IE6 Addiction Inhibits Windows 7 Migrations

eldavojohn writes "As anyone in the industry will tell you, a lot of money went into developing web applications specific to IE6. And corporations can't leave Windows XP for Windows 7 until IE6 runs (in some way) on Windows 7. Microsoft wants to leave that non-standard browser mess behind them, but as the article notes, 'Organizations running IE6 have told Gartner that 40% of their custom-built browser-dependent applications won't run on IE8, the version packaged with Windows 7. Thus, many companies face a tough decision: Either spend time and money to upgrade those applications so that they work in newer browsers, or stick with Windows XP.' Support for XP is going to end in April 2014. In order to deal with this, companies are looking at virtualizing IE6 only (instead of a full operating system) so that it can run on Windows 7 — even though Microsoft says this violates licensing agreements. IE6 is estimated to have roughly 16% of browser market share, and due to mistakes in the past it may never truly die."

Comments

What do you expect?

[obergfellja]

When people get comfortable enough with something, they don't look for new products to replace it. IE is just another reason why people don't change.

Re:What do you expect?

[mark72005]

If a prepackaged, working solution currently existed to virtualize IE6 and solve all these problems with just the receipt of a licensing fee, this would not be a story.

Similarly, if it were cheap to rewrite all these web applications for IE8, it would also not be a story.

Re:What do you expect?

[Critical Facilities]

It reminds me of when I used to work for a very large, US based Financial Services Provider. They would waste so much money doing things in these roundabout, haphazard ways despite being shown very plainly how planning project progression carefully would save them money and heartburn. Of course, they'd never listen. So, we came up with what we felt best summed up their mission statement:

"There's never enough money to do it right, but there's always enough money to do it again."

Re:What do you expect?

[EraserMouseMan]

We're talking about custom corporate web apps here, folks. It doesn't matter if your web app runs perfectly on Mozilla/Netscape/Opara, etc. If your corporate IT standards mandate that all web-apps must run on the standard corporate-supported browser then that is what you develop the web apps for. Period. In that situation, back in the 2000 time-frame, the market share of IE in most large companies was 100%.

Re:What do you expect?

[Monkeedude1212]

You don't need to retrain someone to use the app - its the exact same application it just runs in a different browser. You don't need to worry about re-certifying and supporting because you'll have to do that with a virtualized system anyways.

I made a post a bit futher up about how you basically save money in the long run by simply paying the dev costs for that upgrade as opposed to a licensing solution. Licensing something to run an obsolete product is a terrible idea, there are very few circumstances where it would even make sense financially.

Re:What do you expect?

[jgagnon]

This is probably moot, but Windows 7 can actually be set up to run nearly (very nearly) as well as XP on the same hardware. There are a few things you'd have to shut off and other tweaks to be made, but it actually does run well on the older hardware. As long as you are running the 32-bit version (highly recommended for cases like yours) then drivers shouldn't be a huge deal.

Having said that, I fully understand why your workplace would not upgrade past XP. We're still on XP here even though Windows 7 32-bit runs everything we use here (with a couple minor, limited-use exceptions). Mostly because we have no profitable reason to leave XP behind. We'd still be running the same applications on 7 that we run on XP, so why upgrade?

Re:What do you expect?

[jgagnon]

Windows XP mode on older hardware sucks royally. Windows 7 can run fine in a 1 GB machine, but try running XP Mode on top of that and see what you get. Now try it again on a 512 MB machine (and a lot of companies still have them - we have at least a dozen 2.4 GHz P4's with 512 MB of RAM).

Re:What do you expect?

[Rob Riggs]

One word describes most of the IT departments in the US: underfunded.

Re:What do you expect?

[Rich0]

These are all big corporations who buy tons of windows licenses.

MS isn't going to sue them because they are running a bunch of IE6-apps on copies of IE6 that they paid for via XP licenses, on copies of Windows 7 that they pay licenses for, supported by Active Directory and Exchange servers that they pay licenses for and client licenses for, etc. Oh, and the reason they're doing it is because MS stopped taking their money for XP extended support contracts and instead they're paying for Windows 7 extended support contracts.

