Testers Say IE 11 Can Impersonate Firefox Via User Agent String

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?"

Comments

Hmmm

[BrokenHalo]

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?

Too bad if it does. Their excuses wore out long ago.

Re:Hmmm

[Anonymous Coward]

Or, just use another browser and spoof the user agent back to MSIE

It wasn't the web developers excuses, they're gonna suffer because of MS and they don't deserve that.

Re:Hmmm

[pspahn]

...they're gonna suffer because of MS and they don't deserve that.

Suffer? This just creates more billable hours. I'm not sure what line of work you're in, but the phrase "more work for you" isn't exactly a bad thing (as long as it's paid for!)

Re:Hmmm

[Anonymous Coward]

Yes, yes they do.

Re:Hmmm

[Anonymous Coward]

Too bad if it does. Their excuses wore out long ago.

Exactly, they've had 10 years to fix their old busted stuff.
If Microsoft's going to kill off old Windows apps, they should take old web pages with them as well.

Re:Hmmm

[second_coming]

The main culprits I've seen which do this are telephone system providers (Mitel/iPecs etc).

The issue being that people are very touchy about updating telephony software, primarily following the old adage, "if it ain't broke don't fix it".

Re:Hmmm

[rudy_wayne]

The main culprits I've seen which do this are telephone system providers (Mitel/iPecs etc).

  The issue being that people are very touchy about updating telephony software, primarily following the old adage, "if it ain't broke don't fix it".

The problem is that it is IS broken.

Re:Hmmm

[mwvdlee]

Spoken like a true soulless manager.
If you go and "spraypaint" the wall of your company's toilets, it's "more work for the cleaners" too.
Think they'll be happy with those extra billable hours? I'm sure their managers are.

Re:Hmmm

[rwa2]

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).

Re:Hmmm

[jimshatt]

Except that it wont. If the business app was tied to a specific version of IE, then it still is tied to that specific version of IE. IE6 still sends out MSIE, regardless of the existence of IE11, IE12, or IEFoxHunt.

Re:Hmmm

[jimshatt]

And why is it Microsofts job to do this? No-one is stopping those organizations to do this themselves.

Re:Hmmm

[pspahn]

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: Hmmm

[Anonymous Coward]

I hate to be the one to tell you this, but your boss sounds like a total opportunistic douche.

Re:Hmmm

[YeeHaW_Jelte]

Yes that is all they have to do and surprise, surprise, they do it:

http://www.microsoft.com/en-us/download/details.aspx?id=11575

Re:Hmmm

[Anonymous Coward]

Fuck excuses, make a browser that actually works. A website built in Asp.Net using Visual Studio's horrible WYSIWYG (yeah right) editor won't even fucking render in IE, although Firefox and friends render perfectly fine. I'm not the one making any excuses, but I'm forced to support their clusterfuck of a browser nonetheless.

Re:Hmmm

[MightyYar]

I'm hourly as well, but if I weren't, the analogy would be more like a dog shits on your lawn: you can either leave it there to turn white or you can pick the shit up. Either way, your life sucks a little bit more.

Re:Hmmm

[Bengie]

Your analogy assumes there were toilets in the first place. Some times it's better to bite the bullet and do things correctly instead of complaining about each time some fragile software breaks because of bad design and bad practice.

Re:Hmmm

[Bengie]

Those apps aren't going away (a lot are there to meet contractual/legal obligations and aren't trivial to redevelop / recertify)

I have no sympathy for companies that used bad software. They're in their position because of bad business decisions in the first place.

Re:Hmmm

[jimicus]

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

[DarkOx]

A VM would not work very well; for lots of applications. Most of those things uses some terrible ActiveX controls, Need to touch disk files etc. If Microsoft had been really smart they would just continue to include IE in only the Professional and Enterprise releases of Windows, re-branded IE-Legacy or something as an optional component.

Keep the id10ts for using it as there web-browser by modifying it slightly to only allow it to open sockets with RFC1918 address by default; and some registry keys and (GPO to set them) if you *need* it to access some normally publicly routed ranges.

That would make just about everyone happy.

Re:Hmmm

[rudy_wayne]

Those apps aren't going away (a lot are there to meet contractual/legal obligations and aren't trivial to redevelop / recertify)

I have no sympathy for companies that used bad software. They're in their position because of bad business decisions in the first place.

Unfortunately, it's not that simple.

Browsers and the World Wide Web in general didn't just suddenly appear one day, fully formed with a complete set of perfect specifications and standards. They evolved slowly over time. And while everything was evolving, and while everyone was trying to figure out exactly what those web standards should be, the rest of the world wasn't standing still. Billions of web pages were being created, based on whatever shitty browsers and standards existed at the time.

For a long time, it didn't matter what "standards" there were. Internet Explorer *WAS* the standard, because it was the only major browser -- there was no Firefox or Chrome -- and so that's how web pages were designed. Then when things changed, when there was competition among browsers and more emphasis on adhering to standards, there was a problem. There were all these billions of web pages and applications based on old shitty browsers. Suddenly businesses had all this stuff that worked perfectly fine in IE6 but broke horribly with any other browser. It's easy to make fun of their "lack of foresight" but back when IE6 was the only browser from a big well known company, people had no way of knowing that things were going to change tremendously in just a few years.

And so browser developers were forced to resort to all sorts of hacks and kludges to make sure that their browser properly rendered all those shitty poor designed web pages. Sure you could design a browser that refused to display all those improperly coded pages. (Hey, remember "Strict HTML"?) And you would watch usage of your browser drop to zero. When the average person goes to a page that does not display properly how many of them think "this page wasn't designed properly" versus "there's something wrong with my browser".

Re:Hmmm

[Sulphur]

I'm hourly as well, but if I weren't, the analogy would be more like a dog shits on your lawn: you can either leave it there to turn white or you can pick the shit up. Either way, your life sucks a little bit more.

Of course if it turns white, then it is left there as part of the decor.

Re:Hmmm

[Anonymous Coward]

Spoken like a contractor. Billable hours only works if you're supplying services or contracting...and it's not your job on the line because of inconsistent behavior with the damn browsers.

Re:Hmmm

[YeeHaW_Jelte]

Too lazy to follow the link?

It's a virtual machine, works perfectly fine on VirtualBox and thus on Linux and OSx.

Re:Hmmm

[YeeHaW_Jelte]

Eh pardon yes it's an .exe but it's just a wrapper around a rar file. Unrar, at least on Linux machines, understands them perfectly well.

Re:Hmmm

[MightyYar]

Dung shui?

Re:Hmmm

[CastrTroy]

MS could have alleviated a lot of pain everywhere if they would have just added a supported method of running 2 versions of IE at the same time. This way they could have continued to support businesses that were locked into corporate intranet applications that wouldn't work in anything but IE 6, and also have newer versions of IE be able to adopt proper standards without having to worry about how it affected older websites.

Re:Hmmm

[Bozzio]

Exactly. Thanks for mentioning this. I was going to post the exact same thing.

Re:Hmmm

[Anonymous Coward]

I'm with you man. I have no idea why people are calling you soulless because you're trying to make money from the ultimate soulless machine, the corporation. It's not any of it would be your fault. I'm assuming these posters are college kids who've never worked.

Re:Hmmm

[StripedCow]

This is called the "broken window fallacy".
http://en.wikipedia.org/wiki/Parable_of_the_broken_window

Re:Hmmm

[rvw]

A VM would not work very well; for lots of applications. Most of those things uses some terrible ActiveX controls, Need to touch disk files etc.

A VM has that all. It's just a very basic Windows installation and it can include ActiveX, and has its own virtual disk.

Re:Hmmm

[Anonymous Coward]

This makes the assumption that the money would be spent on something else when the true objective of the game is to hoard as much as possible.

Re:Hmmm

[AmiMoJo]

Actually MS offer a compatibility mode in IE that runs the old IE6 engine in a sandbox. You can create a whitelist of sites thatwill auautomatically use it. No need for a VM.

Re:Hmmm

[Anonymous Coward]

The truth of the matter, is that apps have a lifetime / lifecycle... just like you aren't going to find a computer to support your old ISA or VL-Bus cards anymore, you're eventually going to find that you need to either: Stall software updates/upgrades completely (which is less practical given an eventual need to purchase new hardware), or upgrade the software.

IMHO, this is one of the strongest reasons to consider COTS software solutions instead of home-brewed apps.

Re:Hmmm

[Ash Vince]

Browsers and the World Wide Web in general didn't just suddenly appear one day, fully formed with a complete set of perfect specifications and standards. They evolved slowly over time. And while everything was evolving, and while everyone was trying to figure out exactly what those web standards should be, the rest of the world wasn't standing still. Billions of web pages were being created, based on whatever shitty browsers and standards existed at the time.

For a long time, it didn't matter what "standards" there were. Internet Explorer *WAS* the standard, because it was the only major browser -- there was no Firefox or Chrome -- and so that's how web pages were designed.

Exactly.

I used to work from 2002 to 2005 as a web developer for a company who mostly contracted to graphic designers. At the time they expected to things to work on IE5 (the Mac version of course). They did not really care about Firefox (although it did exist then, but with zero non-techy users).

I threw together god knows how many sites in the 2-3 years I worked at that company. All we did was offer the client a choice: If they wanted firefox support, they paid extra. Almost nobody bothered. We were a budget development house so our margins would not support the extra work of supporting all the IE hacks needed and the more W3C firefox unless the client paid extra. They all required the sites to work perfectly in IE though obviously.

I tried to make sites work in Firefox just out of a sense of professionalism on a few occasions but the problem is that then you appeared to have a far slower work rate than the rest of the team who took the IE only short cut they were told to by the technical manager. He was also a developer, director of the company and joint owner so he made an informed decision not to support anything other than IE from technical perspective and was able to see if you were ignoring it. If you ignored it that was fine, but you still had to keep up with the other devs simply by putting in overtime.

It only took other browser to get a market share above 5% - 10% for things other than IE5 / IE6 under Windows and suddenly clients were interested in supporting other browsers. In 2002 - 2004 though IE was so dominant that nobody cared about anything else in the real world as only geeks bothered to change their browser. Making things IE only remained common place in commercial web development right up to 2005 - 2006.

http://en.wikipedia.org/wiki/Usage_share_of_web_browsers

I remember having to spoof using IE under Linux in order to access my online banking (from HSBC) as they considered all other browser to be too insecure :)

Re:Hmmm

[Anonymous Coward]

spoof the user agent back to MSIE

So that the served content is for IE6-8 and your app still doesn't work because your browser handles the version specific hacks differently.

Re:Hmmm

[SenseiLeNoir]

If I had mod points, I would award you all of them.

Its easy for some "Younglings" today to criticise what happened then. I personally didn't like IE back in 2000(Active X drive by installs, toolbars, and standards compatibility). Hell, life was bad back then (1997 - 2004) when Nutscrape and Exploder each having their own view of "Standards" (I am ignoring the Mozilla Suite and Netscape 6/7 as both never truly gained critical mass, and were not really for the masses anyway).

Until Firefox was released in 2004, there was no credible standards based browser, apart from the rather good Opera, which was pay ware at the time. Even then it was the surge of appalling viruses and drive by installs on IE, that finally got people to notice Firefox with its better security, and seduced by features such as pop up blocking and tabbed browsing.

Thanks to the efforts by Mozilla, which were then cemented later by Apple with Webkit/Safari, Standards Compliance has become a desirable feature for browser developers, rather than a drag.

Even Microsoft has slowly gotten their act together, and improving their own standards support.

Today we are in a better place, but its easy to forget how bad it was. even though Firefox was released in 2004, it still took a while to overtake IE.

In companies, where they are very much into standard installations and frowning upon installing software, it wasn't until Windows 7 came that IE 6 was eventually replaced together with XP. Therefore coding for the crappy standards of IE6 was the only realistic path available.

Its easy for most of us working in Tech companies to say otherwise, but in other companies like banking, and government, its a different environment.

Re:Hmmm

[Runaway1956]

And - why do we CARE whether they are trivial to redevelop / certify? If they need the app, and the app doesn't work any more, they'll pay. If they don't pay, they don't have the app. I see no problem here. The scrambled bullshit is going away, one way or another. And, few of us really give a damn that it's going to cost the corporations something. Eventually, IE6 in a VM will no longer be an "acceptable solution" for getting things done.

As for the triviality - maybe what you're saying is, all the geniuses have passed away or something? Today's crop of hackers don't have the intelligence to reproduce yesterday's results, using the tools available today? Are you the result of the infamous "dumbing down of America"? Are you incapable of doing what your predecessors have done, even with their finished product in front of you?

Oh well. Maybe you can join the illegal aliens in digging ditches and roofing houses. Surely there is a future for a failed IT guy.

Re:Hmmm

[ArhcAngel]

The issue being that people are very touchy about paying for updating telephony software,

We are currently upgrading our phone system (less than 200 PID) because updating JUST the software on our current system to get official support from the vendor is going to cost >$50K and a new system with all licensing and a year of support as well as added functionality will cost slightly >$100K and isn't already EOL'ed.

