IIRC, the Web Sockets implementation was something the WCF team was already experimenting with for their own purposes and originally had not much to do with IE9 or the IE team. Since it was implemented in WCF, obviously it would require a CLR (i.e. Silverlight).
Putting it on the HTML5 Labs site is kind of goofy, and the whole HTML5 Labs thing is sort of a dubious PR exercise to begin with, but I guess it does at least show they're experimenting with the Web Sockets protocol.
As someone who hates IE, I must say that I'm actually really looking forward to IE9. The "worst" of the major browsers just closed the gap (and surpassed?) with fast JavaScript and gpu acceleration.
There are ways to use svg to implement scalable gradients: http://css3wizardry.com/2010/10/29/css-gradients-for-ie9/ . With some clever engineering someone could make a library that automatically converts gradient styles into SVG backgrounds.
Useful to know, but the simple thing I've been working on - a product browser akin to coverflow, with bookmarkable items - needs a 'skew' effect, which is considered a 3D transform.
WebSockets isn't even finished yet. What's the point complaining about the lack of a feature when it's not even ready yet? It's not like all the other browsers support it either.
Is it true that it won't work on Windows XP? I know XP is like a decade old, and I really don't wanna start an upgrade-your-browser-you-noob flamefest, but it seems kinda lame to exclude the people still on XP from having a modern browser and I can't figure out what the rationale might be for that.
There's never been a good reason publicly stated, but I imagine they don't want to be seeing as supporting their decade-old OS.
Chrome, Firefox and Safari have no problem doing it, so it's very likely a political decision.
I think Microsoft has some responsibility to update IE6 on XP, given that they left it out on the market so long and caused an incalculable amount of time to be lost developing for it while the rest of the worst was trying to move on.
Actually, no. The good reason that IE9 does not run on Windows XP has been publicly stated for at least an year, if not actually longer. The FAQ on ietestdrive.com [1] answers this question. The short answer is that the Graphics stack that IE9 uses is not available on Windows XP.
If Microsoft wanted to support GPU acceleration through an alternate rendering path like two other browsers are doing on XP, they would have done so.
I find it more likely to conclude that they just didn't want to support XP and developed a browser GPU acceleration engine that wouldn't support it as a result of this high-level decision.
That's a fascinating conspiracy theory, but having looked at the code myself, I'd have to say someone must be nuts to have gone to extra efforts to create a new engine that specifically won't run on XP.
Huh? I never said they wrote the graphics stack not to support XP. I said that there's no technical reason why you can't have a fallback rendering path to work on XP. Chrome and Firefox show that this is possible.
You can point at the FAQ all day, but you can't deny that there is a way to do GPU acceleration of web content on XP.
Actually, chrome and firefox demonstrate that there is a technical reason, since they both only implement accelerated compositing on XP, not accelerated content layout.
I suspect that it is two reasons they don't have IE9 on XP:
1) They don't want people using XP. If other browser vendors want to take it on, I think MS has decided that they're fine with that. But they don't want to enable XP anymore.
2) The test matrix is a huge test cost. Doing XP means cutting something else. I think they decided that the cost of supporting XP wasn't worth it when they looked at how it effected their roadmap.
I personally wish for a crippled version of IE9 that doesn't use GPU acceleration but reliably backports to XP. But I can also see how that could be a nightmare to develop and maintain.
You don't have to support IE. Just like an iPhone developer doesn't have to support Android. Or an OS X developer doesn't have to support Windows. You give up some market share, but it's a standard tradeoff == more work vs more revenue?
There's nothing to be sad about, it's a decision we make everyday. If IE is so painful that the revenue is not worth it, then don't do it. We currently don't support OS X. I'm not sad when Apple ships a new version of OS X. I just know that the 10% of the market that they own is one that we won't support. I'm fine with that, and have been for a while now.
Here's the light at the end of the tunnel: I have a bug tracking app, and can honestly say that IE is an insignificant part of my traffic - comparable to Opera. Granted, it's a technical audience, but I can safely just not support IE.
I really don't want to have to start thinking of myself as a "Firefox Developer". If it surfs the web, be it desktop browser, mobile browser, screen reader, robot or even IE, it's part of my job as a web developer.
Doesn't mean I have to reward Microsoft's failures to keep up with standards with praise, or by abstaining from criticism. My life would be better if they just gave up at this point.
I don't use IE9 for my browsing and I am sad.
I am a web developer and now we will have 3 or 4 versions (depending on whether IE6 support is required) of Microsoft browser to care for. And they all are outdated compared to Firefox, Chrome, Opera, Safari.
sorry man, i don't know anything in web development. I'm a system programmer works on C\C++ and hardly know internal details of browsers. But IE9 is very cool(compared to ie8), i liked it. Thanks for making me aware of how newer versions affects everyone.
If solving the web development communities problems was only so simple as that, it would have been solved already. I think it's safe to say that essentially nobody here uses IE.
I'm afraid our problems with IE involves other people using it, and their choice to do so is largely beyond our control.