MS would be suicidal to file legal action against companies like this. They're EXACTLY who keeps them in business. No, they're going to do everything they can to make the migration path as smooth as they can. The IT guys at these companies can pick up the phone and have engineers at their beck and call any time they want - they are MS's bread and butter and they know it.

Sure, MS would prefer to leave IE6 behind, and no doubt they'll do what they can to get people to move on. However, the worst they'll actually do is remove official support - they won't be suing their customers.

Re:What do you expect?

[John Hasler]

We are going through the painful process of rewriting and certifying IE6 specific apps and migrating to IE8.

Do you mean that you are going to lock yourself into IE8, or that you are going to rewrite and certify to standards?

Re:What do you expect?

[yuna49]

Everyone wants to blame Microsoft for everything, the real blame is on the standards bodies.

I don't think so at all.

Berners-Lee's original draft for HTML appeared under the auspices of the IETF in 1993. By the end of 1995, HTML was already at version 2.0. The first W3C "recommendation," HTML 3.2, was released in January of 1997.

ASP 1.0 was released in December of 1996. PHP 1.0 had existed since June of 1995, and people were already writing web applications in existing languages like Perl before then. PHP 2.0 was released in November of 1997; I remember writing some PostgreSQL-backed applications in that language soon thereafter. This whole "there wasn't any alternative to Microsoft" perspective on early web application development that I've seen here recently on Slashdot has no basis in reality. If anything, the array of free tools and the pace of their development quickly surpassed anything Microsoft had to offer.

I also don't think you can exempt Microsoft from blame for whatever delays you might think took place in the W3C standards meetings either. After all, Microsoft had, and still has, a seat at the W3C table. At the time they wanted nothing to do with a standards-based Internet. This history of HTML at the W3C site makes for interesting reading about that early period. (Microsoft isn't the only guilty party here, of course. The first HTML "Editorial Review Board," set up when the more open IETF working group failed to reach a consensus, included representatives from IBM, Microsoft, Netscape, Novell, and the W3 Consortium. One outcome of these head-to-head negotiations was Microsoft trading away the MARQUEE tag in return for Netscape giving up BLINK!)

Re:What do you expect?

[sjames]

It's not as if there weren't PLENTY of warnings that sooner or later that would come back to bite them in the ass, they just ignored them all and painted themselves into the corner as fast as they could.

The few apps developed way back then that worked on multiple browsers still work on the latest and greatest.

What forethought by Microsoft

[ByOhTek]

They used IE6 to E^3 (Embrace, Extetnd, Extinguish) Windows 7 long before it even came out!

So sue them.

[Hatta]

In order to deal with this, companies are looking at virtualizing IE6 only (instead of a full operating system) so that it can run on Windows 7 -- even though Microsoft says this violates licensing agreements.

Then Microsoft should sue them. That would teach them, right? After all, violating intellectual property licenses is the same as theft.

Never Upgrade, Never Surrender!

[VoiceInTheDesert]

Just goes to show you that no matter how annoying you can claim Microsoft to be, their user base can be equally so with their instance that decade-old software be their ONLY solution.

You gotta upgrade sometime, people.

Re:Never Upgrade, Never Surrender!

[blincoln]

Companies and governments have massive amounts of custom code which runs only on IE6. The time, money, and effort to rewrite this would be absolutely huge.

Lots of people have massive amounts of 16-bit apps. Should Microsoft have included 16-bit support in x64 versions of Windows as a result? Lots of people still use VB6 apps. Should Microsoft continue to support Visual Studio 6? When I upgraded to Windows 7, I was annoyed that the one 16-bit app I still used at home wouldn't work anymore, but I got over it.

Are you seriously suggesting that organizations just toss out a mission-critical bit of software either because it's old or proprietary?

Organizations need to adjust to a world where they can't just build something that works and continue using it as-is indefinitely. That just isn't practical in an environment that also has internet connectivity of some kind. Long-term supportability must include the ability to continue updating the system over time when e.g. the client OS and/or browser are no longer supported by their vendor(s).