Re:Hmmm

[gbjbaanb]

So they're also invented IETab for IE too now.

Re:Hmmm

[Anonymous Coward]

This makes the assumption that the money would be spent on something else when the true objective of the game is to hoard as much as possible.

It's not about money. It's about the opportunity cost of the repairer. Instead of making something new and increasing the overall wealth (not the same as money) of the community, he is simply bringing the overall wealth back to where it was before the window was broken.

Re:Hmmm

[Bacon Bits]

The main culprits I've seen which do this are telephone system providers (Mitel/iPecs etc).

  The issue being that people are very touchy about updating telephony software, primarily following the old adage, "if it ain't broke don't fix it".

The problem is that it is IS broken.

There's a significant difference between "broken" meaning "functions in an anachronistic or extremely sub-optimal fashion" and "broken" meaning "complete loss of function". If you've got the latter, you'd gladly take the former.

This is why people tend to dislike new technology when it completely replaces an existing old system rather than complimenting it or existing along side it. Systems don't survive to be old if they don't meet the needs of the people who use them, and almost any new system will have some period of time where the new system does not meet their needs.

Re:Hmmm

[jc42]

Suffer? This just creates more billable hours. I'm not sure what line of work you're in, but the phrase "more work for you" isn't exactly a bad thing (as long as it's paid for!)

In economics circles, this would be considered a case of the "Broken Window Fallacy". That's the term for the belief that descructive acts (e.g., breaking windows) adds to the economy because it creates sales of replacement parts and employment for the workers that fix the damage. This is wrong, of course, because it doesn't add to the total wealth; it only shuffles money around while decreasing the total wealth. Time spent repairing damage is time lost that could have been used to create new stuff.

The concept applies in the software business, too. Real social wealth is created when someone builds software that delivers useful new capabilities. The Web as a whole is a good example of this. But software that simply does something in an incompatible way doesn't add to wealth; it merely increases the labor required to do a given job. That's a reversal of the usual "wealth" benefit of computing, which is based on the idea of replacing human labor with the activity of mechanical gadgetry, freeing human time to do more interesting things.

Unfortunately, we have a lot of history saying that people easily fall for the Broken Window Fallacy in most of its forms. In particular, manufacturers routinely "innovate" by intentionally making things that aren't quite compatible with their competitors' equipment. This is a serious drag on advances in the "Human Condition", since it's invariably a sinkhole of human time, trying to deal with the messiness and unpredictability of all the incompatibilities. We have adopted computers because they've freed up our time, not because we want to spend more time doing things that could be done quickly.

Microsoft has a well-understood history of throwing monkey wrenches into the machinery (to use another form of the metaphor), but they're far from the only ones. Pretty much any corporation with the economic clout will do the same thing, as they attempt to lock customers into their brands.

Re:Hmmm

[Anonymous Coward]

souless manager?

hey dipshit, souless managers are the people you work for, and their not a glass is half-full, people.

so shut the fuck up, you souless cubicle monkey.

let the contractors who have a drive and a spirit, and plenty of soul, come in and fix, what you and your W2 little serfs couldn't hope to fix with an infinite number of typewriters and equal amount of years.

Re:Hmmm

[Anonymous Coward]

Yes it does work on Mac and Linux you stupid fucking idiot.

Re:Hmmm

[Anonymous Coward]

How would this break apps? IE6-8 will still have the same user agent.

Re:Hmmm

[jc42]

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.

Re:Hmmm

[Anonymous Coward]

I agree. But as it is I have to keep one boat anchor of a machine around here in order to access a specialized website that only works with IE7 or lower. It doesn't work at all with Firefox, Chrome, or anything else. I have zero control over the remote site's ridiculous choices. All I can do is politely complain.

They've told me that the next version will "solve" the problem and become multi-browser/multi-platform capable. Yay! They're finally moving into the modern, browser-choice-independent age.

They've chosen to implement it with SilverLight.

Excuse me while I scream.

Re:Hmmm

[Anonymous Coward]

Why are you people amazed at the captcha? It uses an algorithm to choose a word based on the text in your comment. When there is not enough text or it's rules cannot parse the conflicting text enough you get a random word. Or did you really think that the ./ server is in some sort of super coincidence loci?

Re:Hmmm

[JDG1980]

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.

The Sega v. Accolade case held that it was not copyright or trademark infringement for an unlicensed third party to use the string 'SEGA' in a cartridge ROM without authorization when it was technically necessary for interoperability purposes. Presumably the same logic would apply to user agent strings.

Re:Hmmm

[Sloppy]

The Broken Window Fallacy represents an overall systemic loss, but that doesn't mean there can't be localized gains. They're just gains at someone else's expense. It harms the economy to go around breaking windows, but it really can benefit someone to go around doing that. That's why "defense" contractors love war. The construction industry probably loves hurricanes. Acknowledging or advocating these localized gains doesn't mean someone fell for the Broken Window Fallacy; it merely means they might be vampiric parasitic assholes.

Re:Hmmm

[POWRSURG]

Actually, IE's ability to use their old rendering engine only goes back as far as IE7, not IE6. It's funny, given that quirks mode is really IE5.5 rendering you can test every version of IE from 5.5 up except for IE6.

Going back to the VM solution, Windows 7 Professional users are allowed to download something Microsoft calls Windows XP mode (using Virtual PC) and can create their own VMs of XP that do not require you to purchase another license. If you truly need to test in IE6, IE7, and IE8 all you need do is run all of the updates for XP for IE6 but do not install IE7, then copy that VM and run the update to IE7 and run the updates for that but do not install IE8, and then create another copy where you upgrade to IE8.

Windows XP mode is not available on Windows 8. Yet another reason why businesses may hold out on upgrading ....

Re:Hmmm

[Sulphur]

Dung shui?

It gives it that lived in look.

--

Business is picking up.

Re:Hmmm

[MightyYar]

true objective of the game is to hoard as much as possible.

You can't build wealth by hoarding money. People only flee to cash when there is (1a) a ton of uncertainty or (1b) very high interest rates and (2) low inflation. The last crisis nearly sank a bunch of companies that could no longer depend on loans. Count my company as one which has spent the last few years de-leveraging and socking away money for the next downturn instead of growing.

Re:Hmmm

[Anonymous Coward]

Was? The place where I work, all but one manager, IT, and other devs still insist if it doesn't work in IE it's broken. I'm lucky I can get people to agree to at least versions of IE shipped in the past 5 years, but even that creates issues.

A prior dev installed and created an app with a dep where the european developers don't even own a windows computer, and aren't willing to bother to test in IE.

Personally and profesionally -- I don't blame them. They're long time professional web devs, they've done their time with microsoft, and their attitude now is "fuck it, I won't be bothered to even look at the company that shat on my profession for a decade".

This shocks people. But what I can say is the app and updates works in *EVERY* other javascript enabled browser I've used -- firefox, chrome, safari, opera, kmeleon.

IE is the outlier, and microsoft is shitting this all up again.

My code never tests for "if IE, don't do X"... but... it will very often check for capabilities, which IE lies about and I end up looking for in IE case by version string. If you're less than IE9, you get minimal augmentation, very little JS or CSS...

That's just the world MSFT created -- I can't help that their JS engine runs 10 times slower than everything else out there.

The IE only world is still out there, but people are beginning to get bit in the ass and moan to their IT department about it. When you tell them it's going to to be 25% more to support IE7 in a new app, and that cost will increase annually in a 6 year project -- people get really really nervous.

It also makes IT departments and CIO's very unhappy -- but they start looking really quickly at the bottom line -- and suddenly one of two things happen:

1) You lose the bid to an asp.net dev and people shit at a modern looking page -- that doesn't work on their home computer, iphone, android, tablet ... and start complaining all over the place (relevant on public sites)

This case nearly doesn't matter unless it's just an inhouse application anyway, in which case you're already locked to a specific target platform.

2) The IT department decides it's cheaper to take a low cost bid and just deploy chrome or firefox somehow.

Re:Hmmm

[Cid Highwind]

Please. Mozilla was lightyears more "credible" than Opera ever has been, and available in 2000. Firefox is a Johnny-come-lately like Chrom(e||ium).

Re:Hmmm

[thinsoldier]

I've been doing the same from 2002 to present. Around 2009/2010 we started charging extra for IE 6, 7 and now 8.

Re:Hmmm

[DarkOx]

You just havent worked with lots of these applications. Its even fairly common for these 'web' apps to integrate with desktop programs like terminal emulators and the like via clipboard or sends keys to trigger events in the user mail client, stuff like that.

The whole point of most of these applications is workflow. Your VM will either totally break that, or be so full of wholes as to provide no meaningful protection. Seriously it would be better in every way just to a specially castrated build (in the way I suggested) of IE6 natively and have stuff actually work.

Re:Hmmm

[Ant2]

Homework: Go Google "Broken Window Fallacy" then come back and consider a reply.

Re:Hmmm

[GrumpySteen]

The broken window fallacy doesn't really describe the situation well. The windows aren't being broken. Businesses can keep using older versions of IE and use a different browser for everything other than the applications that are dependent on the older versions of IE. A broken window isn't something you can keep using internally in your house while using a different window externally.

The company I work for did that for quite a while, in fact, and installed Firefox as the browser that should be used for everything except the internal applications. They eventually replaced the applications with newer ones that work with both Firefox and IE9 and up (and Opera and Chrome, in fact, though those aren't officially supported) and the older IE was updated on all computers.

Re:Hmmm

[Impy the Impiuos Imp]

The purpose of custome extensions was to get people dependent on them so they wouldn't switch to other browsers. In this it is much like many other quasi-os-related things.

Re:Hmmm

[rwa2]

Welcome to your tax dollars at work. There were a ton of mandatory training / accounting / payroll apps that were developed and tested solely on IE6 back in the day, sometimes even as interfaces to mainframe apps.

Just be glad that the 'solution' was not to misspend another fortune redeveloping these large unwieldy government contracts that took years to develop and test to work with browser-du-jour. The 'solution" was to simply mandate that everyone stick to the "company standard" version of IE6 on their desktop. This was actually a big reason that WinXP / IE6 would not die in large corporate / government use, and probably a big reason Firefox was accepted and took off in corporate use as much as it did in order to access modern web apps that IE6 could not handle.

As for the failed IT guys, they got promoted to management, and were recently spouting off their wisdom to "make sure the new app works in Firefox, so we don't get locked into IE". The bright ones might even be champions for funding to test in 2 browsers... at least for long enough so they have something to chop when the next round of budget/efficiency cuts strike.

IE6 in a VM is something a bit beyond their ability. Maybe if they could throw your money at some monthly Citrix licenses to achieve the same thing... yeah, that's much more like what I've seen in practice.

Re:Hmmm

[Acaeris]

The Mozilla tag in most UAs is a compatibility tag for the Mozilla rendering engine and in the case of IE, is a leftover from it's fight against Netscape. So in reality, it's never spoofed Firefox. Everything just spoofs Netscape.

Re:Hmmm

[Anonymous Coward]

The broken window fallacy is that it is ok to ignore crime since it creates business for the windowmakers while ignoring the other aggregate and related problems created. However, we are talking about software here. If you wish your software to be frozen in time go ahead and enjoy your unpatched version of IE6. Meanwhile, Microsoft has a business to run and will continue to support products that propel its business. Microsoft has no obligation to support a product that is outdated. If you hitch your wagon to one vendor like this, don't be surprised when they screw you over for their own needs.

Re:Hmmm

[Anonymous Coward]

Systems don't survive to be old if they don't meet the needs of the people who use them, and almost any new system will have some period of time where the new system does not meet their needs.

Systems are like humans in this sense. Then again humans are systems as well.

Re:Hmmm

[amicusNYCL]

I'm hourly as well, but if I weren't, the analogy would be more like a dog shits on your lawn: you can either leave it there to turn white or you can pick the shit up. Either way, your life sucks a little bit more.

Option 3: ignore it for a few days until the landscapers show up and deal with it with their lawnmower.

Re:Hmmm

[BeerAndLoathing]

The windows aren't being broken

Have you seen the Metro interface?

Re:Hmmm

[Anonymous Coward]

Speaking as a failed IT guy who traded cubical for jobsite as a laborer, carpenter, and now construction superintendent, I can confirm that there is indeed a future for the sedentary desk weenie who leaves the tech world behind for a physical occupation. The intelligence, creativity, and analytical abilities that led you to IT will hold you in good stead in any occupation you choose. But be warned. You may find that your inner geek still compels you to periodically lurk on Slashdot ten years on ...

Re:Hmmm

[GreatDrok]

I have mod points but what the hell - this is bull. Everyone knew IE wasn't even following the standards of the day. The problem was that MS was busy tying IE and Windows together to migrate one monopoly into another and worse, IIS was serving deliberately broken HTML to make IE appear faster since MS had control of both ends of the equation.

