Support for older versions of Internet Explorer ends on January 12, 2016
microsoft.com
microsoft.com
Don't get me wrong - I think it's a good move to discontinue support for older IE versions... There is just something with all the smiling and laughing faces all over the page that make all my alarms go off
It just means Microsoft can't be bothered to ship patches for those older browsers.
Checkout the photoshopped freaks they used for VS2008 http://www.secretgeek.net/vs2008_bugeye
Behind their grins are the questions, "what now?" or maybe even "where do I want to go today."
Their IT krewes are wondering, "how am I going to get the CFO to approve an upgrade to the GE medical imaging suite? we've been out of maintenance for 15 years."
And their info security guy is considering seeking asylum in a foreign country.
Who? Oh, that guy? We let him go six years ago when we couldn't figure out what he did.
http://www.lucasthenomad.com/images/cache/img/5d/5d8664cf92e...
I have a beautiful dream sometimes, where I go to work, and I only have to care about supporting Chrome...
If you still have some IE5 stuff, my guess is that you have to bring it up to the HTML5 table layout zone. And then you need to back polyfill a bunch of stuff to deal with the "IE7 mode" requirement.
I understand the sentiment, but to be clear: Ideally we do not want any one vendor with a majority browser share. Otherwise someone will get complacent and we'll have a repeat of the dark times (IE6). Having multiple browser vendors is good for the web.
A browser's rendering engine is the ONLY part that matters, and if anyone has a monopoly that is "bad" because competition helps innovation.
Each rendering engine winds up, for good and bad, with a small group of people deciding what features do and do not get put in. If they simply decide that feature Y doesn't get put in, and that rendering engine has a monopoly then feature Y is a non-entity as far as the web is concerned.
For example, if Google controlled the web's primary rendering engine do you think Do Not Track would be in it? NO! It took all of their major competition to add it before they did, starting with IE.
Active X would like to disagree with you. We're currently dealing with a massive (revenue/company) product that only runs under IE < 10 because it still uses Active X.
I don't think a common rendering engine is at all a bad thing. Maybe two to keep them honest. But blink + webkit + trident (?) + Gecko (?) seem like wasted engineering talent.
The incompetence was in mandating a certain implementation instead of writing a technology-neutral law requiring a certain security level.
> For example, if Google controlled the web's primary rendering engine do you think Do Not Track would be in it?
Isn't Do Not Track a pretty good example for a browser feature that is independent of the rendering engine?
I don't necessarily agree with your assertion that a browser's rendering engine is the only part that matters. Internet explorer ships with windows by default. Mozilla (firefox) has policies that focus on privacy. Chrome focuses on speed, and, well, serving Google.
These are all things that differentiate the products in the market and aren't necessarily rendering engine related. Yes - if one entity maintained absolute control of the rendering engine, then the needs of each vendor would not be served - but then it wouldn't be open source, either.
I don't want to be too suspicious, but to improve the web I'd rather bet on the Mozilla fundation.
I can dream, right?
I just wish that there was one unambiguous spec that they all supported. How many man-hours on aggregate have been wasted tweaking CSS and futzing with javascript to try to achieve identical results across browsers that have all these slight and not-so-slight differences of opinion about the HTML, CSS and ECMAScript specs?
A whole other world of pain is when you get into browsers that are embedded in GUI frameworks, or as extension points in other products. You really have no idea what they support, until you try something and see it is broken. Does this browserPanel object use the system version of IE, or some other version, or an old, forked version of WebKit, or some other Lovecraftian monstrosity?
you sure? Firefox is ahead in at least 2 of the 3 categories on caniuse (though I will admit, they are _very_ close).
Look under "Summary" on these pages:
* http://caniuse.com/#cats=JS%20API
I'm not affiliated, but I reported issues for Xbox One's Edge browser and got a positive response. Time will tell if it actually gets fixed, though.
Yes I also long for the return of the IE6 days when only one browser mattered and what it did was law. Not.
Seriously folks we've already done software monopolies (not just in browsers either), it always ends up in tears. If a browser is terrible refuse to support it or ask for extra time/money for its support, but do not condone, suggest or recommend browser monopoly.
2015 is really the year where we saw the majority of our clients drop support for IE8 altogether. Most of them showed sub-2% browser share and dropping fast in their analytics..
Replace that with IE and many of us have experienced it, we more or less had to beg for time to test in other browsers.
It was kind of a dream but more of a nightmare and I hope we don't go back there soon, neither with IE nor with Chrome.
This is a (primarily UK) product most of whose users are on internal bank networks. It's come down from about 40% 2 years ago, so it's heading in the right direction - but I'm not sure I see it going away any time soon.