Yes, I realize this is painful, and it's a big shift from the world as it was a few decades ago, but that's the tradeoff when internet connectivity is involved. Hopefully it will also make managers realize that meeting a list of feature/behaviour requirements is not enough. The way an application is built is also critical, because obviously it's possible to build something that meets a feature/behaviour requirements list but that will break as soon as any one piece of infrastructure receives a minor update.

Re:Never Upgrade, Never Surrender!

[Anonymous Coward]

Companies wrote those things because they had an immediate business need to get things done. Over time, they added onto them, and the tools became much more entrenched and something they can't get away from.

Yes. It's called "bad design", or in enterprise terms, "short-term focus". Many of those applications can be linked back to the dot-com bubble, when every kid and their dog was hired instantly to "be part of the hype". And because of the hype maybe, both the people hiring and (most of) the contractors were not expecting the application to last ten years.

But we are now ten years from when the bubble burst. I have zero sympathy for the companies that have eschewed investing in their "business-critical" infrastructure for ten years. It reeks of dodoism.

In the real world, people need solutions now

People that need solutions now have already missed the boat. They make poor businessmen because they lack strategic vision. See, I can do generalized, unfounded and contentless soundbites too ;)

The reality is, when XP and IE6 were corporate standards, that was the toolset you had to work with

Sure. That attitude is fine when you're prototyping, or doing a quick/throw-away/hack job. But when you're designing a "business-critical" application, long-term stability should have been called for. And MS itself had already indicated in 2005 (at the introduction of IE7) that IE7 would only support IE6 as a "compatibility mode" (through META-tags, I believe). That should have been the first clue, even for C?O's, that IE6 was not a suitable toolset (platform) for long-term stability.

XP Mode?

[dilbert627]

Isn't that the point of XP Mode? To run legacy applications that aren't 7-compatible?

Let that be a lesson

[C_Kode]

Let that be a lesson to all those idiots who wrote IE only web applications.

Brilliant!

Re:Let that be a lesson

[Tim C]

Yeah, they must be kicking themselves for having already been paid once, and possibly having been paid again to port their apps over to IE7+...

IE6: bigger mess than Y2K

[CAIMLAS]

IE6 is beginning to be a bigger mess than Y2K. It's not yet such a long-term problem, but the scope is pretty board due to the fact that it's the entire program, not just date fields, which are broken.

Re:IE Patch

[FreonTrip]

It's akin to wading in filth when you haven't had any shots. Paranoia is probably a more appropriate response - on a system with IE6 installed, the only places I visit are Windows Update and Mozilla, to download Firefox. Doing anything else is basically throwing a malware fiesta, and advertising your computer as a buffet..

Huge Success!

[Pentium100]

Well, Microsoft made IE6 not compatible with standards so that people would make sites compatible with IE (because the majority use IE, since it came with Windows) so that the sites would be less compatible with standard browsers that work on other operating systems, so that people would use Windows and IE, since a lot of sites only worked with IE.

Corporate software also requires IE6, since it comes with Windows XP, why make a program that's compatible with other browsers, except IE and then require that browser when all your users have IE6 by default? Now it is inconvenient, but redoing the app to support standards would be expensive.

So, now IE6 is so entrenched in the corporate environment that not only it prevents the company from migrating to Linux or some other OS, but it also prevents the company from migrating to a newer OS made by Microsoft.

Whoever was in charge of the decision to make IE6 non compatible did a wonderful job - XP and IE6 will live for a long time. It will probably even outlive newer versions of Windows.

Re:Huge Success!

[Cederic]

Exactly. IE6 is fantastic, truly magnificent - it's a poster child for any architect.

Why? Because now we have the perfect "here's how to fuck up your organisation by not following standards" example. With the added bonus that almost any organisation I go to work for will have fallen into exactly that trap.

Re:Encapsulating IE6

[mlts]

Very true. However it isn't as easy to get set up and pushed out on an enterprise basis as a single app file. Another downside is that because XP Mode is complete VM that can easily get compromised, it requires an instance of antivirus for corporate IT reasons. Having a single executable that runs in a "jail" is a lot better performance-wise, and means one doesn't have to set up virtualization on company desktops.

Probably the simplest solution for a company that needs IE6 on desktops for one task or application would be to use Citrix or Terminal server, and just keep a well locked down copy of IE6 on a dedicated server.