IE was the standard on Windows, and it was even available in Mac and Solaris although those didn't really use the same code base or rendering engine so to say IE was the standard is disingenuous at best because IE on Windows wasn't even compatible with IE on Mac which actually had far better CSS support. Also, lets not forget that before Firefox, there was Mozilla which was the result of Netscape open sourcing Netscape 5 which was in development in the late 90s. Since they pulled a lot of commercial code out of it, the early builds were pretty badly broken but that was all we had on Linux and the web was a mess mainly because of Active X, rather than HTML. And that is where it really comes down to it, you can work around problems with HTML in different browsers, but Active X was an MS only technology developed specifically to do an end run around Java Applets. Both AX and Java Applets were a terrible idea.

Let's look at 2002 or so when IE6 was king of the hill. MS had IE for Mac still but the Solaris port was long gone. Apple looked at Mozilla's Gecko and KDE's KHTML and chose the latter to build a new browser around and they forked it to produce webkit but contributed the changes back as required by the license. In 2003 Safari appeared with OS X Panther and MS threw a shit fit and took their ball home declaring it impossible for them to develop IE on Mac when Apple clearly had info about the platform that would make Safari better (hint MS, only you did that) so the long slow decline of IE started as Apple pushed Webkit forward towards HTML5 standards, Mozilla stripped all the crap (email and news client) and released Firefox and eventually webkit found its way into other browsers, and then dominated tablets. In the end, it is tablets and phones that have proven the undoing of IE because while Firefox and Chrome have done well on Windows, IE is still quite popular, but really people are using desktops less and less and phones and tablets more and more so sites have to work with those and this means the same sites also work well with webkit and gecko based browsers so the thing that kept MS on top is gone.

People knew way back what MS was doing, but managers and developers using MS tools didn't care and so they put out non-standard sites and now that is coming back to bite them. The question is, are we doing it all again with Webkit at the expense of Gecko? Shouldn't MS be claiming to be Webkit rather than Gecko?

Re:Hmmm

[CrashNBrn]

Really, because when Mozilla/Netscape was available in 1999/2000 it was a 25MB download, which took a damnable long time over a dialup connection. Compare that to Opera 5/6, a 3.5-5MB download that also included an email client, tabs, customizable interface, low memory footprint, etcetera etcetera.

Re:Hmmm

[Anonymous Coward]

Many people and organisations had the foresight not to buy into these lock-in schemes. Should the rest of us suffer for those that didn't? I understand that many--maybe the majority--of the 'victims' of software lock-in probably relied on a salesperson or consultant to make the decision for them, but I have a hard time feeling sympathy for those that don't do a little bit of homework when buying into non-standard platforms. It's not like nobody raised the warning call.

Re:Hmmm

[pspahn]

Super off topic here, but since I started the sub-thread, I feel compelled...

I was waiting for some Chinese take-out the other night, and there was a Feng Shui book sitting on the counter, so I opened and started to flip through pages.

One page had a side bar that had a cute little anecdote. It tells of a story of John and Pamela (mostly paraphrased here, I don't remember it exactly). They had a great life, happy marriage, everything was nice. Eventually, John got a promotion and they were able to afford a new home.

A year or so after they moved, their marriage started to deteriorate. John's job became more difficult, and their son was starting to do poorly in school. As a lark, Pamela hired a Feng Shui expert to check their home and see if there were improvements that could be made. After a day or whatever, the expert came back and told them that the mirror in their bedroom was bad and that it should be removed. The bathroom was also located in the Southeast corner of their home, a location that apparently represents marriage. In essence, their marriage was literally going into the toilet.

The expert advised them to, "cease using the toilet entirely". Which they did, and it was not long before things started to turn positive, all because they stopped using the toilet.

Re:Hmmm

[pspahn]

This is pretty much what I was going to respond with. As a developer, it is my job to find clients who want to have work done. I generally care little what the work is (though I do have standards). If they want to make sure their product is compatible across all versions of IE between 6-11, that is their prerogative.

It's not so much a broken window fallacy at this point, but rather a "difficult to close window" fallacy.

The only company directly losing money in a situation like this is the company that requests this work to be done. After all, they could leave the window as-is and instead deal with the consequences. It could easily be argued, however, that this is merely an investment on their part, since having this type of cross-browser compatibility is a feature they wish to provide to users. This feature may be what helps distinguish their company from the next guy, and thus may be the reason they receive more business in turn.

Re:Hmmm

[pspahn]

Yeah, thanks! Please see responses above!

Re:Hmmm

[Anonymous Coward]

Yes that is all they have to do and surprise, surprise, they do it:

http://www.microsoft.com/en-us/download/details.aspx?id=11575

I love it:

Released on 11/9/2012
Supported operating systems: Windows 7, Windows Vista, Windows XP

Even MS seems to know Win 8 isn't worth supporting... LOL

Re:Hmmm

[inglorion_on_the_net]

Your story makes sense, except that Internet Explorer never was the standard. Microsoft was relatively late to come to the Internet.

Before Internet Explorer existed, there was Mosaic. Before Internet Explorer was the most commonly used web browser, Netscape was the most commonly used web browser. Internet Explorer became the most commonly used web browser after an extended "browser war" with Netscape, where both introduced new features that sometimes were and sometimes weren't copied by the other and/or adopted by W3C, the standards body created to increase compatibility across web browsers.

Before Netscape completely disappeared from the radar, they released (I think it was) the next generation of their software as open source, and Mozilla was born.

All this was quite widely reported on.

Certainly, people who chose to only support one web browser were well within their right to do so. However, I have no pity who people or companies who made that choice and are now stuck with the consequences. Claiming that they couldn't have seen this coming is ignoring reality. At that time in the history of the Web there were already multiple browsers and dominant browsers had already fallen. No foresight required, this stuff already happened. And there were already standards to protect you against using features that would later not be available. And there were people at the time telling this to the people making the decisions.

The decision to write non-standard code was a deliberate choice, and the non-standard code not working in other browsers was a consequence that was not only predictable, but already true at the time.

Re:Hmmm

[AK Marc]

That's macro economics. For me, personally, if I have a job fixing windows, and I go breaking windows, I make more profit. That's not a fallacy, that's reality.

Re:Hmmm

[AK Marc]

My favorite was when I switched to Opera, and so many sites would refuse to serve up a page. I'd edit the browser string, and the web page would work fine in Opera. Why would anyone enforce a browser against a user? Today, it's much more civilized. You get the warning, but it'll let you try anyway much more often.

Re:Hmmm

[Anonymous Coward]

Spoken like a true soulless manager.

Spoken like someone who is still at school.

Re:Hmmm

[Anonymous Coward]

A polished turd is STILL a polished turd. IE will always be the crappiest unstable browser POS out there, no matter what M$ thinks, says or does. Just say'n ..

Re:Hmmm

[SenseiLeNoir]

Mozilla suite was not credible for the normal users. Don't get me wrong, I used it as my main browser, hell I even used phoenix as my main browser on my laptop, but I just could not recommend it to most people at the time. Firefox was just right.

Really?

[lesincompetent]

I Say Firefox Can Impersonate IE11 Via User Agent String.

Re:Really?

[Anonymous Coward]

curl --user-agent foo ?

Re:Really?

[Anonymous Coward]

IE impersonating FF, FF impersonating IE... How about IE impersonating FF impersonating IE?

Re:Really?

[a0me]

We Need To Go Deeper.

Re:Really?

[Anonymous Coward]

How about IE impersonating FF impersonating IE?

You forgot the final "impersonating Mozilla", because ultimately that's what all browsers do. Every version of IE has been impersonating Mozilla.

Re:Really?

[Chrisq]

How about IE impersonating FF impersonating IE?

You forgot the final "impersonating Mozilla", because ultimately that's what all browsers do. Every version of IE has been impersonating Mozilla.

Which impersonated Mosaic (Mozilla was the "Mosaic killer").

Re:Really?

[Anonymous Coward]

Brad Pitt can impersonate Shia LeBouf. Doesn't mean that a good move.

Re:Really?

[Anonymous Coward]

Yeah?

Well I installed Firefox with IETab with Chrome Frame, beat THAT.
I'd run Firefox in JavaScript so we go full circle, but that's an awful thought to even think of, never mind think of implementation meth-OH GOD DAMN IT.

Re:Really?

[Anonymous Coward]

So first of all, IE has had the user-agent mocking feature in the development tools since IE 9. But disregarding the fact that this news is 2 years old, it's even more funny when you imitate the Chrome user-agent. Because now you're IE mimicking Chrome mimicking Safari, mimicking Firefox, mimicking Netscape.

Re:Really?

[AK Marc]

I've done Opera impersonating Opera impersonating IE. The idea being I wanted to show up in stats as Opera, but I didn't want to get banned from sites because I wasn't running Opera. I never found the right balance. I ended up as Opera sending out IE browser strings.

Amazing how it can boomerang

[Anonymous Coward]

Microsoft thought they could subvert the web by creating their own standards, and it worked for awhile, and now that same strategy ended up biting their own behinds. I'm enjoying this popcorn. It has Karma written in the container.

Re:Amazing how it can boomerang

[Anonymous Coward]

Oh, tasty, tasty irony. I remember the days when websites refused to work (for no reason) unless your browser was identified as IE, meaning it was Firefox (Firebird, Phoenix) that was routinely masquerading its user-agent string...

Re:Amazing how it can boomerang

[ConceptJunkie]

Microsoft is never the ones who get bitten. It's all the users who can't use the web apps developed for broken versions of IE, and the developers who have to figure out which dead chickens need to be waved over their sites to get them to work in each different (and differently broken) version of IE that comes out.

Microsoft just sits back and laughs.

Re:Amazing how it can boomerang

[Anonymous Coward]

Microsoft is never the ones who get bitten. It's all the users who can't use the web apps developed for broken versions of IE, and the developers who have to figure out which dead chickens need to be waved over their sites to get them to work in each different (and differently broken) version of IE that comes out.

I disagree. It's the IT support people who has absolutely no power over the developers, and has to deal with both the users and the corporate who keeps insisting that the Windows world revolves around them and refuses to fix issues. I've ran into these situations many times, to the point where:

• Mouse-over menu items do not work unless it is IE7 (IE8 with compatibility view does not work either). We ran right into the situation where we had to deploy brand new Windows 7 machines with no more downgrade rights to XP. Their suggested workaround is to use XP mode which increases the users' learning curve.
• A user's training progress is not saved if he/she is using IE9 (even with compatibility view). Their suggestion is to downgrade to IE8. Eventually, we will run into the same situation as above where we can only get Windows 8.
• Youtube rejects the old and outdated Flash v10.1 (yes, 10.1. How long ago was that??!) because one of their application depends on those version. Good thing Chrome comes with its own Flash, as that's about the only way to get around this problem. Even their customer-facing application does not work properly (geez). Earlier this month, we were finally given the green flag to update Flash to 11.x.
• Certain applications (not necessarily web apps) require Java 1.4, and have Windows Update deliberately disabled because they are not tested with the application. Normally, I would not have any problems with this arrangement if it does not connect to the network. However, the application gets its own updates from the Internet and the users sometimes need to look at wiring diagrams from the corporate site via the Internet. Oh, did I mention that their application requires admin privileges to run? I found one of the computers had some browser bar installed that I have not even heard of.

[/rant]
The worst part of it all is, what good is the compatibility view if it doesn't actually work?

Re:Amazing how it can boomerang

[Beorytis]

It has Karma written in the container.

It may say karma, but it smells an awful lot like poetic justice to me.

Re:Amazing how it can boomerang

[Anonymous Coward]

So Apple, Mozilla, and Chrome are all currently "subvert[ing] the web"? Maybe the problem is that standards do not yet exist for the innovation they want to provide in their software. If you want to call anybody out on the problem it should be the slow pace of the W3C or improper support for the open-closed principle within their standard.

Re:Amazing how it can boomerang

[CrashNBrn]

While there is a lot to bitch about IE over the years, a lot of the workarounds that you see floating around the web can be made redundant and unnecessary by using "Strict" html which prevents IE from rendering into compatibility mode.

Also it's quite interesting to see where the HTML spec has headed with the display property --- see inline-flex, flex, and flex-box.

One thing MS actually got right - and should of been the standard from the beginning is:

Padding & Margins were included in an elements size (width and height).

It is a complete pain in the ass trying to position elements in HTML when something like this:

[div style="width:600px;margin:10px 10px; padding:5px 5px;"]

is actually 630px wide...and to get an actual 600px width (total) you need to make the div be 570px since the element does NOT include layout spacing in the height/width settings. That aspect of HTML is nothing short of ass-backwards.

Headline and Summary Mismatch

[Internal Modem]

Wouldn't a better headline be "IE 11 user agent string changes from MSIE to IE," since most of the summary is about that?
The headline isn't even discussed in the summary.
However, it's obvious the standard ability of browsers to report a different user agent for dev and testing has been sensationalized here just for click generation.