There's huge organizational inertia inside these banks around IT systems - if we want to serve them, we need to support IE8 for the foreseeable future.
http://techcrunch.com/2015/04/16/google-will-still-support-c...
Firefox (as far as I can tell) still supports XP SP2, but who knows for how long.
You are assuming that they have a choice. A large chunk of them most likely don't.
The only way enterprise IT departments will upgrade is when they start having to take responsibility for security problems. As long as they can lie to their users and claim there's no need to upgrade, they will.
In practical terms it's still IE9 and IE11 for a while (so far as I can tell)?
https://support.microsoft.com/en-us/lifecycle#gp/Microsoft-I...
I'm sure there is still a load of Windows Vista SP2 machines around...
Windows Vista has support through 2017 and Windows Server 2008 has support through 2020.
https://support.microsoft.com/en-us/gp/microsoft-internet-ex... provides a clearer overview of what the supported version where is. (Note that it excludes OSes that are currently supported by end support before 12 Jan 2016.)
Unsupported XP is, in practice, probably more of an issue than Vista.
"Does this mean Microsoft is changing the support lifecycle policy for Internet Explorer?"
"The latest version of Internet Explorer will continue to follow the component policy, which means that it follows the support lifecycle and is supported for as long as the Windows operating system on which it is installed. Focusing support on the latest version of Internet Explorer for a supported Windows operating system is in line with industry standards."
A simple yes or no will do.
Haven't they done better than that? Except they've dealt with the possible issue that the reader has an incorrect understanding of the support lifecycle policy for Internet Explorer.
https://www.netmarketshare.com/browser-market-share.aspx?qpr...
IE8 still has over 10%
http://gs.statcounter.com/#desktop-browser_version_partially...
Browser trend will probably be more influenced by region and public. If your public is stay home dads in China you should definitvely be looking into iE8, if your public is tech workers in the Bay Area you shouldn't even care about iE11.
Netmarketshare counts each user only once per day (per site?). If you assume users if Chrome/Firefox etc are more active on the web than IE8 users then it fits.
Which number is more important when deciding what to support on your site is a complicated question. Probably your own user demographic has a greater impact though.
> What is end of support? Beginning January 12, 2016, only the most current version of Internet Explorer available for a supported operating system will receive technical supports and security updates. Internet Explorer 11 is the last version of Internet Explorer, and will continue to receive security updates, compatibility fixes, and technical support on Windows 7, Windows 8.1, and Windows 10.
Not a single mention of Vista/etc.
"End of support refers to the date when Microsoft no longer provides automatic fixes, updates, or online technical assistance. [...] you will no longer receive security updates"
Scroll down to the table on this page: https://support.microsoft.com/en-us/gp/lifepolicy
Given Vista's reputation, I'm willing to bet Microsoft doesn't feel a need to address the few Enterprises who are stuck on Vista SP2 until its support ends.
I'm curious if anyone knows a company that plans to ride out Vista SP2 until its support ends. Seems like most companies did the XP to 7 hop (and are currently hoping to ride 7 until it sunsets).
What bugs me is so many negative comments along the lines of "What about all those poor helpless trillion dollar megacorporations that are stuck on an old version because they use proprietary custom software? Microsoft is evil because they don't guarantee to support and repair everything that they've ever made for free until the heat death of the universe!"
Those companies and organizations are stuck not because of Microsoft, but because they mismanaged their own proprietary software or web applications. For more than a decade, they never even bothered to plan for maintenance, although they routinely plan, budget, and schedule depreciation and maintenance for everything else in the business - things that move much slower than software and far far slower than web development. I'm no fan of Microsoft, but that's their fault, not Microsoft's.
My Windows machine is still on Windows 7 now, and not ready to upgrade to 10 yet due to compatibility issues. But I have 4 more years to get things working in a virtual environment, or find replacements. Software that I run dates back to the 70s, so some didn't work by default in 7, but in the end there was only 1 program that I couldn't replace or get working with an upgrade or VM. I reverse engineered the data to migrate it (it used an obsolete floating point format from before floating point ops were standard in PCs). If I can do that, a company or large organization can do it.
For web software, the pace of the state of the art is even faster. Most anything over a few years old probably needs a rewrite and data migration if it hasn't been maintained properly (and possibly even if it has). Especially anything that used old plugins.
But the focus on browser versions is weird. I strongly dislike the practice of web developers/designers speaking of 'supporting' IE6/IE8 or whatever. That's a Microsoft product, Microsoft supports it or not, you and I don't. We support the websites that we build, which we build to current industry standards (such as they are). If you try to access it using a nonstandard client like IE6, Arachne running on DOS, Hyperlink on a C64, or a line mode browser on a teletype, then your experience will be different. The server side will work fine, but your client may not do what it's not capable of.
