Site Ready WebSockets


The Web gets richer and developers are more creative when sites and services can communicate and send notifications in real-time. WebSockets technology has made significant progress over the last nine months. The standards around WebSockets have converged substantially, to the point that developers and consumers can now take advantage of them across different implementations, including IE10 in Windows 8. You can try out a WebSockets test drive that shows real time, multiuser drawing that works across multiple browsers.

What is WebSockets and what does it do?

WebSockets enable Web applications to deliver real-time notifications and updates in the browser. Developers have faced problems in working around the limitations in the browser’s original HTTP request-response model, which was not designed for real-time scenarios. WebSockets enable browsers to open a bidirectional, full-duplex communication channel with services. Each side can then use this channel to immediately send data to the other. Now, sites from social networking and games to financial sites can deliver better real-time scenarios, ideally using same markup across different browsers.

What has changed with WebSockets?

WebSockets have come a long way since we wrote about them in December 2010. At that time, there were a lot of ongoing changes in the basic technology, and developers trying to build on it faced a lot of challenges both around efficiency and just getting their sites to work. The standard is now much more stable as a result of strong collaboration across different companies and standards bodies (like the W3C and the Internet Engineering Task Force).

The W3C WebSocket API specification has stabilized, with no substantive issues blocking last call. The specification has new support for binary message types. There are still issues under discussion, like improving the validation of subprotocols. The protocol is also sufficiently stabilized that it’s now on the agenda of the Internet Engineering Steering Group for final review and approval.

Site-ready technologies

The Web moves forward when developers and consumers can rely on technologies to work well. When WebSockets technology was shifting and “under construction,” we used HTML5 Labs as a venue for experimentation and feedback from the community. With a prototype we gain implementation experience that leads to stronger engagement in the working group and the opportunity to collect feedback from the community, both of which ultimately lead to a better, and more stable, design for developers and consumers. We’re excited and encouraged by how HTML5 Labs helped us work with the community to bring WebSockets to where it is today.

— Brian Raymor, Program Manager for WebSockets

Comments (52)

  1. frymaster says:

    Much as I hate people who hijack random blog articles, is there any plan to improve the RSS reader in IE for IE 10? I'm mainly thinking of a way to tell there's a feed available without needing to have the command bar open…

    (making the rss platform roaming-profile compatible would be nice as well…)

  2. Tom says:

    Great news, thanks Microsoft!

  3. @IEBlog says:

    Please insert Microsoft Translator Widget. Thanks.

  4. PhistucK says:

    Why can you not follow you own recommendations?

    I went to see the test drive and it told me that my browser was not supported.

    What browser am I using?

    Chrome 16.0.882.0 canary.

    Does it have support for the latest version of the WebSocket protocol?

    Yes, it does.

    Does it have support for binary frames, in case it matters?

    Yes, it does.

    Why is it not supported?

    function isBrowserChrome() {

       try {

           //debugString(navigator.userAgent.toLowerCase());

           return (/chrome/.test(navigator.userAgent.toLowerCase()) && /15./.test(navigator.userAgent.toLowerCase()));

       }

       catch (ex) {

           return false;

       }

    }

    (In non developer words – because you are using browser detection rather than feature detection.)

  5. Mario says:

    One of the Modern day Web Browser standards I would like to see Internet Explorer to support in version 10 are.

    •WebM (If possible)

    •Support for text shadows

    •More rounded tabs in the User Interface instead of a rectangle shape tabs like

  6. Chocolate Face Monster says:

    wheee!

  7. Commercial says:

    you know what is sad.. Microsoft is got spell checking in IE.

  8. johnnyq3 says:

    @Mario

    Text shadows are supported in IE10 PP3 which is included with the Windows 8 dev preview.

  9. Mario says:

    @Johnnyq3 Oh, I haven't used the Preview Cause I dislike platform previews..so i didn't know. Thanks!

  10. @mario says:

    Rounded interface elements are so 2001 (as in the XP teletubbie time)

  11. Me says:

    @Mario

    You can't be that interested in IE if you haven't even bothered to use any Platform Previews.

    @@Mario

    The slightly-rounded tabs in IE7 and IE8 looks fine to me. It's a pity Microsoft didn't allow the IE8 executable to launch but use the IE9 rendering engine (like it did with IE7 where an IE6 exe was retained inside a folder inside C:Windows). I'd love to upgrade to IE9 or IE10 but the lack of a separate search box, how empty the status bar is and how ugly it looks when tabs are put below the address bar and the Favourites Bar is enabled just puts me off. So currently I just stick with a mix of IE8 and Firefox 6.0.2.

  12. Mario says:

    Hello again, rounded tabs aren't so 2001… The Square like tabs looks alright but… round tabs looks better

  13. @Mario says:

    Lack of support for rounded tabs is yesterday news. Now they are giving support. What else do you want?

  14. Dave says:

    Server-Sent Events went Last Call in March.  Will they make it in too, I wonder?

  15. @Mario says:

    You can get WebM support for IE here:

    tools.google.com/…/webmmf

  16. Ralph says:

    @@Mario – I'm pretty sure that @Mario meant native out-of-the-box-by-default support.  We don't want to have to install a plugin to support a video format designed for the Web.

  17. hAl says:

    WebM with VP8 codec is bad for the environment.

    It lacks hardware support on all current devices which is making making playback on devices very energy inefficient.

    To put this into perspective let say that we use Youtube as an example.

    Youtube serves about 2 billion video's a day of about 3 minutes average lenght.

    A VP8 encoded video probably uses about 5 watt more to play on an average computer.

    This means that for the users of the Youtube site alone the usage of WebM exclusivly would amount to 100,000,000 Watt/hour extra electricity use a day.

    A pure waste of energy as there is no need to use that inefficient codec.

  18. hAl says:

    Shortened the previous comment compared to an original comment which was lost in the 'comment hell' here.

  19. hAl says:

    Lost site of the math in retyping the comment.

    Should be 500,000,000 Watt/hour lost for users of Youtube if they would serve WebM only.

  20. DanglingPointer says:

    FWIW, Acid3 on pp3 scores 96, html5test scores 300(+6 bonus) and DOM manipulation (nontroppo.org/…/Hixie_DOM.html) is sill very slow! 45038ms !! Reported this issue at the time of pp2 connect.microsoft.com/…/a-dom-manipulation-test-ie-performance

    IE team, I love the way you guys implement not only the spell checking but the auto-correct! Please revise the way you guys manipulate DOM.

    P.S. I tested these result from desktop browser of win8-dev-preview. (because metro browser is not launching even after the fresh install)

  21. Eduardo Valencia says:

    Dea IE developer team

    Sometimes when i have several tabs open in IE9,i close the tab i will use by mistake,it is possible to add a "Lock function" in the tab to avoid this type of situation?

    It can be a great feature right?

    Thank you

  22. DanglingPointer says:

    IE10pp3 scores 100% on ACID3… i updated windows dev preview, restarted my computer and checked on BOTH metro and desktop versions. MS rules! 😀

  23. Gérard Talbot says:

    @Phistuck

    > Why can you not follow you own recommendations?

    > (…)

    > you are using browser detection rather than feature detection

    I fully agree with you. Microsoft IE blog often makes recommendations that all of microsoft.com, including IE test center, including IE blog, including MSDN do not follow and do not comply with. The culminating point of such blatant incoherence, inconsequence is to be invited to ditch Windows XP and IE7, to buy Windows 7, to install IE9… just to view the exact same webpages (always with invalid markup and invalid CSS code) at microsoft.com which start with

    <meta http-equiv="X-UA-Compatible" content="IE=EmulateIE7" />

    or

    <meta http-equiv="X-UA-Compatible" content="IE=EmulateIE8" />

    Microsoft webpages say, recommend and claim things that it denies to itself, by itself, for itself.

    Microsoft asks and requires from others to do things that it refuses to itself.

    Gérard Talbot

  24. meni says:

    From the blog:

    "The Web gets richer and developers are more creative when sites and services can communicate and send notifications in real-time. WebSockets technology has made significant progress over the last nine months. The standards around WebSockets have converged substantially, to the point that developers and consumers can now take advantage of them across different implementations, including IE10 in Windows 8."

    IE team, you could at least apologize for the FUD you were spewing about a year ago regarding Websockets, Mozilla and chrome/webkit. I am of course talking about this seemingly funny video, which was pure unnecessary propaganda:

    http://www.xtranormal.com/…/web-sockets-we-are-the-first

  25. meni says:

    One more thing, regarding this blog software, I once saw a t-shirt with "friends don't let friends use web-forms" printed on it 🙂

  26. @meni says:

    you are nothing but a troll. Live with it!

  27. iSaad says:

    i don't use IE9 because it saves pages slow, and when is saving a page, i can't scroll page and i can't switch tabs. Please fix in IE10.

  28. @meni says:

    go away troll. that "unnecessary propaganda" is 100% true.

  29. meni says:

    I am a troll.. whatever kind of software MS come up with, I will keep trolling… becuase I am steve-jobs' buddy 😀

  30. Klimax says:

    @Meni: Well, until you apologise for baseless accusation and missinformation, then I don't think you have rigth to demand anything from them.

    Like saying "FUD", when in fact they linked to that article which directly contradicts your comment, making your comments mainly FUD-like.(ironically) At least if you tried it with blog post, which didn't link to previous one for anybody to see…

    So when do you apologise?

  31. meni says:

    I'm a pathetic troll and a pathetic and f***ing loser. Lol. I'm just so sad that LinSux has only 1% market share and Windows 7 has 90% market share. Ha ha, open source sucks.

  32. Xero says:

    IE-team. Nice work coming with IE10. I have two concerns:

    1- Please improve the spellchecking as its relatively slower as well as if you manually correct the word, it doesn't recheck it unless you press space after the word. Imo, the spellchecking thread should be continuous and fast-and-fluid (optimized).

    2- Here is the test of innerHTML vs add for select options => jsperf.com/options-add-vs-innerHtml . Compare he result in he browserscope graph. You guys need to look at the Safari performance in that test. Is it related to the one reported by DanglingPointer (nontroppo.org/…/Hixie_DOM.html)?

  33. Tom says:

    The change in the score for Acid3 is a result of changes in the Acid3 test and not IE10pp3.  Don't believe me?  Try running Acid3 in IE9.  You'll get 100 now.

  34. Tom says:

    Sorry about the duplicate post, I thought that the comment system has lost my comment.  Nope, it was just being slow.

  35. @Tom says:

    Same goes with Firefox. This means Acid3 is corrected after implementing the correct w3c standards implementation. Chrome 12 is failing 1 test now, well seems like they were implementing some non-standards alongside just to pass some joe's test!

  36. TomG says:

    @Tom

    changed my name to avoid confusion.  Tested GC14 and it passes with 100.  Also Firefox 5.  That's all the browsers that I have installed. So, now it's time to move forward and find new dragons to slay.

  37. Nathan says:

    I only have 1 comment.  FIX this Blog Software!

  38. Gérard Talbot says:

    @hAl, Tom and Nathan

    The problems with the IE blog software are

    1- it allows unmoderated comments, even name-calling types of comments, off-topic comments, purely disparaging comments, abusive comments, uselessly rude, abrasive comments, advocacy comments, etc

    2- it allows total anonymity and anonymity breeds contempt

    3- it often does not work, it hangs, it fails, it is slow, etc

    hAl, Tom and Nathan: please find the link "Email Blog Author" in this page and send an appropriate request. If you lack inspiration, just copy and paste this message.

    Gérard Talbot

  39. Todd says:

    @Gerard – point well taken, but if you've emailed the blog author like we have, you'll know that the email doesn't get any attention paid to it… a reply or resolution.

    The viewers of this blog thus indicate in the comments (on their 2nd or 3rd try) that the comments on the blog are busted.

    It serves several purposes.

    1.) It gets visual attention from all readers – thus new posters learn to be weary

    2.) The squeaky wheel gets the grease – if someone at Microsoft finally pays attention, someone at Microsoft will finally realize that the blog software sucks and that the longer they don't address the issue, the longer it continues to ruin Microsoft's reputation as a software corporation committed to the web.

    3.) The complete lack of response helps other developers realize that Microsoft is slow to respond in general – don't be surprised when bug reports filed on Connect don't get any attention either.

    4.) It highlights to new developers that there are either flaws with the .Net platform or the Community Server software – either way, there is likely better alternatives to developing web applications on other platforms in better languages.

    PS In case it shows up soon, various comments about this software/blog (and the lack of a fix in 5 years) have been submitted to TheDailyWTF.com it shouldn't have come to this, but if you don't make an effort – you can't be surprised.

    Todd

    (this is post attempt #2)

  40. yellowstone says:

    IE10

    *.jxr(JPEGXR) Image support support

  41. @yellowstone says:

    <b><i>Submit the feature request at connect.microsoft.com</b></i>

  42. Gérard Talbot says:

    @Todd

    I got a reply from Ted Johnson [MSFT]. I could post the content of such email but this would not be polite and fair. I would rather have him or someone responsible for IE blog management and IE blog software voluntarly make a statement on their own and by themselves.

    > if someone at Microsoft finally pays attention,

    I think someone at Microsoft was already aware of problems in IE blog software as soon as 2006 or 2007. I remember Eric Lawrence [MSFT] mentioning known issues with the IE blog.

    > someone at Microsoft will finally realize that the blog software sucks

    > and that the longer they don't address the issue, the longer it continues to ruin Microsoft's reputation as a software corporation committed to the web.

    I agree with you. It's kind of stupid to be trying to demonstrate and convince people that IE10 may be technologically speaking and web standards-wise speaking the best browser around … when its IE blog is frustratingly bad, its underlying code mediocre and non-compliant.

    > The complete lack of response helps other developers realize that Microsoft is slow to respond in general – don't be surprised when bug reports filed on Connect don't get any attention either.

    From my own experience, I would say bug reports filed on connect got attention; the most urgent and most important ones were addressed accordingly.

    I stopped reporting bugs and issues at connect IE beta feedback because such Bug Tracking System was not well designed, was not a model of good working interoperable code and it was frustrating to work with. Microsoft should improve their general webpage coding habits and clearly+definitely move toward *using* the web standards they claim they implement in their IE9+.

    > there is likely better alternatives to developing web applications on other platforms in better languages

    Yes, there are better alternatives. I could easily list well over 50 Bug Tracking system websites from big/major companies which can list bug reports without requiring a login, without requiring registration, which allow searching through the bug reports without requiring registration.

    Any/all Bulletin Board websites generally do that too.

    Gérard Talbot

  43. steve_web says:

    @Gérard Talbot, @Todd

    Its sad to say but I agree with both of your comments. I too used to submit bugs to Connect but I gave up.  I found the UI/Search to be frustrating to use and about as far from "inviting collaboration" as a system could be.

    This blog suffers the same fate.  The issue with posting comments has been known for 4-5 years yet little (if anything) has been done about it.  There was a re-vamp at one point to improve the styles and hook in a *slightly* better search but the underlying problems with commenting, too-long index page posts, and archive titles vs. useless dates still exist.

    @Sean Jenkin [MSFT] did momentarily drop into the conversation once to "look into" the issue but he was not heard from since.

    The problems all still exist to this day (and as I type this, I know that I have to copy my comment to the clipboard because I'm willing to bet $50 that my comment will not be saved – that's how well I trust the IE blog with comments)**

    The other issue (which this comment will suffer the fate of) is that once a new blog post has been made, comments on the previous thread effectively die and get no attention from Microsoft (from my experience)

    All of this is quite sad because I actually enjoy coming here to be informed of the changes coming up and the discussions about implementations that we might otherwise not understand.

    Steve

    ** yup! comment was lost – surprise, surprise…. man I wish there was someone here willing to take the bet – I could make a fortune!

  44. Sean Jenkin says:

    Hey, I'm still here. I haven't forgotten about looking at it. Now let's see if this comment posts.

  45. SeanJenkin says:

    It did, that's good to know. 🙂

    Now for those of you who have frequent trouble posting to the IE blog, do you have reproducable steps that demonstrate the problem? Can you document them? Can you screen capture the experience? I'd like to be able to help work out what's going on, but I need data to go from.

    For those of you it's slow for, can you do a network trace so we can see what is slow?

    You can reach me by going here: blogs.msdn.com/…/contact.aspx

  46. alvatrus says:

    @Sean Jenkin

    The comment will not be saved if you spend more than a certain amount of time on the page after refresh.

    Someone here has already experimented with this and determined that 30min. will do the trick.

  47. hAl says:

    The loss of comments is more common the longer they are. Probably also relates to the time spent on the page.

    Double posts frequently happen because after posting nothing happens for 30 seconds (or more) and I reclick on [Post] button until the page makes some kind of indication that the button has been activated. The blog has no double post detection nor any other basic filtering that most common blogs have.

    Anoying is that while you are posting below all the comments the blog gives a "waiting for moderation" message above the comments.

    Also anoying is that the ieblog frontpage loads all media in all articles on the page leading to 5MB or more downloads for the frontpage if it has old articles with video in them even though you are not going to watch those old videos in old articles. This makes the blog terrible for mobile access.

    (repost as orginal effort has been lost, I'll try to create a double post now)

  48. hAl says:

    Oops, a quadruple post. A bit overenthausiatic clicking on the[Post] button after the original effort to post the above wast lost.

    The losing of the comment looked a page refresh directly after clicking the [Post] button.

    De double post is just waiting for an indication that you have clicked the [Post] button and then reclicking after ages of waiting just because absolutly NOTHING seems to be happening on the page.

  49. hAl says:

    Last comment it took 21 seconds for the page to react after posting (measured on stopwatch WP7 app) and then about 4 second to complelty refresh with the green "awaiting for moderation" bar but without the comment.

    Then I refrehed the page again myself to see if the reaction had actually posted which took another 8 seconds.  

    A sequence that actually felt rather fast for this blog as I have seen it slower with posting times over a minute.

  50. testing says:

    not only are comments not saving properly – but now successful comments are being deleted afterwards!

    Blog Fail!

  51. MVC4 says:

    Mycrowsoft guys. Why did you launched MVC framework at all if you are not intended to use it on your on network.

    ~ Cut the crap and use MVC4 for this simple blogging app!