Re:Headline and Summary Mismatch

[Anonymous Coward]

Wouldn't a better headline be "IE 11 user agent string changes from MSIE to IE," since most of the summary is about that?
The headline isn't even discussed in the summary.
However, it's obvious the standard ability of browsers to report a different user agent for dev and testing has been sensationalized here just for click generation.

No. The addition of "like gecko" is far more relevant. "gecko" is the string that most sites performing content customisation use to identify firefox and/or chroe (chrome uses "khtml like gecko", so this may result in iee bring detected as a khtml derivative)

It also doesn't seem to be an optional thing: this would be the default user agent string, although I assume it will dll revert to previous strings if compatibility mode is enabled.

Re:Headline and Summary Mismatch

[Anonymous Coward]

Wouldn't a better headline be "IE 11 user agent string changes from MSIE to IE"[.]

Yes, that would be a much better headline seeing as how that is what the story is actually saying.

TimmyFail!

Re:Headline and Summary Mismatch

[Eirenarch]

Simplest explanation (and therefore most likely to be correct) is that they were hiding. Until yesterday's leak this was all confidential and they did not want to leak information via user agent stats.

Re:Headline and Summary Mismatch

[Beriaru]

It's not a mismatch: It was impersonating other headline.

Re:Headline and Summary Mismatch

[Anonymous Coward]

"KHTML, like Gecko" - right. This just goes to show that UserAgent strings are an abomination and should never have become standard, They just serve as a way for websites to serve different content (style or otherwise) to the end user, which just means that everyone has to mimic the UA of the most popular one. Add in those vendors that think they can win more extra content than they lose by having their own little nugget in the UA, and we just get longer and longer nonsensical UA strings, which all mean the same thing: zilch.
You know what would be good? If the major browser vendors would decide one day to simply stop sending the damn thing.

Re:Headline and Summary Mismatch

[Anonymous Coward]

http://webaim.org/blog/user-agent-string-history/

Re:Headline and Summary Mismatch

[Anonymous Coward]

Browser user agent strings are an important part of your daily tracking.

Re:Headline and Summary Mismatch

[Xest]

It doesn't even make any sense whatsoever:

"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."

Why? Just because IE11 is coming out, doesn't magically make existing business apps suddenly change themselves to stop working with the "ancient and specific version of IE" they've always worked with. If it only ever worked with a specific version of IE, it'll still only work with a specific version of IE, no matter what Microsoft does with new versions. If IE11 now renders HTML5/CSS properly and completely then why on earth would they want to receive some hacked together version that's no longer relevant to the new version? Or is the implication that Microsoft should keep their browser broken so people can continue to use broken sites whilst always enjoying features of each new browser for ever? If so then that's stupid. Really stupid.

"Or does IE10+ really act like Chrome or Firefox and this will finally end the hell of custom CSS tricks?"

No it wont, because even if IE11 now works exactly like Firefox (which it probably doesn't) you'll still have a million custom CSS tricks to make Firefox and Chrome display a site the same.

Or what, you thought Firefox and Chrome consistently implemented the HTML/CSS standards? Oh, sorry to burst your bubble - no, Firefox/Chrome/Safari et. al. all require just as many hacks as modern IE versions to ensure consistency across all browsers to the greatest extent possible.

I don't really like IE, god only knows I haven't used it as my browser in about a decade now. It has a lot to complain about like the stupid compatibility mode settings that broke far more than it ever fixed, but this story is full of troll, makes little sense, and is very much wrong.

Re:Headline and Summary Mismatch

[DarkOx]

You are missing the point. Its not about supporting older it apps its about not having older apps try to support broken versions of IE.

There are lots of apps out there that see MSIE and send different pages; if they don't see IE they send a (probably) somewhat more standards compliant version. Microsoft thinks IE11 should now have behavior similar enough to the other main line browsers that users of IE11 will have a better experience with pages targeted for them. Its about IE11 users not have a degraded experience being fed pages full of workarounds designed to accommodate older versions of IE. Meanwhile applications that do look for MSIE in the string; visited by older version of IE likely to actually need the workarounds will continue to serve them.

Seriously this is a good idea.. Everybody wins, it won't break anything that probably would not have been broken by the new release anyway; except for the very most brain dead apps that look for MSIE and refuse to do anything if they don't see it; but that should be trivial to fix in most cases; unless after testing it does not work right and then you are back to it would have been broken anyway.

Re:Headline and Summary Mismatch

[cgimusic]

Not really. The browser still identifies itself as being IE11 although it now claims to render "like Gecko". This user agent doesn't hide anything. It is most likely for exactly what the article says it is for: to prevent servers sending it broken IE specific code created when IE couldn't render anything properly.

Re:Headline and Summary Mismatch

[Xest]

I'm not sure if you misread my post but you seem to be telling me I'm missing the point and then pretty much agreeing with what I said.

Re:Headline and Summary Mismatch

[Anonymous Coward]

Oh, sorry to burst your bubble - no, Firefox/Chrome/Safari et. al. all require just as many hacks as modern IE versions to ensure consistency across all browsers to the greatest extent possible.

Any evidence for the "just as many hacks" assertion or did you just pull this out your ass?

Re:Headline and Summary Mismatch

[Xest]

Yes, a ton of first hand experience dealing with clients who were brutally anal about it having to look as close to identical as possible between browsers, something that's just as hard to achieve between Chrome and Firefox as it is Firefox/Chrome and IE9.

Re:Headline and Summary Mismatch

[JDG1980]

No it wont, because even if IE11 now works exactly like Firefox (which it probably doesn't) you'll still have a million custom CSS tricks to make Firefox and Chrome display a site the same. Or what, you thought Firefox and Chrome consistently implemented the HTML/CSS standards? Oh, sorry to burst your bubble - no, Firefox/Chrome/Safari et. al. all require just as many hacks as modern IE versions to ensure consistency across all browsers to the greatest extent possible.

That has not been my experience. What I usually find is that if I develop using standard HTML/CSS and test in Chrome, I almost always get the same results in Firefox. IE9 sometimes displays a bit differently and requires a workaround of some sort with trickier CSS layouts and/or features, but usually not that much. IE8 requires extensive hacking (e.g. css3pie if you want rounded corners or gradients) and any version below that I won't even bother attempting to support.

Re:Headline and Summary Mismatch

[Anonymous Coward]

I create high visibility websites, my company traffic reports ~100,000 visitors a day. I spend roughly 60% of my time building the site, 20% testing, and 20% IE testing. While there can be slight differences in rendering between Firefox and Chrome, the vast majority of the time they can only be detected by overlaying screenshots. This even alleviates much of the IE tweaks, particularly with libs like html5shiv. If you've done things stupidly, then yes you'll have extremely noticeable differences between every browser, and "patching that up" with css is gonna be a nightmare and will take considerably longer than just fixing your errors to begin with.

My secret?
1. Avoid HTML5 unless absolutely necessary. This may not sound like it makes sense, but the simpler the page output the less problems will exist. Granted I will use HTML5 for things that will otherwise requires browser hacks or tons of obtuse javascript + extra libs just to fake (like rounded corners), but the less of it in use, the less chance any browser can crap a brick.

2. Use as little javascript as possible. You would think this would be assumed, but I've met way too many devs who thought the solution to a browser specific JS problem was... more javascript. Javascript (for most uses) only adds flash, and very rarely adds necessary functionality. Nowadays it's hard to get away from AJAX calls being made 6-8 times per page, but I will often find the source of IE trouble stemming from an AJAX call.

3. It's mostly a solved problem. Sure, there will be innovators who will turn the web programming world upside down with their offerings, but most likely you're just making a blog about your cat (or something of equal value, like a 100,000 user magazine website). If you have to live on the bleeding edge of web technology, rampant bugs is the price you pay for that. Sticking to tried and true methods can save so much time on having to rework issues because of browser specifics. I still have forms and pluggable dropdown menus I've used since early 2000's that STILL work today, and I still use them because they do everything they need to do. Whenever they cease to work I will either fix them to work or abandon them, but until that day comes they are solved problems.

Re:Headline and Summary Mismatch

[sexconker]

Oh, sorry to burst your bubble - no, Firefox/Chrome/Safari et. al. all require just as many hacks as modern IE versions to ensure consistency across all browsers to the greatest extent possible.

Any evidence for the "just as many hacks" assertion or did you just pull this out your ass?

True story: Chrome gives me far, far more issues than IE 9 or IE 10.
And if your shit works in Chrome on Tuesday it'll be busted on Thursday, only to be partially fixed on Wednesday.

Even from a user perspective I have no idea why people use Chrome outside of Google fanboyism. It's got ads and tracking out the ass with limited options to block/remove that shit, but it runs some shitty javascript benchmark a little bit faster? Who gives a shit? I'd much rather turn off shitty javascript on shitty sites via NoScript than have it complete 2 milliseconds faster. The time I save by not downloading ads (ABP in FF) vs. downloading them and then hiding them (Chrome) is orders of magnitude more significant anyway.

Re:Headline and Summary Mismatch

[Xest]

I completely agree with you and work exactly as you do, but therein lies the problem - sometimes clients demand bleeding edge so whilst in an ideal world you could stick to the simple long tried and tested XHTML1.1 + CSS2 with minimal Javascript using it where only very necessary, but for some clients that's not an option and when that's the case, it rapidly becomes clear that there's as much disparity between Firefox and Chrome, as there is between IE and Chrome/Firefox.

My point being, that most people complaining about IE and standards/differences of rendering nowadays (IE9+) and holding the likes of Firefox/Chrome up on a pedestal are just parroting old fashioned anti-MS rhetoric without actually having any justification for doing so - as you said, bugs are the price you pay for bleeding edge, but the fact is, the bugs are there, whatever browser you use.

Re:Headline and Summary Mismatch

[shutdown -p now]

I use Chrome because its UI sucks less than all other browsers except Opera (with major configuration necessary for the latter first), except Opera was unbearably buggy in the last two major releases.

You don't say!

[T-Bone-T]

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:You don't say!

[davek]

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)

I think this world needs to find a quiet room in which to do long division. That is all.

Re:You don't say!

[ConceptJunkie]

I say that this is the kind of nonsense up with I will not put.

Re:You don't say!

[PPH]

Grammar Nazi!

Like gecko

[Anonymous Coward]

Or that they've added 'like gecko' which is actually the point of the article. Although they might have been better impersonating Chrome since it's more popular:

http://en.wikipedia.org/wiki/Usage_share_of_web_browsers

Final Resolution

[marienf]

Ignore User-Agent and redirect any request containing one or more headers starting in "X-MS" to http://www.kmfms.com/

Re:Final Resolution

[Anonymous Coward]

That site is about as outdated as IE4.

Sigh

[ledow]

The day that the first website was able to detect what client was being used to view it, we were in trouble.

Whether it was people trying to "fix" ancient Opera (and still some sites had such tests until very recently), people telling you what browser to use (i.e. not accepting Netscape / IE of certain versions - I still know of a UK bank that stops you logging in as certain browsers, but fake the user agent string and it works 100%), or just plain faffing about (i.e. iPlayer detecting the user-agent to see if it's "allowed" to download the iPad streams, etc.).

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. If someone has a client that can't render a standard page, then that's their problem and we should have left them to it - eventually they would have complained to the relevant person and their browser would become closer to the standard. We would also have killed off abominations like non-standard HTML tags and everything else.

If you have CSS, in this day and age, that does detection of the user-agent, then that's your problem - you manage it and if it ever affects my usage of it, I'll be complaining and going elsewhere. If you have a browser that can't change the user-agent at will and still work, then that's a crap browser (purely because the user should be in control of the website they are displaying and not the other way around). Precisely because we're all too stupid to just make browsers and websites conform to a common standard.

Personally, I use Opera - have done for nearly a decade now. If it doesn't work in Opera, I move on and go somewhere else. The number of times it's stopped me doing something I wanted is vanishingly small (probably 4-5 incidents in all that time), and I've blamed the website every time - not Opera (because in every instance, faking the user-agent to something else has fixed the problem, so it's not the browser). It's cost several small companies my custom (not that they would be able to tell, or care).

Fact is, my life is too short to play games with accessing your website. If I can't, I move on. End of. I've even moved my bank accounts because of it (NatWest, in the UK, had a website that refused to work with anything but ancient versions of IE or Netscape - yes, it actually said Netscape even in the era of Firefox - and they refused to fix it "for security reasons", so I moved on. Presumably they've fixed it now, but I don't really care because the damage was done by not being able to log into it at my convenience).

You have a website because you want people to come to it and see your content. Hiding that content, because you don't know how to properly display it, is so counter-productive, I can't even begin to explain it. If the fancy shit you're pulling messed up my browser (which conforms to all the ACID tests and general compatibility with EVERY OTHER SITE on the planet), maybe you should take that fancy shit off?