People will debate endlessly whether IE9 is a "modern" browser based on its HTML5, CSS3, JavaScript, etc support.
The real smoking gun that it still isn't modern is that it doesn't include a spellcheck by default. You can't claim to be doing what users want while ignoring that elephant in the room.
My guess is that the real reason for this is that Windows 8 will add a native spellcheck API (technically Vista already did, as part of WPF, but that generally isn't used by consumer applications for various reasons), and in the meantime to avoid future confusion they don't want to implement it separately in IE.
Launch, as in end of product, it's "finished", it's in the wild, you're on your own we will now work on IE 10?
Or launch like the rest of the software development world now operates and push out updates on a regular basis, with new features and better standards support, leading to something that actually has a chance of competing with the browsers the rest of us love to use?
'Cause I think it was around 2000 that people started realizing that monolithic development and release cycles lead to failed products. Software needs to be, and can be, updated constantly, or it falls behind. If this launch is like all other IE launches, it will be a day we will all look back on with disappointment. Don't get me wrong, I would love to see Microsoft actually become a modern development house, but I have huge doubts that it could ever happen.
Exactly twelve months from the first Platform Preview of IE9, on Monday March 14th we will celebrate the developers and designers who are making a more beautiful web for all of us. We will release the final version of Internet Explorer 9 for download beginning at 9 p.m. Pacific.
I guess I earned the downvotes for the bad attitude. I'm just amazed that in a time when everyone else is iterating and pushing out updates in a Big Hurry, they're announcing (celebrating?) the fact that they took a year to go from preview to release.
42 comments
[ 4.7 ms ] story [ 146 ms ] threadIt still doesn't work in Firefox4 though. Not sure if I'm surprised.
EDIT: Argh, didn't take long to find an "install Silverlight" while browsing through the HTML5 demos site:
http://html5labs.cloudapp.net/WebSockets/ChatDemo/wsdemo.htm...
Why do they keep doing this?
Putting it on the HTML5 Labs site is kind of goofy, and the whole HTML5 Labs thing is sort of a dubious PR exercise to begin with, but I guess it does at least show they're experimenting with the Web Sockets protocol.
Hopefully, there's more to come.
There are ways to use svg to implement scalable gradients: http://css3wizardry.com/2010/10/29/css-gradients-for-ie9/ . With some clever engineering someone could make a library that automatically converts gradient styles into SVG backgrounds.
Chrome, Firefox and Safari have no problem doing it, so it's very likely a political decision.
I think Microsoft has some responsibility to update IE6 on XP, given that they left it out on the market so long and caused an incalculable amount of time to be lost developing for it while the rest of the worst was trying to move on.
[1] http://ie.microsoft.com/testdrive/info/FrequentlyAskedQuesti...
http://hacks.mozilla.org/2010/09/hardware-acceleration/
http://blog.chromium.org/2010/09/unleashing-gpu-acceleration...
If Microsoft wanted to support GPU acceleration through an alternate rendering path like two other browsers are doing on XP, they would have done so.
I find it more likely to conclude that they just didn't want to support XP and developed a browser GPU acceleration engine that wouldn't support it as a result of this high-level decision.
You can point at the FAQ all day, but you can't deny that there is a way to do GPU acceleration of web content on XP.
1) They don't want people using XP. If other browser vendors want to take it on, I think MS has decided that they're fine with that. But they don't want to enable XP anymore.
2) The test matrix is a huge test cost. Doing XP means cutting something else. I think they decided that the cost of supporting XP wasn't worth it when they looked at how it effected their roadmap.
I hope Mozilla and Google take advantage of this opportunity.
There's nothing to be sad about, it's a decision we make everyday. If IE is so painful that the revenue is not worth it, then don't do it. We currently don't support OS X. I'm not sad when Apple ships a new version of OS X. I just know that the 10% of the market that they own is one that we won't support. I'm fine with that, and have been for a while now.
"Well, sir (or ma'am), your new website will only work with 'Modern' browsers."
"I see. And what is a 'Modern' browser?"
"Anything but IE."
"Ah. Nice meeting with you. I'll keep your quote in mind."
No IE support != less revenue. No IE support = no revenue.
Our day is coming...
I'm afraid our problems with IE involves other people using it, and their choice to do so is largely beyond our control.
The real smoking gun that it still isn't modern is that it doesn't include a spellcheck by default. You can't claim to be doing what users want while ignoring that elephant in the room.
Spelling words properly, just for one day
Or launch like the rest of the software development world now operates and push out updates on a regular basis, with new features and better standards support, leading to something that actually has a chance of competing with the browsers the rest of us love to use?
'Cause I think it was around 2000 that people started realizing that monolithic development and release cycles lead to failed products. Software needs to be, and can be, updated constantly, or it falls behind. If this launch is like all other IE launches, it will be a day we will all look back on with disappointment. Don't get me wrong, I would love to see Microsoft actually become a modern development house, but I have huge doubts that it could ever happen.
twelve months
twelve months
twelve months