Testers Say IE 11 Can Impersonate Firefox Via User Agent String 252
Billly Gates writes "With the new leaked videos and screenshots of Windows Blue released, IE 11 is also included. IE 10 just came out weeks ago for Windows 7 users and Microsoft is more determined than ever to prevent IE from becoming irrelevant as Firefox and Chrome scream past it by also including a faster release schedule. A few beta testers reported that IE 11 changed its user agent string from MSIE to IE with the 'like gecko' command included. Microsoft may be doing this to stop web developers stop feeding broken IE 6-8 code and refusing to serve HTML 5/CSS 3 whenever it detects MSIE in its user agent string. Unfortunately this will break many business apps that are tied to ancient and specific version of IE. Will this cause more hours of work for web developers? Or does IE10+ really act like Chrome or Firefox and this will finally end the hell of custom CSS tricks?"
You don't say! (Score:4, Interesting)
Business apps designed specifically for IE6 might not work with IE11? I'm shocked! That's terrible! What is this world coming to? Or should I say, to what is this world coming?(don't answer that)
Re:Hmmm (Score:5, Interesting)
Heh, all they had to do was offer IE6 in a VM to allow all the businesses and government organizations to still run all of the old crappy homegrown locked-in apps to run. Those apps aren't going away (a lot are there to meet contractual/legal obligations and aren't trivial to redevelop / recertify).
So Microsoft hurt itself? (Score:3, Interesting)
Well... yeah but no. Their being different enough to make everyone else think all the other browsers were broken worked. Only web deveopers knew differently. And the business apps only worked under MSIE thing ensured people wouldn't migrate their client machines from Windows.
I have to wonder what Microsoft will pull next. As their game ran its course and more and more things went the standards route, what's next?
Parsing user agent strings = bad. (Score:5, Interesting)
I've been developing web applications full time since 1996 and I've never once had to resort to browser detection via user agent strings. It's just bad practice.
The fact that some people have been doing this has led to the very convoluted user agent strings we see today, rather than a simple description of the browser / rendering engine and version.
It's perfectly possible to write code that works cross-browser without having to detect browsers via user agent strings. The closest I've come to any sort of browser specific code is occasionally including IE specific CSS to work around IE bugs, but this included in an IE specific way and is ignored by other browsers.
A browser vendor should be able to put whatever they like in the user agent string, and if that breaks a web site or application, then so be it. It's the fault of the developer for making assumptions.
Bork Bork (Score:5, Interesting)
Back in 2003 msn.com deliberately sent Opera a faulty style sheet that broke the page, in response and to make a point Opera released a Bork version of their browser that turned msn.com into Swedish Chef talk. http://news.cnet.com/2100-1023-984632.html [cnet.com]
Karma is a Bitch.
Re:Hmmm (Score:4, Interesting)
Acutally, yeah, I manage the entire team here at my office... at home... consisting of myself... and, oh wait it's just me.
You can either bitch and moan about corporate lack of vision (or bureaucratic weight, or whatever you want to call it) or you can knuckle down and fix the shit they pile on everyone else's plate... and get paid for it.
I prefer to be the guy people can call when they want someone else's shitty mess fixed, rather than be known as the 'unapproachable tech guy'.
I've spraypainted nothing... But if someone wants to pay me to come clean it up, I have a contact form I can direct them to.
Re:Sigh (Score:4, Interesting)
> The day that you were able to tell what someone was running and make a decision based on that, we basically lost the point of a standard
Well, sort of. If the browser gets the standard wrong, and the options are:
1. It doesn't work for that browser.
2. Degrading the result for everyone.
3. Implementing a browser-specific work-around.
Which would you really prefer? Yes, user agent testing is heavily mis-used, but it's not the terrible idea it's made out to be.
I'll give you a specific example; we had an issue with file uploads with Safari over SSL. For some reason if the connection was kept alive, Safari would frequently start uploading the file but never complete. The work-around was to force connection close for Safari; it wasn't perfect, but it massively reduced the frequency with which the issue appeared.
Re:Hmmm (Score:5, Interesting)
Too bad if it does. Their excuses wore out long ago.
They did, but business apps that are tied to specific versions of IE are endemic and quite often it's not as simple as paying money and getting the software updated. We're not talking one or two apps here that need updating; we're talking hundreds if not thousands of applications, some of which quite clearly haven't had any major UI work done in five or ten years.
In the last fortnight, I've seen - and this is in just one small business:
- A web app that requires a specific ActiveX plugin to print - evidently a stylesheet for printing or even generating a PDF is too difficult. This plugin only works on 32-bit versions of IE; under 64-bit versions the plugin installer silently fails to work. (The plugin developer does have a 64-bit version available, but it's commercial software. You can't just download a 64-bit version from the developer's website yourself).
- This web app is provided for franchisees by their franchisor. (I won't name the franchise, but I guarantee you've heard of it). As with any franchise-type arrangement, the franchisee can ask their franchisor nicely but cannot force anything - and in this case, the franchisee simply cannot say "In that case I won't use your tool; I'll find something else to do the same job", using it is a condition of the franchise.
- Several web apps that require you to explicitly click the "broken mode" button in IE - they're generating IE6-only HTML when IE is used but IE isn't detecting this and automatically downgrading.
- Quite often these apps will work just fine with Chrome, Firefox et al. It looks like they're detecting an IE User-Agent string and generating IE-6 specific HTML rather than checking the IE version.
- These apps are provided by a third-party and you have to use them otherwise you can't do business with that third party. The business itself doesn't care about your idealistic attitude that IE-dependant websites must die; they need to meet payroll this month and one of the ways they do this is by working with various third parties.
- Web applications that quite simply do not function in anything but Internet Explorer in any form, no matter what you do with your user-agent string. You'd be amazed (and faintly disturbed) how many project managers read as far as "no need to deploy your own client app" when first considering web development and didn't get the bit about "with careful development, client platform independent".
- Much of this is actually Microsoft's own doing - they purposely encouraged this sort of behaviour back in the days of IE6.
Re:Hmmm (Score:4, Interesting)
spoof the user agent back to MSIE
So that the served content is for IE6-8 ... [which doesn't work with IE11]
Part of the fun here is that IE has "spoofed" FF and the other Mozilla browsers all along, by including "Mozilla" in most forms of its User-Agent string. I see this all the time, when I test my web sites against various versions of IE. This has always been a minor problem for web developers, since it's easy for software to misunderstand such things. You might think you've got a test that successfully distinguishes real Mozilla-type browsers from IE, but then MS releases a version with a tweaked User-Agent string that your RE doesn't parse quite right, and your code sends the wrong style of HTML to the browser.
I've occasionally wondered why the Mozilla gang hasn't charged MS with trademark infringement for such monkey-wrench tactics. After all, if I were to start providing a browser whose default User-Agent string included the "MSIE" token, MS's lawyers would be all over me. But they use their main competitor's brand name with impunity. If the Mozilla crowd weren't such nice guys as to allow this, life might be a bit easy for web developers everywhere.
Actually, quite a lot of browsers provide a list of User-Agent strings, and let a user choose one. This is probably legal, and is occasionally useful, especially to developers. But it's annoying and a waste of developers' time when vendors are allowed to install a lying User-Agent string as the default. It would improve matters for a lot of us it this were legally considered consumer fraud, trademark infringement, or whatever other legal terms apply.