Re:Sigh

[Anonymous Coward]

eventually they would have complained to the relevant person

The problem is/was that the relevent person was often mistaken for the website owner rather than the browser maker, so the complaints were directed at the wrong people. The website owners cared about their users so they started inventing tricks to get around the browser differences for their users.
The browser makers however were always just implementing their "own standard" AKA as making shit up as they went along as they had very little financial risk invested in their users so the users needs could basically be ignored. In the early days it was "my way or the highway" from the browser makers. This is why the environment we have now is what it is. It is actually caused by a lack of corporate will to cooperate with others. I do believe this is latently still the case. Even though we have better compatibility across modern browsers now, it is not done for the philanthopy benefit of the users but for the convienence of chasing the dollar.

Re:Sigh

[Spad]

NatWest, in the UK, had a website that refused to work with anything but ancient versions of IE or Netscape - yes, it actually said Netscape even in the era of Firefox - and they refused to fix it "for security reasons", so I moved on. Presumably they've fixed it now, but I don't really care because the damage was done by not being able to log into it at my convenience

They haven't. You still have to fake UA strings to use most browsers with their online banking site.

Re:Sigh

[Anonymous Coward]

maybe you should take that fancy shit off?

unless someone high up (your boss, in marketing, etc) requests it, then you have no choice. and PHBs love fancy shit.

regardless, your post is spot on!

Re:Sigh

[Xugumad]

> 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:Sigh

[ledow]

"I'm sorry, your car doesn't have a standardised fuel cap. Is the fix to:

1) make your car have a standard fuel cap?
2) force everyone to use your new fuel cap ?
3) make pumps sense by the numberplate which model of car they are filling up and change the fuel cap to the right one each time?"

Whatever option you choose, 3) is really incredibly stupid and puts the onus on fuel stations to make the changes rather than the idiot that wanted to be different for no good reason. It might be *A* solution, but *THE* solution is to just look at the guy who can't fuel up their car with a "You pillock" look until they realise they've bought a turkey - and then let Ford / GM / whoever supply an adaptor to him rather than you having to carry 20 adaptors for all the different types of fuel cap there are.

All you've done is encourage Safari to be the exception to the rule, with a broken implementation that now doesn't have to be fixed (because you "fixed it" for them on your end).

By way of analogy, if - say - a browser can't upload more than a 2Gb file, then you're choosing to detect the browser that can't, chop the file up into little bits just for them, and pass it on. You're fixing their crappy browser for them, so you have to take all the burden for their mistakes. That's just not sensible compared to say "Sorry, you're browser is crap and can't handle downloads the size of your average DVD from 5 years ago. Maybe you should investigate alternatives."

Re:Sigh

[jimicus]

The day that the first website was able to detect what client was being used to view it, we were in trouble.

You haven't needed to do that ever if you want to detect IE.

IE automagically turns backslashes in a URL into forward slashes. AFAIK, no other browser does this.

Re:Sigh

[Secret Agent Man]

If someone has a client that can't render a standard page, then that's their problem and we should have left them to it - eventually they would have complained to the relevant person and their browser would become closer to the standard.

Not every client can get their browser changed. Not every client is willing to lift a finger to improve their standard either, and can happily move on to a competitor who's more than willing to accommodate them.

If you have CSS, in this day and age, that does detection of the user-agent, then that's your problem

Tell that to clients who want a particular design. They don't care about standards, they care that the pages look correct across all browsers, including legacy versions of IE dating back to 7 (or, even in some cases, 6), and on mobile devices. They're not going to say "oh, IE 7 doesn't support [standard]? Okay, you can cut that piece out then. ^^" They're going to not appreciate that your pages don't look right, and they won't care why.

Personally, I use Opera - have done for nearly a decade now. If it doesn't work in Opera, I move on and go somewhere else.

Yeah, filtering one's web traffic based on user-agent string is stupid. I think it was a poor attempt to say "we only have to design for one web site!" by "forcing" users to use "correct" browsers. That being said, I've not seen this practice used for some time (maybe because I use Chrome? I've not seen it at work either...)

And as others have said, the fancy stuff usually is mandated by people over the site designers' heads, either for the client, or because Mister Boss Man wants it, or for some other reason. Yes, in a perfect world, we'd only use HTML, JavaScript, CSS, etc that we knew worked 100% of the time across any medium that can display a web page, but that just isn't how the world turns.

Re:Sigh

[twdorris]

If someone has a client that can't render a standard page, then that's their problem and we should have left them to it - eventually they would have complained to the relevant person and their browser would become closer to the standard.

Are you new here? You may not remember the days when this mess all started. IE was king and you *had* to work around it. You couldn't just let it be "their problem" and "left them to it". That's so "counter-productive, I can't even begin to explain it". These customers (sheep running IE) would come to *you* in droves asking why they couldn't view your website. And your response was going to be "because IE doesn't display my standards-compliant page"? Wow...no...that doesn't work.

Nowadays, things are clearly different. Which is great. But to suggest developers should have never used the user-agent tag to distinguish browser differences is ludicrous.

Re:Sigh

[Kjella]

Sure, in an ideal world it's not needed. But in the real world browsers have bugs and no matter how much you blame the browser to the average user your website is broken. Unless you want to limit yourself to the minimal subset of HTML that no browser has managed to screw up you want the ability to work around browser bugs, which means you need to know what browser they're running. That is why so many file formats have a field to let you know what software wrote it. Honestly if you haven't had to implement an IE-specific hack, then you can't have done any web development in the last decade. Because of IEs conditional comments that serve much the same purpose, we could provide Firefox/Opera (and if it were today, Chrome) with a standards-compliant page while still making it work with IE. Otherwise the standards compliant web would have died on the drawing table.

Re:Sigh

[RaceProUK]

NatWest, in the UK, had a website that refused to work with anything but ancient versions of IE or Netscape - yes, it actually said Netscape even in the era of Firefox - and they refused to fix it "for security reasons", so I moved on. Presumably they've fixed it now, but I don't really care because the damage was done by not being able to log into it at my convenience

They haven't. You still have to fake UA strings to use most browsers with their online banking site.

Bollocks - the website works perfectly fine in Firefox without having to fake anything. I bet if I tried it in Chrome, it would also work fine.

Re:Sigh

[RaceProUK]

The day that the first website was able to detect what client was being used to view it, we were in trouble.

You haven't needed to do that ever if you want to detect IE.

IE automagically turns backslashes in a URL into forward slashes. AFAIK, no other browser does this.

Firefox does - just tested it.

Re:Sigh

[ConceptJunkie]

And your solution puts the onus on the user to become an expert in software. The whole point of the exercise is to give the user what he wants/needs. If some part of the system is broken, it usually ends up being someone downstream or upstream that gets the blame. After all, if you are refusing to support broken browsers, then you might be encouraging a gigantic company to fix their software (yeah, like that ever happens, short of a world-ending cataclysm), but you are also jeopardizing your users' experience and ultimately your own business. Playing hardball like you suggest, however tempting it may be, and it's _very_ tempting, costs your website users, and that means fewer customers. At some point you might have to decide to cut your losses, but as a website developer, just like any other software developer or anyone providing a service or product, you _must_ bend over backwards to give your customers the best value/most convenient experience/best service possible.

It's an ugly situation, but we should be thankful that Mozilla and other competitors have gained enough ground against the Evil Empire that there _are_ alternatives... and that the Evil Empire has had to react in response to chinks in their monopolistic armor. Unfortunately, that leads to the temptation of "This website works best in $BROWSER."

I guess the ultimate question is: Should we be able to treat browsers like telephones? No matter what kind of telephone you have, at some point it either allows you to call other telephones or it doesn't. Beyond that, they are all essentially identical to the person placing or receiving the call.

Re:Sigh

[Anonymous Coward]

Standards lag behind the usage. Standards are created when the ISO body sees various parties resorting to different tricks to do something similar. The alternative would've been to publicize and standardize an extension before implementing it. But that would've slowed the innovation and everyone would be pissed at the glacial pace.

Re:Sigh

[Inda]

I second your bollocks.

Yes, there were early problems. Problems like "Firefox 11 is not supported" a day after I replaced FF10 with it.

Natwest soon learned. I can't remember the last time I had browser issues with them.

Natwest trying to get people to install CleanMyPC or whatever it's called, sure, I'll rant with you all day about that, but not browser support.

Re:Sigh

[Anonymous Coward]

The problem is that standards are slow to adopt, and we wouldn't get anywhere without browsers implementing their own features. For example, javascript started out as a netscape specific feature and xmlhttprequest started out as an IE6 specific feature.

Re:Sigh

[Inda]

Redirects. I'll rant about their redirects too if you like. Why can't I stay on their domain to do my banking? Why must I be redirected? Who knew that https://www.nwolb.com is actually part of Natwest? Not me and not their internet security staff when I phoned them (tell me I have malware again! I dare you.)

Re:Sigh

[Xugumad]

Have you ever tried telling a userbase that there's a problem with their browser and they should change? If you're lucky enough that they read the notice instead of just hitting reload a few dozen times then complaining it doesn't work, generally they'll tell you that it works elsewhere, and why not on your site.

It also presumes they can move browser; less of an issue with Safari, but we've had to put in work-arounds for IE6/7 for users who are locked into those browsers by their employer (who really, really doesn't care enough to change).

Oh, and unless you either don't have to support the users, or have a very generous allocation of support staff, telling 20-30% of your total users to change browser is going to involve the support staff being hopelessly swamped with related questions and issues.

Re:Sigh

[Anonymous Coward]

If someone has a client that can't render a standard page, then that's their problem and we should have left them to it - eventually they would have complained to the relevant person and their browser would become closer to the standard.

We used to, and they did. Except back then, the people to which they complained had no reason to act the way you or we expected. That led us to the mess as it stands now. So by all means, please take your suggestions, devoid as they are of real-world historical context and how the real world loves to severely punish wide-eyed utopian idealists, and kindly fuck off before you make it worse.

Re:Sigh

[Nemyst]

There used to be a time where yes, detecting the browser through the user agent string was pretty much the only way to fix inconsistencies.

Nowadays, there really isn't much of an excuse for it. The good practice is to make a standards-compliant website, even going as far as using CSS3 and HTML5 gimmicks (they degrade gracefully if unsupported) if you want.

If a significant proportion of your expected users have a non-compliant browser, or if you're using something that's fairly cutting-edge and necessary for your website, the way around it is not browser sniffing. Browsers can change, and you don't want to end up having a fix break things in newer versions. Instead, you'll use something like Modernizr, which detects features. It's a much more logical thing to do, too: instead of, say, implementing an IE sniffing because drop shadow doesn't work on it, why not just detect whether drop shadow is supported, and bypass the logical leap of assuming IE doesn't? This future-proofs your website and often simplifies things, as you don't have to check for all possible browser variants that don't support a feature you want.

If the user doesn't have JavaScript enabled, he/she is either using an extremely old or locked down browser, in which case it probably wouldn't render well anyway, or he/she is tech savvy enough to know how to do that and thus likely runs a modern browser, in which case it'll render fine since it's standards-compliant in the first place.

Re:Sigh

[Anonymous Coward]

and this is most of my thinking in regards to Noscript being set to DENY ALL by default. I prefer to whitelist only those sites I actually have to and forget about broken websites that simply refuse to work because their coded using Flash or absolutely depend on J-Script for everything. To me these site are broken as they aren't accessible to the disabled (screen readers or other means). Pitiful and if it's a major corp, I tell them while passing the complaint to the proper federal agency that they are violating the Accessibility rules (EU/US). They tend to fix things pretty damn quick once the Gubermint gets involved by at least having their system detect/redirect to an accessible website.

Re:Sigh

[Anonymous Coward]

not on mac. just tested it.

Re:Sigh

[RaceProUK]

not on mac. just tested it.

OS X doesn't use backslashes in file paths, so it's no surprise.

Re:Sigh

[NewWorldDan]

I specifically prohibit my developers from looking at the user agent string. Heck, I use IE9 as a browswer, and I've changed the user agent string to 'null'. Anyway, if it doesn't render correctly in Chrome, Firefox and IE, then we find a new way of doing things. Browswer specific code is never allowed. Not even for detecting mobile browsers. There's already a link to the mobile version. If they want to use the main site, that's their business.

Re:Sigh

[cybernanga]

NatWest works fine with Safari on OS X, and has done for quite a while now.

Re:Sigh

[FuzzNugget]

And your solution puts the onus on the user to become an expert in software.

Oh, baloney.

You're not asking them to download some source code and compile it (and all the headaches of requisite compilers and dependencies and so on)

Just provide them with a direct link to a non-shitty browser and the installation process holds their hand the entire way.

Re:Sigh

[Anonymous Coward]

Aside from encouraging bad developer behavior, User Agents are a privacy risk. Point releases in browsers contribute to having a unique set of headers which may specifically identify your computer. It doesn't work as well as a cookie, but you have no control over it. Head over to Panopticlick for a demonstration.

Re:Sigh

[AK Marc]

Personally, I use Opera - have done for nearly a decade now. If it doesn't work in Opera, I move on and go somewhere else. The number of times it's stopped me doing something I wanted is vanishingly small (probably 4-5 incidents in all that time), and I've blamed the website every time - not Opera (because in every instance, faking the user-agent to something else has fixed the problem, so it's not the browser). It's cost several small companies my custom (not that they would be able to tell, or care).

And you never go to sites that require ActiveX?

Wow...

[Curupira]

...karma really DOES exist!

So Microsoft hurt itself?

[erroneus]

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.

[danhuby]

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.

Re:Parsing user agent strings = bad.

[Hal_Porter]

The last time I did any HTML and Javascript was back in the IE3/IE4 days i.e. around 1997. We supported IE3 and up and Netscape and up. Netscape support was unofficial because the management didn't think it was necessary.

We had verification on the server. If the browser was recent - i.e. IE4 or later or Netscape we enabled verification on the client too. So using the app on a modern browser was faster and placed less load on the server. But the server was set up so that even a completely dumb client with no Javascript support was still supported. The server generated the pages using ASP - so it could simply not send Javascript to a downlevel browser.

Of course it was horrible primitive HTML - basically tables and 1 pixel gifs for layout - because CSS wasn't supported by the browsers people actually used. Still it looked OK on the browsers we supported. The actual UI was a bunch of HTML form elements - mostly combo boxes, input text fields and checkboxes.

Re:Parsing user agent strings = bad.

[gl4ss]

sure, but if you managed to get ie 6.0 into the support contract then you'll keep it as that and refuse to serve pages to anything else.

why? so you can bill for removing that requirement.

Re:Parsing user agent strings = bad.

[Anonymous Coward]

Mod parent up ...

You can use this as vetting criteria. Developers that depend solely on parsing User-Agent string should not be hired. You can also check what kind of shoes they wear ... lace-up's or slip-on's. If they can't tie their shoes, that may be an indication of something! ;-)

Re:Parsing user agent strings = bad.

[Anonymous Coward]

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.

Completely agreed. The mangled UA strings we have today (in all browsers) are basically the result of a fifteen year arms race against poor quality sites doing UA sniffing.

The whole thing is basically a reaction to the kind of code described here: browser-version-checker-malfunctions-for-versions-under-or-equal-to-ie8

Re:Parsing user agent strings = bad.

[tepples]

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.

Say you want users of web browsers on full-size devices (PC, 10" tablets) to see a headline and a 1-sentence abstract on your site's front page, while users of web browsers on phones' smaller screens get only a headline. You could send the abstracts and hide them in the small-screen CSS activated through a media query, but then that'd slow down page loads and cost the user money, as U.S. wireless users pay by the bit to send or receive data. You end up having to look at the user agent to determine whether to redirect the user to the URL with abstracts or the URL without them.

Or say you're designing a shopping cart application, and it shows a list of the items that the user most recently viewed within a browsing session. The web application distinguishes one user's browsing from that of another user using an anonymous session cookie. If a search engine's bot doesn't send the session cookie, and they don't, then the web application will create a new anonymous session for each page view. On a site with 100,000 products, this can clog up the session table with 100,000 different anonymous sessions for each pass by each bot. Distinguishing "Googlebot" and the other major spiders from users interested in products improves performance of anything that hits the session table.

Re:Parsing user agent strings = bad.

[danhuby]

In my view these are both examples of the bad practice I mentioned in my post.

Your first example requires a continually maintained list of phone browser user agents. The second example, a list of search engine user agents. There are a vast number of devices and browsers and you're unlikely to stay on top of this list.

Creating a unique table row for every visitor to your site isn't the best application design for the reason you point out - a browser without cookies is going to add a new row with each access. My approach would be to generate a session key but store that in a cookie only, with session related data stored only as necessary (e.g. basket rows). Search engines are unlikely to POST so you won't get any basket rows from them (if you're adding to a basket via GET then you will have other issues).

Re:Parsing user agent strings = bad.

[tepples]

Your first example requires a continually maintained list of phone browser user agents.

Then what's the correct way to determine how much mark-up to send to the user so that the user doesn't end up paying per bit to download the mark-up for a bunch of elements that are just going to end up hidden with display: none?

My approach would be to generate a session key but store that in a cookie only, with session related data stored only as necessary (e.g. basket rows).

You still need to store something that states whether or not a session has been invalidated and when it expires. Otherwise, you leave yourself open to a replay attack like the one we discussed recently.

Search engines are unlikely to POST so you won't get any basket rows from them (if you're adding to a basket via GET then you will have other issues).

You still add an item to the session's list of Recently Viewed Items via GET, unless you want all links to a product page to be form submissions so that the user can't open a product page in a new window and can't easily recommend a product page.

Bork Bork

[TheP4st]

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

Karma is a Bitch.

Re:Bork Bork

[Anonymous Coward]

This affected all 17 Opera users, especially the 3 of those that actually went to MSN.com...

Re:Bork Bork

[TheP4st]

Wow... original, did you come up with that yourself?

Re:Bork Bork

[Anonymous Coward]

No. Everyone knows there are only 3 MSN.com reading Opera users.

Re:Bork Bork

[TheRaven64]

Actually, in 2003 Opera was not that obscure. The wikipedia statistics have 94.43% for IE. Netscape and Mozilla each had about 2.5% and Opera had 0.66%. Back then, on any *NIX system, Netscape was old, Mozilla was crashy (and slow), Opera was fast. On Macs, the same was roughly true, except you had IE 5.5, which had transparent PNG support (which Windows lacked), but implemented an old draft of CSS standard. On Windows you also had IE, which was ubiquitous, but Opera was the best choice technically. They also released binaries for FreeBSD and NetBSD, on a few architectures. It was sufficient of an improvement over the competition that I actually paid for it.

Re:Bork Bork

[Anonymous Coward]

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

Karma is a Bitch.

Most stylesheets are sent 'deliberately'. What MSN did there was exactly what IE10 has been on the receiving end of from many web sites. When MSIE is detected, it gets an old faulty IE stylesheet instead of the same stylesheet as Chrome/FF gets which would have given much better results. Same happened with Opera on MSN. It's laziness on the side of the web site programmer, who don't bother to properly update and test their code.

Re:Bork Bork

[Kjella]

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

Of course the actual story is that Opera had a bug which that style sheet worked around, when they fixed it in a new version the page looked broken because they still got the modified style sheet. So yes it was deliberate but not malicious, in fact someone had made extra effort to make it work on Opera however the PR opportunity was far too good for Opera to pass up. That's one problem with browser-based hacks, if you're not around to maintain them should you assume the next version of IE will be 100% standards compliant or that most the IE6 hacks would also be required for IE7. It wasn't as obvious as you'd think, to the clients it looked like your site was incredibly fragile when it broke horribly on any new browser version. Those were dark days, long before real standards compliance.

Re:Bork Bork

[Anonymous Coward]

Opera was awful on OS X in 2003. Also, by that point we had Chimera and Safari.

Re:Bork Bork

[Anonymous Coward]

Your mother uses Opera.

Re:Bork Bork

[Anonymous Coward]

I remember Opera being the browser I'd see most often, after IE, back in 2003. I also remember one of its most-useful features being the ability to easily impersonate UA strings. Just have the prowser pretend to be IE and broken websites would suddenly work. This may have inflated IE's reported popularity just a wee bit, while concealing some of Opera's.

Re:Bork Bork

[Poingggg]

No, Opera did not have 'a bug which that style sheet worked around'. I am too lazy to find a link, but when Opera changed the user agent to 'IE' (or Firefox, I'm not sure) without changing the renderer or anything else, the pages rendered perfectly. So there was no bug in Opera, MS borked the stylesheet they served to Opera.

Re:Bork Bork

[Anonymous Coward]

According to the linked article, it also didn't work in IE6.

Re:Bork Bork

[rarumberger]

You clearly did not read Kjella's post. Your account and Kjella's do not differ, except in that you assume malicious intent, and Kjella is rational.

Re:Bork Bork

[Poingggg]

Sorry guys, but I believe this more:

https://people.opera.com/howcome/2003/2/msn/

Microsoft NEVER plays fair, never has and probably never will.

Re:Bork Bork

[Kjella]

No, Opera did not have 'a bug which that style sheet worked around'. I am too lazy to find a link, but when Opera changed the user agent to 'IE' (or Firefox, I'm not sure) without changing the renderer or anything else, the pages rendered perfectly. So there was no bug in Opera, MS borked the stylesheet they served to Opera.

If only there was a tech site full of geeks running in 2003 that could provide insight into this, oh wait here...

Yeah but the stylesheet for opera is designed to work with Opera v6. The bug in the stylesheet was hidden by a bug in v6 which was only last week fixed in v7. sure, there's a bug in the stylesheet, but that doesn't mean it's new or that it was placed there specifically to break Opera v7.

Hmmm... although the thing is, the item in the stylesheet which they claim is broken (ul tag style) is the same in the Nav6 stylesheet.

Mozilla renders the page perfectly, and it gets the nav6 stylesheet with that very same ul {} declaration... here's all three listed together:

Opera: ul {list-style-position: outside; margin: -2px 0px 0px -30px; list-style-image: url(http://msimg.com/m/8/bullet-black.gif);}

Nav 6: ul {list-style-position: outside; margin: -2px 0px 0px -30px; list-style-image: url(http://msimg.com/m/8/bullet-black.gif);}

IE6: ul {list-style-position: outside; margin: -2px 0px 0px 8px; list-style-image: url(http://msimg.com/m/8/bullet-black.gif);}

It seems my memory was a bit off (hey it's been 10 years), it was Netscape Navigator that had the bug but Opera v6 rendered fine anyway - but would have with the standards compliant CSS too. Then they released Opera v7 and Opera broke because it got the fix for Netscape Navigator that now caused it to render 38 pixels off.

And just for laughs, one of the sigs from 2003:

World IPv6 Launch

Maybe in another 10 years, mate...

Re:Bork Bork

[Poingggg]

OK, looks like you are right after all. I am not so at home in these matters, but if I interpret this right than Opera 7 got served the style sheet for Opera 6, which contained a bug that Opera 6 worked around but 7 didn't (as you wrote in your first post).
I do have a tendency to suspect MS in all they do, and as far as I could tell from the link in my previous post it looked like foul play all the way. But I guess a site full of geeks has more authority in these matters, so all I can do is humbly admit that I was wrong.
(Dang, I hate being wrong! :-) )

(BTW: Nice sig, I totally agree with that one!)

Re:Bork Bork

[Billly Gates]

In 2003 the writing was all the wall.

IE 6 was the wave of the future. Our memories become distorted based on opinions if you ask any psychologist! But lets look at the statistics leading up to 2003. No one and I mean no one survived competing against MS. Opera, Netscape, others were dying month after month. IE 6 was the best browser as Mozilla and Netscape had more quirks in that time frame. No I am not trolling, but rather serious. They improved afterwards.

Webki/khtml, opera, netscape/Mozilla, and IE all failed the ACID test. Not one browser was good back in 2002 and 2003. This was not an IE only thing we believe as today.

So it makes great business sense to standardize on IE 6 10 years ago as Opera was probably going to die and so was this silly Mozilla hack called Firefox. IE integrated with active directory and was eventually going to hit 100% marketshare whether we liked it or not soon. That 3%? Who cares in business I.T. decides your browser.

Of course today things are better. Even IE does not exhibit bizarre behaviors. But Opera in 2003 sucked as well as khtml and even Mozilla had many Netscape quirks in it that required workarounds.

Re:Bork Bork

[Anonymous Coward]

Holy "Citation needed", Batman!

Re:Bork Bork

[toddestan]

The way it played out as I remember was:
1. Opera had a bug in it that caused msn.com to render badly.
2. Someone at Microsoft hacked up a custom style sheet to make msn.com render properly in Opera.
3. Sometime later, Opera fixed their bug.
4. Now, msn.com renders badly in Opera due to the custom style sheet that is no longer needed.

Really, it was one of those cases where both parties did something wrong. Opera shouldn't have had the bug, and Microsoft either shouldn't have created the custom style sheet, or made it so it only sent it out to specific versions of Opera so any new version of Opera would get the standard sheet (and if needed, that version could be later added to get the custom sheet).

Re:Bork Bork

[Poingggg]

You are right. As I already said in another reply, the only thing I can do is humbly admit that I am wrong. Thanks for your clear explanation.

Release Schedule?

[craigminah]

The OP says Firefox and Chrome have a faster release schedule but fails to mention the content of those releases by Firefox and Chrome are minor where the updates of IE are major. You can't say that because Firefox and Chrome release a new "version" every month that they're better or more innovative. They just consider everything worthy of a version upgrade. Add a built-in pdf viewer, increment the version number; change the font style, bump the version up one; add some other good or inconsequential feature, release a new version. It is really misleading but it obviously tricked the OP into thinking it's better.

Re:Release Schedule?

[Cenan]

It is better... In an ideal world, every tiny change would get it's own unique version number. Fix->Test->Release
If you have a fix, why sit on it till 999 other issues have cropped up and are (supposedly) fixed? And no, a version number plus a list of KB articles for updates is not the same thing. It doesn't matter if all you change is the font, if that one thing is what breaks it for someone. But it makes it a hell of a lot easier to figure out why it's not working anymore, as opposed to a mammoth 1000 item changelog.

Re:Release Schedule?

[Merk42]

Oh yes, Firefox 137 sounds much better!

Re:Release Schedule?

[Cenan]

Who cares? It's a number that distinguishes one incarnation from another. Only a moron would think they are comparable between distinct applications.

Re:Release Schedule?

[craigminah]

But that's part of the argument the OP raises in support for Firefox and Chrome as a benefit...obviously to most people on ./ it's a nonfactor but to the general public (the the OP) it's a perk.

Re:Release Schedule?

[Imagix]

Because big users can't afford to test every little point release that you make every 2nd day. Acceptance testing may take weeks.

Re:Release Schedule?

[DMUTPeregrine]

Really, it's a matter of where you put the separators, or if you have them at all.
1.3.7 vs 137 vs 1.37 etc. A version number is just a way to determine if two copies of the software are the same, and if they're different it can be used to look in the changelog and find what the differences are. Distinguishing major and minor features isn't always very useful, as even minor changes can break things.

I only use 1 IE-specific conditional.

[Anonymous Coward]

<!--[if lte IE 9]>We have detected that you are using Microsoft Internet Explorer which may be running the plugin called StealYourCreditCardInformation.virus.B. That plugin tends to break the layout you see on our site because our site is very secure and that plugin can not operate on our site. If things look broken, we suggest you uninstall that plugin, or use a good web-browser like Google Chrome, Mozilla Firefox, or Opera.<![endif]-->

That goes on the top of the page. I then go out of my way to make use of all CSS that triggers IE-specific bugs.

Re:I only use 1 IE-specific conditional.

[CaptQuark]

LOL!! Thanks. I may never actually use that on my web pages, but it did make me smile.

User-Agent is broken

[Anonymous Coward]

The format is not well-defined, hence hard to impossible to correctly parse. JavaScript should query the navigator object and not use the UA at all. A very long time ago the UA was used by web statistics software but has since become a relic. Just ignore that it even exists. That header should have been deprecated years ago.
So if an application breaks because of the UA, beat the developer with a stick

Same old story

[gidoca]

Nothing particularly new.

Not surprising

[DrXym]

Bad JS has code such as "if (document.all || /MSIE/.test(navigator.userAgent)) isIE = true;" or some variant thereof. So changing the user agent and also removing any IE specific extensions like document.all, CreateObject() etc would be a good way to force browsers down the other path which is presumably more browser agnostic. IE could implement a whitelist test which enables the cruft on intranets which absolutely refuse to work otherwise.

I suppose we have to be grateful for MS in doing this providing they're now supporting standards rather than half implementing them. Sites shouldn't be testing for Gecko or Webkit either though or they'll be creating a problem for themselves down the line just like the one with IE 6/7 now. They should be programatically testing the features they need and avoid what the browser engine is as much as possible.

Re:Not surprising

[Anonymous Coward]

You don'r need to test UA strings to branch code, you object detect, so for example

if(window.applicationCache) { // browser has this capability
} else { // use a fall back
}

You should never (there are edge cases, maybe) test the UA string but capabilities.

Re:Not surprising

[DrXym]

Hence my last sentence.

Re:Not surprising

[Anonymous Coward]

If I remember correctly, the only way to make AJAX work is to detect the user agent. If it is IE, it has to be a certain way. The other browsers use a standardized method. That's about the only thing that I find acceptable for detecting the user agent. Anything else can use the <!--if IE--> tags to fix IE.

Firefox does it with a plugin

[houbou]

Firefox can do the same thing with a plug-in, used it alot for testing mobile devices as I would enter their user agent string.

Re:Firefox does it with a plugin

[Anonymous Coward]

That's not a plugin; it's an extension.
Why the hell are there so many alleged geeks who still don't know the difference, all these years later?

It always did that.

[Anonymous Coward]

Every version of IE, at least back to IE4, and probably even older, had the user agent string "Mozilla/4.0" and later "Mozilla/5.0".

For those who don't know, Mozilla was the code name of Netscape Navigator, later it became the name of the new open source version of Netscape, called Mozilla, and ended up as the name of the organization behind Firefox.

So no, they didn't start impersonating Firefox. They just updated the string to impersonate a newer version.

Summary is wrong

[onyxruby]

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.

The faster release schedule doesn't have a damn thing to do with Firefox and Chrome gaining ground over Internet Explorer. In fact the fast release schedule has blatantly hurt Firefox gaining ground over Internet Explorer as enterprise after enterprise has blatantly refused to distribute Firefox until they pull their head out of their ass on their release schedule. This causes a logistical support nightmare and the idea that this is going to /increase/ exposure through peoples works places is nonsensical. Firefox, go back rapidly releasing patches, instead of new versions, got it?

Internet Explorer losing ground has everything to do with two things. Experience and exposure.

When you use another browser you have a better experience because you can use things like Ad Block Plus or Flash Block to prevent your computer from being taken over by ads. You can use plugins to customize your experience for any number of other things in ways that you like. Internet Explorer simply doesn't support extensions that let the user improve their experience. You, the user, remain in /much/ greater control and as a result have a much better experience.

As for exposure that is an inevitable result of people being exposed to additional browsers through things like the browser choice menu in the European versions of Window and through smart phones. People are no longer in the position of having a browser monopoly and being quite so limited to a single default browser.

History repeating

[Bert64]

IE has always identified itself as Mozilla, i believe the current versions still identify themselves as mozilla/4.0 (ie netscape 4).

uhm..

[SuperDre]

it's not like you don't have to check with Chrome or Firefox if it supports the features you want to use, that's with every browser..

Re:uhm..

[ferret4]

you don't have to check any such thing if you code with graceful degradation. For example, a browser either supports CSS3 rounded corners or it ignores what it doesn't understand - nothing breaks, the user just gets square corners instead. If you design and code with that in mind, there is no need to perform checks on the browser vendor and version.

Business smarts?

[guyniraxn]

If it's the user agent string and not the updated browser functionality that breaks the "business app" then why bother to use the user agent string at all? The companies I've worked for have always stuck with the oldest version of IE they can get away with and if you managed to somehow get admin rights and update IE, the app will load but not work. User agent strings should not matter in a professional, managed software environment so I fail to see how a new string in a new browser that corporate isn't going to deploy anyway will break anything.

Sure

[Murdoch5]

This is why I don't develop for IE at all. I need a web platform that doesn't break CSS and HTML code. If Microsoft wants to change the string from MSIE to IE then they better have finally fixed CSS and HTML handling. If they have I'm more then happy to support IE as part of my web platform, if not then I'll just have to put an update to block this new IE 11.

Re:Sure

[Anonymous Coward]

And YOU and your ilk are part of why this is all a mess. Putting browser specific fixes, code paths and blocks is why all this shit has had to happen, incompetent developers that resort to detecting UA strings instead of writing sites to the standards and letting browsers fix there own bugs.

Re:Sure

[Anonymous Coward]

Hey, I developed my site only targeting Chrome using HTML5 and CSS3. When I tested it in Firefox and IE9/IE10 it rendered exactly the same in all three browsers. It's amazing what good web programming conventions gets you. The only "hack" I had was for touchscreens so :hover would work properly (and done with a capability test in js).

Re:Sure

[Merk42]

What sort of CSS/HTML issues are you having with new versions of IE? This isn't IE6 anymore.

Re:Sure

[TheSkepticalOptimist]

I hate web developers like you.

First, I am getting really tired of websites that don't work on mobile platforms. How hard is it to pick up a tablet or phone and test your website on it before releasing it to the masses. It's the 21st century more people are likely to browser your website on a mobile device then a desktop so stop only creating websites for desktop browsers. More phones ad tablets don't need the mobile version either, most people do not test full desktop sites on a tablet which is why most of them don't work. My Nexus 10 has more pixels then MOST desktop workstations, fucking deal with it properly and stop formatting a website like I am using a VGA phone.

Second, as mentioned, any web developer worth their salt will not target a specific platform or browser. You want to piss off and make your site unusable by a large percentage of the market, go right ahead, but you are not going to be successful in the long run. It doesn't take that much effort to do cross browser support, often it just means that instead of doing some stupid little web animation or div magic you found an example of somewhere you just have to make your website a little more "standard" and stop with the stupid little tricks that might only work on Chrome or IE or something.

If you get paid to develop a website then get rid of your prejudice and do your damn job properly. Realize that there will always be different browser and platforms and focus on making a great website that works for everybody rather then building a site out of prejudice and ignorance.

Re:Sure

[Murdoch5]

I stopped reading after I read this:

First, I am getting really tired of websites that don't work on mobile platforms. How hard is it to pick up a tablet or phone and test your website on it before releasing it to the masses. It's the 21st century more people are likely to browser your website on a mobile device then a desktop so stop only creating websites for desktop browsers. More phones ad tablets don't need the mobile version either, most people do not test full desktop sites on a tablet which is why most of them don't work. My Nexus 10 has more pixels then MOST desktop workstations, fucking deal with it properly and stop formatting a website like I am using a VGA phone.

I never said I don't develop for mobile platforms, a mobile platform is still a web platform, the fact you didn't realize that is fine but it doesn't mean you can call me out for not targeting mobile devices.

Secondly, I do all development for free, I don't change for the web sites I create because frankly they're not meant for professional installations, if you take what I give you and you leave the scope I place on it then don't cry to me when it doesn't work.

So what I'm sick of is people who assume outside the box abstracts that make developers like me look bad. If you want a website to work on IE then hire a web developer who will build you a site that will work on IE, I don't promise that and I don't even tell you I'll do that. I make websites to fit the need of a need and not for the general public.

Re:Sure

[Anonymous Coward]

So what I'm sick of is people who assume outside the box abstracts that make developers like me look bad.

You're doing a good enough job of that yourself.

Re:Sure

[Synerg1y]

Porting a large scale website to a mobile version can most millions of dollars. I'd say go buy a clue, but I'm afraid you're going to be broke after the port. Also, who browses your website is determined by it's function, not by current trends.

I agree that cross-browser consistency is important, but most business have requirements around that: here's what we support, if you're not on it... contact IT.

Even e-commerce websites monitor their user base and only support the top browsers in relevant versions.

So... you're obviously not a web developer, why are you talking critiquing us?

Do I critique you on how you interact with your mom on a daily basis?

Re:Sure

[Anonymous Coward]

My Nexus 10 has more pixels then MOST desktop workstations, fucking deal with it properly and stop formatting a website like I am using a VGA phone.

You might want to learn how to use your device and get a browser that has the option to "always request the desktop version of websites". Also keep in mind that not all mobile device users browse the web through a WiFi connection. Many of them use mobile data plans with download restrictions ranging from 100MB/month. These people would not appreciate it, if every page they visit requires them to download 10MB total in pictures, thus the use of the stripped down versions.

Why should it matter?

[Anonymous Coward]

Didn't we all stop doing user agent detection a decade ago? Any web page that still relies on UA string is just doing it wrong. If it weren't for browser stats, I'd say all UA strings should be set to blank to enforce proper feature detection.

IE10

[Anonymous Coward]

Just looked at IE10... It shows Mozilla as well. Nothing new here.

Re:IE10

[Merk42]

Yes, IE10 shows "Mozilla", the new thing is that IE11 shows "like Gecko"

Re:Hmm

[Anonymous Coward]

It was working. The update breaks it, therefore the update is broken. That's how people explain it, EVEN if it was broken to begin with.

assinine

[Anonymous Coward]

custom CSS is not hell you fucking whiny douche-nozzle

This User Agent madness needs an ending.

[u64]

During my Privoxy experimental years i tried life without sending any HTTP User Agent header.
It mostly worked fine. Some pages looked wierd. And a few pages refused, and barfed errors;
"Danger! Danger! Server is confused. Evil user for sure."

The only way to end these endless UA detection games is to NOT play.

There will be some rough sailing for some years. But once we've crossed that problem,
the benefits makes it worth it.

transform (the former -ms-tranform)

[Anonymous Coward]

I have a special printing page where the text is rotated 90 degrees. All "good" browsers look like this:
-o-transform(-90deg)
-moz-transform(-90deg)
etc.

And they all work. IE uses an image generated on the server. Well come IE10, MS took away the IE conditionals (!--[IF IE 10]) and I was forced to do something to fix this. I added -ms-transform(-90deg) and found the DIV rotated, but not the text inside the div. Who is right? I am inclined to believe MS is WRONG but I haven't looked at the specs. Adding insult to injury, MS touted that you could now use transform instead of -ms-transform. GRRRR.

It Should Cause A Bunch Of Firings

[StormReaver]

Hopefully, the managers and developers who thought that tying their business processes to a specific version of Internet Explorer was a good idea cause such incredible expenses to their companies that they get fired for being the incompetent idiots that they are.

Re:It Should Cause A Bunch Of Firings

[Anonymous Coward]

Netscape 4.0 was even worse and more buggy than IE 6 back then! Lets look realistically what 2003 was like rather than have our biases to corrupt our memories here? Studies show peoples memories do change based upon opinions today.

