bad, and I wish people would use more descriptive terms in its stead. What, exactly, is bad about having several release streams with varying degrees of speed and stability?




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: