Share via


Browsing Without Plug-ins

Because more and more browsing takes place on an ever wider variety of devices, and browsers on them, more and more consumers are browsing without plug-ins. Delivering a great site experience to consumers who browse the Web plug-in free is important work that sites must do to reach the widest audience. With HTML5, modern browsers and sites can deliver a great consumer experience even without plug-ins.

“Plug-in” refers broadly to browser extensions that run native client code using low-level browser interfaces.  For example, here is a basic description of Webkit’s approach; the IE equivalents are ActiveX controls and Browser Helper Objects. Web sites use a wide variety of plug-ins; Adobe Flash is one of the most common.

More and more common

Lots of Web browsing today happens on devices that simply don’t support plug-ins. Browsers that do support plug-ins offer many ways to run plug-in free.  IE9, for example, includes ActiveX Filtering. Other browsers offer add-ins to control plug-ins, like this one or this one. While plug-ins can be compiled to run inside 64-bit browsers, many developers only release versions that run inside 32-bit browsers; running a 64-bit browser is another way that running plug-in free is becoming more common:

64-Bit Windows includes both 32-Bit and 64-Bit Internet Explorer.
64-Bit Windows includes both 32-Bit and 64-Bit Internet Explorer.

Better and better experiences

Many sites today already offer a good experience when plug-ins are not available. For example, visiting Hotmail in IE9 with plug-ins unavailable (not installed, for example with the 64-bit browser, or disabled through ActiveX Filtering) works just fine:

Hotmail Inbox without plug-ins in IE9.
Hotmail Inbox without plug-ins in IE9.

Some sites require some consumer action before they work plug-in free. For example, YouTube without plug-ins works after visiting https://www.youtube.com/html5 and clicking “Join the HTML5 Trial”:

YouTube without plug-ins in IE9. On the left before, and on the right after, joining the HTML5 Trial. The context menu on the right shows that the video playing uses HTML5.
YouTube without plug-ins in IE9. On the left before, and on the right after, joining the HTML5 Trial. The context menu on the right shows that the video playing uses HTML5.

Other sites block some or all functionality with plug-ins disabled. For example, MSNBC.com and CNN.com work except for the videos; Gmail currently blocks IE9 with plug-ins disabled because it checks for the XHR ActiveX object rather than using the Web standard XHR that’s been available since IE7. The Web has evolved quite a bit since IE7 and so sites will want to go back and revisit code that is specific to older browsers or old versions of standards.

Showing how some sites disable some or all functionality without plug-ins.
Some sites disable some or all functionality without plug-ins.

Feature Detection: How site developers can make this even better

Many sites already offer great experiences when browser plug-ins aren’t available. There is a problem for consumers when sites do this based on the particular device or browser in use. For example, MSNBC.com videos work without plug-ins in a browser on a PC sending the User-Agent string of a device, but not in the same browser on the same PC sending a different User-Agent string:

MSNBC.com videos don't work in Apple Safari without Flash installed, and work fine on the same browser identifying itself as the iPad.
MSNBC.com in Apple Safari without Flash installed, and on the same browser identifying itself as the iPad.

Consumers are better off when developers use feature detection and fallback rather than hard-coding a site to specific browsers and configurations. For example, detecting and using HTML5 video in the absence of plug-ins gives the consumer a better experience. Many sites already perform the equivalent of this fallback when serving ads in the absence of plug-ins, showing that this approach is a practical and scalable solution.

The consumer experience is better when sites follow best practices to test for the standards-based feature first and fall back to plug-ins only if necessary. For example, here are good and bad patterns for feature detection of XMLHttpRequest:

// BAD PATTERN: Don't do this!

var xhr = window.ActiveXObject

? new ActiveXObject("Microsoft.XMLHTTP")

: new XMLHttpRequest();

 

// Best Practice: Use Native XHR, if available

if (window.XMLHttpRequest) {

// If IE7+, Gecko, WebKit: Use native object

var xmlHttp = new XMLHttpRequest();

}

else if (window.ActiveXObject) {

// ...if not, try the ActiveX control

var xmlHttp = new ActiveXObject("Microsoft.XMLHTTP");

}

else {

// No XMLHTTPRequest mechanism is available.

}

Falling back gracefully to other content is important to avoid consumer frustration. For example, Gmail video chat degrades gracefully in the absence of Adobe Flash. Some sites provide fallback only for specific devices based on the user agent string or other browser-specific properties. This short-term solution doesn’t account for changes in browsers and how consumers browse.  Building an app is another approach that some sites (for example, IMDB.com) take for browsers on devices that don’t support plug-ins. User agent strings have not been a reliable or robust way to determine the specific HTML and script to serve for quite some time.

HTML5 capabilities make it possible for consumers to experience the Web today on more devices and in more configurations than ever before.  For Web developers, this means more opportunity for people to visit their sites, and more motivation to make their sites work even if plug-ins are not available.

—John Hrvatin, Program Manager, Internet Explorer

Comments

  • Anonymous
    August 31, 2011
    The comment has been removed

  • Anonymous
    August 31, 2011
    The comment has been removed

  • Anonymous
    August 31, 2011
    This is very cool. What about that "disable your plugins" message?

  • Anonymous
    August 31, 2011
    Can anyone explian me for what possible reason, IE9 does not display the video when we click the button "Edit and Click Me" (whether we change anything or not...) at; www.w3schools.com/.../tryit.asp ??? Even if we open the embedded iframe (www.w3schools.com/.../tryit_view.asp) seperately, IE9 is unable to launch the video...

  • Anonymous
    August 31, 2011
    The comment has been removed

  • Anonymous
    August 31, 2011
    The comment has been removed

  • Anonymous
    August 31, 2011

  • Resolution for the error reported in my prior comment - For www.w3schools.com/.../tryit_view.asp, try refreshing page couple of times then it displays the video. For www.w3schools.com/.../tryit.asp, (after updating the code and clicking the button and getting the redcross) right click the screen near the video area and click refresh (to refresh that specific frame) and the updated video shall be displayed. This isn’t cool, but it works this way. I am trying to confirm the support for "media" attribute of HTML5's source tag in IE9. Can anyone confirm if the source tag's media attribute works fine for IE9? Any reference or testbed to confirm it’s support? The best source I get is MSDN library msdn.microsoft.com/.../ff974686(v=VS.85).aspx. But it doesn’t tell if the browser under consideration support all aspects of this attribute listed here www.w3schools.com/.../att_source_media.asp
  • Anonymous
    August 31, 2011
    @Peter_Pan There is a Recommended Windows Update to address blurry fonts in IE9 (and other applications) using DirectWrite. Eric Lawrence wrote an article on it here: blogs.msdn.com/.../ie9-cleartype-improved-clarity-for-tahoma-verdana-and-arial-fonts-fuzzy-blurry.aspx You can download the update manually here: support.microsoft.com/.../2545698

  • Anonymous
    August 31, 2011
    I like the cross browser philosophy but when is the IE Blog going to implement it on this site? None of the videos posted on the blog work in Firefox without plugins as there are never any WebM versions of the video. On the IE Testdrive site the Santa's Workshop, HTML5 Blizzard, and Preschool demos have audio but not in formats supported by Firefox. Strangely, the HTML5 Blizzard and Preschool demos refer to Ogg Vorbis and wav audio respectively in the page source but the files don't exist on the server. Follow John Hrvatin's example and make cross browser browsing a reality on these sites.

  • Anonymous
    August 31, 2011
    And the Fireflies demo on the IE Testdrive site has the same problem as the other demos: no video compatible with Firefox or Opera. It's really easy to fix and, after all, "[d]elivering a great site experience to consumers who browse the Web plug-in free is important work that sites must do to reach the widest audience." And we mustn't forget that "[f]or Web developers, this means more opportunity for people to visit their sites, and more motivation to make their sites work even if plug-ins are not available." You talk the talk, but can you walk the walk?

  • Anonymous
    August 31, 2011
    Can we have a built in PDF reader, and Word (doc + docx) reader, so we can access most web content without the need for plug-ins or apps.

  • Anonymous
    August 31, 2011
    The comment has been removed

  • Anonymous
    August 31, 2011
    The comment has been removed

  • Anonymous
    August 31, 2011
    Also IE team get IE10 to get a 100/100 on the acid3 test! Don't release IE10 into final release till you guys get it to 100/100. please!

  • Anonymous
    August 31, 2011
    certainly you are just another hatter... let me prove it to the audience Do you have any idea whats the difference between Safari, FF, GC and IE in terms of HTML5 video support ? GC is (completely) dropping the support for H.264 soon (blog.chromium.org/.../html-video-codec-support-in-chrome.html) and will continue support for OGG Theora and WebM VP8. FF is not supporting H.264 (at all) but the other two as GC. Safari is supporting H.264 and other two via manual install just like IE. None of the browser is giving you leeway to enjoy all/any kind of HTML5 video format.   So your "iPhone" and "iPad" runs Safari which has the same combination of support as IE does. Again, you are just another hatter with poor homework done for bashing the product you will never like whatsoever. Now, may be you are the neutral guy. Did you extend the same sort of comment on Safari blog/forum? No you didn't .. because hatter will remain the hatter whatsoever.

  • Anonymous
    August 31, 2011
    @ Lex Mitchell Thx for trying to help me. But the update you suggested (kb2545698) has no effect - I still get blurry fonts in IE9. (Software rendering turned on or off) IE8 screenshots (zoom in): imageshack.us/.../ie8x.jpg imageshack.us/.../ie8b.jpg IE9 screenshots (not as blurry as I remember them, but still not good) (zoom in): imageshack.us/.../ie9c.jpg imageshack.us/.../ie9b.jpg Still no luck. Back to IE8 again :(

  • Anonymous
    September 01, 2011
    The comment has been removed

  • Anonymous
    September 01, 2011
    The comment has been removed

  • Anonymous
    September 01, 2011
    Even I am of the same opinion that the message saying 'Disable your plugins' should be around :) Good to see HTML5 gathering steam and taking care of all users from the world of plugins! But any <a href="www.mohanbn.com/.../">solutions to get rid of Ads?</a>

  • Anonymous
    September 01, 2011
    The comment has been removed

  • Anonymous
    September 01, 2011
    I take that back. I found 2 websites where the font was OK. I give up - back to IE8 again. 3 hours wasted. (Sorry for my off-topic postings)

  • Anonymous
    September 01, 2011
    @Peter_Pan: You might want to check graphic card's drivers - nothing is bug-free... (Must note that was generally unable to reproduce those issues.)

  • Anonymous
    September 01, 2011
    My wishlist, from a user's point of view, for IE10:

  • An "Open image in another tab" command when right-clicking on an image in a webpage
  • A "Past and Go" command when right-clicking on the address bar
  • A built-in PDF reader
  • A more obvious way (ie. a checkbox in the settings window's privacy tab) to enable the "Do Not Track HTTP Header" The first two features should be really easy to implement and would be great for usability. Every other major browser has them and I happen to use them all the time; I really miss them when running IE9. Also, I don't know if it is a bug or just the way it is implemented, but in IE9 the arrow-shaped mouse cursor keeps blinking when typing in the address bar. This has never happened in any other application and it is a little bit annoying. Please fix this in IE10. Thank you.
  • Anonymous
    September 01, 2011
    more to Andrew's wishlist: Ctrl+Click_Back_Button should open the previous page in new tab (same goes to forward button)... many browsers support this shortcut.. would be great if IE10 enable this shortcut.. Also the CTRL+V & CTRL+Shift+V on IE's download-manager to create a single and multiple-downloads entries, respectively, from the clipboard: connect.microsoft.com/.../create-download-in-ie-download-manager

  • Anonymous
    September 01, 2011
    You're forgetting that Silverlight 5 today released a 64bit version of their plugin.

  • Anonymous
    September 01, 2011
    @Peter_Pan. Could you please contact me via email.the alias is IE9PQ and the domain Microsoft.com

  • Anonymous
    September 01, 2011
    .. i wonder if we ever have smooth webpage scrolling in IE, like the one recently introduced for SL4 for WP7... connect.microsoft.com/.../ie-10-ui-smooth-scrolling-and-better-tab-management Same goes for Windows Explorer...

  • Anonymous
    September 01, 2011
    Browsing without plug-ins, Microsoft says! I wonder, now. That is of course because plug-ins for Internet Explorer are impediments instead of bless. However, if you must look at the competition, add-ons in Mozilla Firefox are one of the primary sources of its popularity. Microsoft should take plug-ins more seriously. A lot more seriously...

  • Anonymous
    September 01, 2011
    Folks eager to learn more about the next release of Internet Explorer will be happy on September 13 when the //BUILD/ Windows conference (http://www.buildwindows.com/) commences in Anaheim, CA.

  • Anonymous
    September 01, 2011
    @Morten Was it the plugin or just the SDK RC?  I haven't managed to find the actual plugin.

  • Anonymous
    September 01, 2011
    The comment has been removed

  • Anonymous
    September 01, 2011
    Fixed: if (window.ActiveXObject) alert('Upgrade your browser please, thanks');

  • Anonymous
    September 01, 2011
    Testing to see if I can figure out why commenting on the IE Blog is broken (not the spam above) as many attempts to post are completely ignored.

  • Anonymous
    September 01, 2011
    Just testing again, noticing some odd parameters past in the post back as undefined.

  • Anonymous
    September 01, 2011
    what is with the Spam....what microsoft can't afford spam filters on the Internet Explorer blog or something?

  • Anonymous
    September 01, 2011
    @Steve Did you have a link in the posts that were lost in posting ?

  • Anonymous
    September 01, 2011
    Test #5 - I think I've figured it out, there's something that is checking the time between the page load and the page submit.  If you spend "too long" reading the page it automatically ignores your posted comment. Whatever this value is - the check for it should be completely removed (or the logic reversed)... e.g. if the submission time is within 5 seconds of the page load, then yeah, likely a spam bot.

  • Anonymous
    September 01, 2011
    @hAl - comment tests #3 and #4 were both lost... this one is #6.  I can't post a link to the comments that were lost (e.g. they weren't lost after being successfully added - they were never accepted/received by the system) The comments were all "normal" comments without any spam type references, link dumping etc.

  • Anonymous
    September 01, 2011
    Test #7 - Waiting 5min between loading and posting a comment.

  • Anonymous
    September 01, 2011
    Test #8 waiting 20min between loading and posting a comment

  • Anonymous
    September 01, 2011
    Steve... keep up the good work for the betterment of humanity... but I doubt if MS guys are even willing to spend a minute to fix this bug ever. Its being years people complaining about this issue.. i guess some old guy from MS logs into his system and publish the comment, that was previously authored in MS Word or WL Writer, and logs out... That’s the only experience MS has with these rusted, abandoned and orphaned blogging systems.. To the guy disguising as ieblog, if you are really giving some attention or "following up" with the comments, I have a question for you; will you guys fix this issue: connect.microsoft.com/.../a-dom-manipulation-test-ie-performance when/ETA??

  • Anonymous
    September 01, 2011
    Microsoft: you guys to Listen to the users of Internet Explorer.. without them there might not be an Internet explorer.

  • Anonymous
    September 01, 2011
    Test # 10 I waited more than 30 minutes on test #9 before posting. and it failed miserably. Again, I'm not sure where the exact threshold is, but please remove it - it is massively frustrating!

  • Anonymous
    September 02, 2011
    The comment has been removed

  • Anonymous
    September 02, 2011
    @Danglingpointer The DOM manipulation test makes so many DOM manipulations that it is not a realistic real world scenario. So why would IE optimise to support that weird scenario.

  • Anonymous
    September 02, 2011
    @hal, one the other blog, where they are introducing Windows8 and talking about the programming paradigm for desktop-based win8 apps would be supporting HTML5+CSS3 for the enterprise/commercial apps... most of the people would agree that DOM manipulation is one thing which should "atleast" be as efficient as other browsers (and which should NOT be "240 times" slower than safari), because when it comes to commercial apps, user requirements are considered subjective and the "non-real-world" scenarios you are inkling upon, can be the very requirement of the potential client --- Exactly that’s why!

  • Anonymous
    September 02, 2011
    The comment has been removed

  • Anonymous
    September 02, 2011
    The comment has been removed

  • Anonymous
    September 03, 2011
    We are waiting eagerly for the BUILD conference. We hope that WebGL support will be announced. We keep our fingers crossed.

  • Anonymous
    September 03, 2011
    I hope the Build conference thingy is when IE team people releases IE10 beta..

  • Anonymous
    September 03, 2011
    i ope the fix this blog system before build

  • Anonymous
    September 03, 2011
    @Mario Should be IE10 beta time at the BUILD conference as it has been exactly a year by then since the IE9 beta.

  • Anonymous
    September 04, 2011
    Dear Microsoft, There is Spam from "Anonymous" Above. I would recommend doing something about that cause, it might cause slow loading time for some viewers. By the by you can delete this comment if you want to as it is just a notice to you guys. Sincerely, The Spam Hater

  • Anonymous
    September 04, 2011
    we don't need plugins, but we need native code for the web, and by that I am not talking about js as an assembly... There is a reason there are so many languages out there, you can and should not try to convince all to use js for everything!

  • Anonymous
    September 04, 2011
    @danglingpointer, I think some 3rd party development corp. can possibly develop this kind of reverse extension. Bcoz, big corp. only work on the solutions which are futuristic in nature (need of the hour). HTML5 is the future and flv fallback would be supported unless a major lot of users are not upgraded to the HTML5-supporting browsers.

  • Anonymous
    September 05, 2011
    Make the tabs in IE10 more round looking... the Tabs look in IE9 looks to square looking its getting annoying to me.. round tabs would look good on IE.... you guys had round tabs in IE9 beta but changed it in IE9 RC

  • Anonymous
    September 05, 2011
    I agree with Mario here. Also, the selected tab (especially when there are few tabs opened) has a somewhat trapezoidal shape/lighting, which does not look that great to me, frankly; with just one tab opened, at first glance the tab's vertical edge just seems misaligned with the address bar, before actually realising that that effect was perhaps intended.

  • Anonymous
    September 05, 2011
    The comment has been removed

  • Anonymous
    September 05, 2011
    Re XHR detection: native XHR in IE7 and IE8 is restricted to certain HTTP methods, thus it's understandable if sometimes people prefer the ActiveX variant.

  • Anonymous
    September 06, 2011
    @DanglinPointer, here is another test confirming IE10pp2's ability to deal with DOM is far too slower than Safari:  jsperf.com/options-add-vs-innerhtml innerHTML: 160 vs 16718 & options.add: 34 vs 757 IE really needs to improve in this area !!

  • Anonymous
    September 06, 2011
    Browsing without plugins is something I will never go. Hopefully the IE team will revert to their senses and allow pinned websites to be used with plug-ins.

  • Anonymous
    September 06, 2011
    The comment has been removed

  • Anonymous
    September 06, 2011
    I agree with Jacob cause most websites are starting to use png favicons more these days and firefox and google chrome supports it but Internet Explorer doesn't yet?

  • Anonymous
    September 06, 2011
    @jacob You can alteady download spelchecker addon like SpeckIE Microsoft originally created the Favicon using the .ICO image container. All browsers therefore support the original .ICO container as it is the original favicon. .ICO containers can hold multiple images for different icon sizes. .ICO containers can also contain PNG images for the icon

  • Anonymous
    September 06, 2011
    Add shadow Drop support! IE would get a 100 easier with shadow drop support.

  • Anonymous
    September 06, 2011
    Also, IE really needs to support things like HTML5 History API (www.w3.org/.../history.html), so that websites can make use of proper URL rewriting and do not have to use hashes and stuff ("/#!/") in the URLs (thus creating very long and ugly URLs, as the navigation goes by). A great example of this is Facebook: stackoverflow.com/.../how-does-facebook-rewrite-urls-in-chrome-firefox

  • Anonymous
    September 07, 2011
    I just typed a long comment and clicked "Post" and the comment has not appeared. What happened?

  • Anonymous
    September 07, 2011
    Yep, comment dropped. Way to go guys. And I see the "Post" button URL: javascript:WebForm_DoPostBackWithOptions(new%20WebForm_PostBackOptions("ctl00$content$ctl00$w_47041$_d66745$ctl00$ctl00$ctl00$ctl05$bpCommentForm$ctl05$btnSubmit",%20"",%20true,%20"BlogPostCommentForm-ctl00_content_ctl00_w_47041__d66745_ctl00_ctl00",%20"",%20false,%20true)) Great job there. Why can't you let the default submit button alone? Wht foul it up with terribly-written javascript?

  • Anonymous
    September 11, 2011
    Commenting on the IE blog is STILL COMPLETELY BROKEN!!!! Please remove all the stupid broken JavaScript for comment submission And just use a regular 100% guaranteed to work submit button!!! Were getting really tired of this issue not getting fixed!!!!

  • Anonymous
    September 12, 2011
    I'd like to extend an olive branch in good faith to Microsoft regarding fixing this blog once and for all. Is there someone at Microsoft I can talk to about fixing the comment system on here so that posts are not constantly lost. It is currently not presenting Microsoft in a positive light when their flagship developer blog doesn't even work! We realize that the majority of blame points to 3rd party blog software but there are many simple solutions to fix it. If you plan to fix it on your own that's fine, please indicate you plan to do so and provide a timeline for the fix. If instead you would rather have some assistance please specify an email address where we can contact you to discuss (free as in beer help). Thanks steve_web PS sorry for cross posting on multiple threads but this issue needs serious attention ASAP.