People forget back in 2003 there was no Firefox. Even in 2004/2005 Firefox was a hack on Mozilla, which was a hack on the proprietary Netscape code which only owned 7% of the market. IE 6 owned 93% of the market and was going to be around forever so why change? Everyone used IE and in an office environment the I.T. department controls what everyone else runs. It is not like MS would change the standard to follow this hack job that only 1% of users which are geeks used in the office right?!

If you could get a time machine you would see alternatives were not quite there? It was not until 2008 until Firefox finally brought IE under 80%!.Companies making business software take 2 years to adjust so it wasn't until 2010 that they finished their obligations from 2008 in their projects specced from 2006. Now these customers have $$$$ software that wont run on anything else and upgrading either is not an option as it does not raise the shareprice or the vendor is long gone/product discontinued.

What other option was there? 2008 is very recent and business software around that time started requiring IE 7 and sometimes even Firefox. But most software pre-2008 is still requires IE 6 and maybe IE 7 as the requirements were made in 2006 and earlier. How about being pissed at Microsoft for changing its standards then trying to be the good guys and follow W3C? MS is the standard in the office PERIOD. We are in a much better place now and the office is in a hiccup because of it as they can't change like we can in a blink of an eye.
 

Re:It Should Cause A Bunch Of Firings

[Anonymous Coward]

For real! Did you know that the comment slider on these pages does not work in IE9/10. BUT if you change the UA to this new one it does!? Slashdot is full of idiots.

Why not?

[TheSkepticalOptimist]

Firefox has been doing a pretty good job of impersonating bloated IE for the last few years.

100% in favor of breaking old business apps

[rainmayun]

Any app that still depends on an ancient, deprecated version of MSIE should be sunsetted. Developing new, modern replacements for those apps is exactly the kind of capital investment that companies have been putting off since 2002. It is long overdue, and it is the kind of investment that will actually get the economy growing at a much better pace. Good riddance to old cruft.

What goes around ...

[PPH]

... comes around.

Back when Netscape ruled the Web and Microsoft still didn't 'get' the Internet, they (Microsoft) had to impersonate a Netscape identity to keep web sites from rejecting it. Looks like we've come full circle.

MOD UP!

[fritsd]

Hark the word of Andersen the UA string chronicler!
Jay verily, thine blog hath shone a bright light upon this string's murky history.

Finally... for slashdot...

[Anonymous Coward]

Finally, the comment slider on these comment pages actually works when using the new User Agent (as far back as IE 9!). Hey Slashdot, hire someone who knows how to code for the web and stop serving up crap that's not necessary.

MS already lost the battle

[Anonymous Coward]

When big companies stop supporting native IE [1] it means that Microsoft already lost the battle. It's funny how things have changed during 10 years.

[1] For example: http://www.cisco.com/en/US/docs/net_mgmt/prime/infrastructure/1.3/release/notes/cpi_rn_13.html#wp75302

But if IE 10+ is standards complient, what breaks?

[Mike Van Pelt]

If IE 10+ is actually standards complient, what will break if it's "Firefox impersonation" causes web sites to treat it as if were standards complient?

I really don't see the problem here. Those antique versions of IE from the last century will still identify the same way, so the web sites designed to cater to crufty old browsers from the previous century won't treat them any differently than they do now.

The obvious potential problem would be that if Microsoft went back to do their old trick of making up their own standards incompatible with everyone else, then web pages wouldn't work with the new browser. But I think the web environment has changed enough that if they were to do that, the rest of the world wouldn't go along with it. I suspect the folks at Microsoft are aware that such a move would accelerate the switch to Firefox, Chrome, etc.

Problems with linking to a non-shitty browser

[tepples]

Just provide them with a direct link to a non-shitty browser

That would be useful if you can answer these questions: How are you going to do that when you don't even know whether the user is running Windows, Mac OS X, GNU/Linux, or Android because you haven't looked at the user agent? You could put up a form where the user clicks the logo of the operating system, but given the fact that people fall for the "I'm calling from Windows" telemarketing scam, I'd imagine that a lot of inexperienced computer users don't even know which operating system their computer runs. Besides, how do you propose to treat users who aren't the owner of the machine that they're using, such as users of a machine owned by a business, school, library, or head of household? And how do you propose to treat users of iOS or a game console, where even the owner of a device lacks privileges to install software through "a direct link" and the manufacturer refuses to provide browsers using competing rendering engines on its app store?

Also an issue with Safari

[tepples]

It also presumes they can move browser; less of an issue with Safari

I'm not a paid member of the iOS developer program and therefore not privy to the latest version of the iOS App Store Review Guidelines, but an older leaked version states that no rendering engine other than that of Safari is permitted on iPod touch, iPhone, or iPad devices.

Tables have turned?

[aklinux]

Doesn't seem that long ago people were coming up extensions so that Firefox, and other browsers, could masquerade as IE...

Object detection

[tepples]

But the server was set up so that even a completely dumb client with no Javascript support was still supported.

Couldn't you have sent the JavaScript anyway and had it disable itself using object detection if the DOM objects not needed for client-side validation were not present?

Re:Object detection

[Hal_Porter]

We found that code using Date objects would crash and burn horribly if we sent it to IE3 which didn't really support the DOM. So in the end we'd just remove great gobs of Javascript and do almost everything on the server in that case.

That's the only example I can remember but there were other very odd things with IE3 like not handling HTTP PUT or session variables correctly. Luckily the project was very server heavy anyway and I had functions on the server side VBScript to encapsulate variable access and pass them between pages - the deeper you got into the application the more variables you needed to keep track of. So on IE I just swapped the one that didn't work for the one that did - i.e. either HTTP PUT variables became session ones or vice versa depending on which one worked.

Like I say the only reason to use client side Javascript was to save a round trip to the server.

It's actually funny hearing people whine about having to support IE6+ now when you think people hacked together web applications that could run on IE3. In fact most of the stuff we did would probably have worked on Lynx too which has no CSS, no Javascript and no graphics.

Re:Object detection

[tepples]

We found that code using Date objects would crash and burn horribly

Then have your JavaScript run a short test suite for the appropriate Date behavior when the page loads, and turn off client-side pre-validation if the test fails. Or by "crash and burn horribly" did you mean that a page using Date could cause IE to "perform an illegal operation and be shut down"?

Like I say the only reason to use client side Javascript was to save a round trip to the server.

I figured as much.

In fact most of the stuff we did would probably have worked on Lynx too which has no CSS, no Javascript and no graphics.

How would you have distinguished a user of Lynx from a bot scraping your site? Visual and audio CAPTCHAs don't work in Lynx.

Re:Object detection

[Hal_Porter]

Then have your JavaScript run a short test suite for the appropriate Date behavior when the page loads, and turn off client-side pre-validation if the test fails. Or by "crash and burn horribly" did you mean that a page using Date could cause IE to "perform an illegal operation and be shut down"?

IIRC once you had an error in some code that used Date objects IE3 stopped running any other Javascript code in event handlers.

How would you have distinguished a user of Lynx from a bot scraping your site? Visual and audio CAPTCHAs don't work in Lynx.

It was a site for booking hotel rooms. You needed to log in to access it. Also to book a hotel room you needed to enter a valid credit card number.

Re:Object detection

[tepples]

It was a site for booking hotel rooms. You needed to log in to access it. Also to book a hotel room you needed to enter a valid credit card number.

Yet Ticketmaster has been known to use a CAPTCHA so that the operator of a scalping robot can't use the valid number of the robot's operator's credit card to automatically book a whole bunch of seats and then resell the seats to someone else.

Adaptation without script

[tepples]

They should be programatically testing the features they need and avoid what the browser engine is as much as possible.

How do you "programmatically test features" when JavaScript is turned off? For example, you might want to send a less detailed page to users of mobile phones but then test the screen size and retrieve more detailed information if the screen is larger. But when JavaScript is off, your script won't run and thus won't have a chance to retrieve the more detailed information and add it to the DOM.

And how do you programmatically test for cookies? There are some features that you might not want to enable for a bot that doesn't support cookies, such as history of recently viewed product pages on an e-commerce site. If you don't detect the bots through User-agent, you end up with a separate anonymous session in your session table for each page view.

Re:Adaptation without script

[DrXym]

a) Javascript is rarely turned off except for the most security conscious / paranoid, and b) I'm mainly talking of client side where JS and CSS cover most of the use cases, including mobile vs desktop, c) server side covers the rest and without using the user agent.

Arguably, web sites that test against user agents to redirect mobile / desktop are evidence as to why using the user agent is such a fucking stupid thing to do in the first place. It's infuriating when my 10" tablet redirects to some mobile site when I could easily render the desktop version. Many user agents are now tantamount to saying "I'm everything" to persuade broken sites to work. It's dumb and user agents just get bigger and bigger as a result.

The correct behaviour is to run a filter on the server which tests for some layout cookie. If the cookie isn't there, redirect the user through a discovery page which finds out some basics about their browser (testing for screen res, HTML5 features etc.) before issuing a cookie and sending them to the most appropriate layout. The url doesn't change regardless of mobile or desktop. The user could change the preference themselves from that point. And anyone who has disabled cookies or scripts gets the desktop version.

User agent should be an absolute last resort. Sites which are hardcoding this mobile / desktop or iPad / Android shit in are impeding the web as much as those testing for IE 6 / 7 did in their day.

Re:Adaptation without script

[tepples]

a) Javascript is rarely turned off except for the most security conscious / paranoid

Then I must have a myopic view, because "the most security conscious / paranoid" tend to hang out in disproportionate numbers on Slashdot. But JavaScript is also turned off by search engines.

Arguably, web sites that test against user agents to redirect mobile / desktop are evidence as to why using the user agent is such a fucking stupid thing to do in the first place. It's infuriating when my 10" tablet redirects to some mobile site when I could easily render the desktop version.

I can think of one workaround: Include a "Show desktop version" button, and use CSS3 media queries to move it above the fold if the display is at least 32em wide. In fact, use CSS3 media queries to rearrange as much of the layout as you can into an approximation of the desktop layout.

The correct behaviour is to run a filter on the server which tests for some layout cookie. If the cookie isn't there, redirect the user through a discovery page

Googlebot would hit the discovery page every time, creating an anonymous session in your site's sessions table for each page that the bot hits.

The url doesn't change regardless of mobile or desktop.

Sending different content on different URLs in this manner breaks intermediate caching proxies if different kinds of device are behind one proxy.

And anyone who has disabled cookies or scripts gets the desktop version.

How would the browser even get from the discovery page to the page with the work if cookies and scripts are off, such as a bot?

I thought IE had plenty of add-ons

[tepples]

Internet Explorer simply doesn't support extensions that let the user improve their experience.

How not? I thought IE had plenty of add-ons. For example, Google Chrome Frame is an IE add-on, a "browser helper object" to be specific. Or do you complain that the API that IE exposes to add-ons is incomplete in some way?

Browser monopoly? In my smartphone?

[tepples]

As for exposure that is an inevitable result of people being exposed to additional browsers through things like the browser choice menu in the European versions of Window

"Window 8" is an apt name for Windows 8 thanks to the all maximized all the time policy that Microsoft imposes on "immersive" (formerly Metro) applications.

and through smart phones. People are no longer in the position of having a browser monopoly

"Browser monopoly? In my smartphone?" It's more likely than you think. Or since when did Apple start allowing HTML rendering engines other than that of Safari on the iPhone or iPad?

From caniuse.com

[tepples]

What sort of CSS/HTML issues are you having with new versions of IE?

For one thing, the inability to install them on the versions of Windows that came with old, paid-for computers. IE 9 or later excludes Windows XP, and IE 10 or later excludes Windows Vista.

For another, by "new versions of IE", plural, do you mean IE 9 and 10 (new released versions) or IE 10 and 11 (latest release and a leaked pre-release)? If the former, than IE 9 still doesn't support File API, explicit caching of offline portions of web applications, IndexedDB, Page Visibility API, requestAnimationFrame, efficient numeric arrays, CSS3 multiple column layout, form input placeholders, or several HTML5 form input types. Even IE 10 doesn't support WebGL, cross-document messaging between tabs or Windows, or getUserMedia. (Source: caniuse.com)

Well over a thousand dollars hard

[tepples]

How hard is it to pick up a tablet or phone

To test on an Android 2.3 phone, an Android 4 phone, an Android tablet, an iPhone, and an iPad, it's well over a thousand dollars hard, plus the cost of the cellular voice and data plans for the phones. This can become cost prohibitive for a student, hobbyist, or anyone else who doesn't develop a web site as his full-time job.

It's the 21st century more people are likely to browser your website on a mobile device then a desktop

Even if the desktop PC market share is declining, I thought laptops acted more like desktop PCs than like tablets or phones in this respect.

My Nexus 10 has more pixels then MOST desktop workstations

But it still has the same number of physical inches as a 1024x600 pixel netbook. How should the server know whether to send a high or low resolution image if JavaScript is off?