You can't expect to stick a blu-ray disc on a record player or shove a flash drive full of MP3s into an 8-track tape deck and have the end-user experience be the same. It just doesn't work that way. Sure we can try to build custom client-side workarounds, but is that worth the cost (especially given that new browsers are free)?
Overheard at a meeting once:
Client: "Our stats say that'll be fine for our customers, but our executive team mostly has older versions of IE, so this won't work for them. We don't want it to look bad to the CEO."
3rd Party Agency Manager: "We'll order new laptops for them and have them delivered, just tell us how many and where to send them. It will be much cheaper for you and save us all a lot of time."
Can anyone disagree with this analogy? If not, I'm going to start using it all the time!
This person was a genius, or was bitten by this far too many times.
Client: "Oh, no. We can't do that. That would violate policies ABC, XYZ and 123. Team FOO would never sign off and ...".
They will be vulnerable to attacks because of this and they will refuse to pay money to upgrade from XP or Vista to go to Windows 10. It will be because their business software only works with older Windows versions. They can't afford someone to migrate to Windows 10 or they lost the source code after programmers retired or got fired and took it with them on their personal laptops because they lacked a source control system.
It is a big mess out there. Colleges are even worse as is the federal government who use outdated Windows Updates by three years and expired antivirus products.
I know a lot of businesses still use DOS or OS/2 and very old PCs because they don't trust Windows.
I would get offers to volunteer for non-profits that still use old PCs with older operating systems on them and they need help fixing them or maintaining them.
There are a lot of third world nations using old surplus PCs because they cannot afford anything else. So they get our leftovers.
For example in Thailand they still use Windows XP, a pirated version but they put it on every PC. They have a skin for it that looks like Vista but it is still XP SP3. A copy of legit Windows is too expensive so they used the pirated XP because they can buy it on the streets for $1 due to a lax copyright law.
Buying or developing technologies that rely on weird proprietary quirks of free versions of software isn't very sane. There's not even a business model in place you can rely on to continue to want to support you since it's free.
That brings up an interesting point though, a special NPAPI plugin that aimed to provide perfect compatibility with older IE versions for a small per-seat license fee could do really well...
It's all very well saying this now, but I wouldn't be surprised if at the time these technologies were seen as JavaScript is now. I remember Adobe Air was when it came out...
It's sort of like having a codebase which you can't change which makes use of a bunch of weird proprietary C extensions from some old compiler, and the compiler itself not only has the capability to introduce bugs and security flaws, but has a track-record of doing so. You know it's a ticking time bomb now, and you should have seen it in the beginning (although so many didn't :/ )
I had open sourced this script to detect the OS and web browser so that it would run JavaScript code using if statements for each platform to solve compatibility problems. It should still work today but the ActiveX detection is broken because of the way JavaScript changed.
http://www.javascriptsource.com/repository/javascripts/2005/...
It took quite a while for some standard JS features to be added[1], but a mostly stable subset was possible (as evidenced by the sites that worked in all the major browsers of the time). The real problem was companies not wanting to devote the time to doing this and taking the shortcut of using the non-standard IE features MS had included at the expense of browser interoperability (which would have provided some protection against using features that would be dropped).
1: https://msdn.microsoft.com/en-us/library/s4esdbwz(v=vs.94).a...
In my experience I had to write JavaScript code that detected the OS the web browser type and even the processor so that I could use if statements to run the correct JavaScript code for each platform that worked.
I've been out of it for a decade due to a disability, but I see that Microsoft has updated JScript and has used some form of JavaScript in Windows 8 Apps as well. I never really liked Metro/Modern UI and Windows 10 kind of changed it to Universal Apps.
I like that they have made improvements to JavaScript like Node etc and even run JavaScript in the OS and the backend as well as the frontend.
Now, I know MS's version of javascript (JScript) had it's own quirks and issues (such as no trailing comma allowed in objects and array syntax), but it shouldn't have been too hard with a minimum of testing to make sure it worked on both IE and Netscape and/or Firebird (before it was renamed to Firefox). The problem is that many didn't both at all, and didn't even attempt to figure out which bits that they used were standard parts of Javascript or the ECMAScript standard and what were MS extensions.
That said, there was a major benefit to people forging ahead with MS specific extensions. The wide acceptance and update of XMLHttpRequest was spurred by MS's introduction of that type of call through custom ActiveX objects[2].
1: https://en.wikipedia.org/wiki/JScript#Comparison_to_JavaScri...
2: https://en.wikipedia.org/wiki/XMLHttpRequest#Support_in_Inte...
Management required that I use VBScript and ActiveX controls, but I had to use JavaScript in cases when VBScript was too slow. Each new version of Internet Explorer would break compatibility and require a rewrite. For example using and programming for IE 4 would break when IE 5 came out and even the AcitveX controls would not load properly. We had to switch from some ActiveX controls Kodak made to use Lead Tools instead to load TIF images of the employee pictures. Then IE 6 came out and broke that and required a new version of Lead Tools and other stuff.
There were people with Macs, OS/2, and Linux that could not use the ASP made websites because they couldn't run ActiveX and VBScript. So I suggested to management that we replace the ActiveX objects with Java objects and use JavaScript instead of VBScript and we could then make it so that the Non-Windows systems could access the Intranet web apps. Management hated that idea and ordered me to never use Java because it wasn't made by Microsoft and they didn't understand it.
But yeah each new IE version broke compatibility in the way management wanted me to develop web apps. I can see why companies want to stay with an older IE version and older Windows, because of the time and money it takes to upgrade the websites to the newer IE versions.
Actually, chances are such an environment will also have the other security settings tightly locked down too, and even outside the browser executables may be whitelisted, so it's not as bad as it may seem at first.
Just disabling JS probably gets rid of 90%+ of the attack surface of any browser including IE.
In most cases the hardware vendors never address the browser compatibility issue (due to focus on new products) and only provide a warning sign upon login. Likewise the hardware isn't easy to replace in some environments.
This use case is completely absent from the web browsing traffic usage patterns on the internet which predominate in browser usage statistics.
My thinking is just that most larger companies will have zero chance of hitting a January deadline, and once the deadline passes, there's really no urgency to upgrade by any particular time. I would think a ~12 month deadline might be more effective in getting people to upgrade (as well as resulting in fewer compromised browsers).
The amazing thing is that Microsoft took this long to correct what should have been an obvious mistake: that IE versions apply like other Windows components and they should only need to support the most up to date one for that OS version.
The kind thing here is that Microsoft has taken ownership of that mistake, publicly admitted that they made it and that the fix was coming, and given companies more than a year and a half (and advertising to that effect, such as the site linked here) to adjust to the fix.
The whole point of microsofts' action here is to say that it doesn't work like that. If you have an infrastructure of webapps/intranets/whetever then you must carry the continuous cost of making sure they sure they work on new browsers all the time, because old browsers will be quickly become insecure.
If you look at the numbers and include China they're quite bleak. If you exclude China it doesn't seem quite so bad.
But China has a massive Windows XP piracy problem that isn't going away.
Otherwise they end up like this French airport which was shut down just few weeks ago because their Windows 3.1 failed.
As Vista is still supported, the IE9 lives on.
Theodores: I only want to support IE11+
Client: No. What if our users have older browsers?
* later *
Theodores: I only want to support IE11+
Client: No. What if our users have older browsers?
Theodores: That should be exceedingly rare, Microsoft have actually pulled support for those versions recently. See this link.
Client: Oh, ok then, great. IE11+ it is.
Assuming your application does not need IE specific features, what are some reasons that prevent you from nudging users this direction?
our customers want to sell to these IE8 users. The response to telling them to upgrade is "we'll use a different vendor then - one which does support IE8"
https://support.mozilla.org/en-US/kb/firefox-no-longer-works...
Windows XP already lost support for their latest release of IE years ago, so this doesn't impact them.
This may impact Windows Vista and Windows Server 2003 but both are out of mainstream support, and Vista's numbers are extremely low overall. Supposedly 1.74-1.77%.
IE 11 ends with Windows 10. Edge the Browser isn't classified as a Windows component, to my knowledge, and so also shouldn't classify for "extended support" like IE has in the past. It certainly sounds like the intention is to drop IE entirely once Edge has feature parity and user adoption on its side.
(Things are a bit trickier when talking about EdgeHTML the Renderer which I would think is a Windows component still, given that it runs some of the Windows Universal Platform, but here also the benefit of forking it from IE is that they can keep the support policies clear from day one as opposed to having to fix up things after the fact like they did here.)
Beyond that, if an enterprise has not gotten the message not to build LOB applications as browser plugins and not to build for one and only one browser model any time since IE6 has been deprecated, then IMNSHO they deserve what they get when IE11 dies as the last "IE". Given the support timeline expected for Windows 10, enterprises have presumably roughly a decade of extended support to adapt before IE dies. Really hope enterprises are paying attention to that (but of course we know a bunch will not and will complain in ten years about lack of support and those of us that got the memo today can proudly mock them then).
To IE 9. :(
My definition of a modern browser is one that self-updates, this is essential for security. There are plenty to choose from, but if you're using a browser that doesn't keep itself up to date you're asking for security problems.