What do you expect? I expect standards

[Kludge]

If your corporate IT standards mandate ...

That's the point: standards.

Unless your company is developing its own browser and its own OS, making it's own corporate standard on browsers is stupid.
The standards that should have been followed here are the W3C standards. Not the "standards" of one company with one browser on one operating system.

Before 2000 there were computer standards in place. Not following those standards is now an obvious huge failure and now companies will be paying for it.

Re:What do you expect? I expect standards

[gorzek]

IE was (and still is, in some places) a de facto standard. Calling it "stupid" doesn't change that fact.

Back around 2000 the browser wars were very much alive and compliance with W3C standards was largely mythical, to say nothing of fractious JavaScript implementations.

Corporations had to settle on something. Microsoft won out primarily because the browser was a) bundled and b) made by the same company as the operating system. It was just less hassle all around to go with IE at the time.

We can look back now and say it was stupid to standardize on a browser with such a non-standard implementation, but that's because we have the benefit of various standards-compliant browsers now, and the notion that you should be able to view a particular site with any browser you choose has achieved wide penetration. At the time, it was thought one browser would "win" and control the standards for all practical purposes, and most people banked on Microsoft. It was an understandable gamble at the time even though it looks foolish in hindsight.

Re:What do you expect? I expect standards

[gorzek]

Actually, we didn't all know that Microsoft would lose the browser wars. Remember, at one time they had very close to 100% market share in the browser space. Had they maintained such dominance they would have effectively controlled Web standards. It wouldn't matter what the W3C said, it would only matter what Microsoft implemented.

People also assumed Microsoft would maintain backward compatibility as they'd done in large part since the MS-DOS days. It was not clear that MS would eventually abandon their custom HTML implementation in favor of W3C standards--and they only did so because adoption of other browsers forced them to.

Once again, it's unfair to take the situation in hindsight and say everyone who settled on IE6 was stupid. From a corporate standpoint, it was the most attractive of limited options.

Anyone choosing to standardize on a specific browser now, I would call a fool: the implementations are similar (and compliant) enough at this point that if your site doesn't work on all major browsers, you're doing something wrong. We also have much better tools now. The situation today is just a lot different than it was when IE6 came out, and it should be examined accordingly.

Re:What do you expect? I expect standards

[TheRaven64]

Back in the distant past, there used to be the notion of a second source. That is, for every product that you buy - especially ones that your business depends upon - you should have at least two potential suppliers, even if you never actually bought anything from the second one. There are several reasons for this. If the first supplier goes bust, you have a backup. If there is a second supplier, then the first supplier can't raise prices too much or they will suddenly find that they are no longer the first supplier.

Back when IBM made the PC, they insisted on a second source for every single component, with two exceptions. The BIOS, they wrote in house. The operating system, they regarded as a commodity, which therefore didn't need a second source. You'd think that other companies might learn from this mistake.

Part of the economic attraction of open source is that it automatically comes with a second source; any open source product that you buy (by definition) comes with the rights to get someone else to maintain it for you.

If you build your internal infrastructure on top of one company's products and do not have an alternate supplier, then you are saying to that company 'we are willing to pay whatever you decide to charge in the future'. This was known well before IE4 was released, and I was certainly not the only person at the time saying that building intranet sites depending on a particular browser was a stupid idea. I have absolutely no sympathy for companies that decided to save a small amount of money in exchange for a large cost later on.

Re:IE Patch

[somersault]

I live in Scotland.

I'm happy for them to ban smoking anywhere. If it wasn't dangerous to health then of course they shouldn't ban it. But smokers do not, and should not, have any right to force their smoke onto other people.

Car exhausts aren't great either, but at least they don't smell so bad, and usually aren't right there next to you.

Re:This is why "integration" is bad. Hmkay?

[Anonymous Coward]

I'm running IE9 beta on my Windows 7 machine at home.

So? The point was that the development teams doing the patching on these crappy apps will just switch the IE6 non-standard behavior for IE8 non-standard behavior which will break instantly in IE9 ironically which is actually compliant by default for a change.

Prove that the problem isn't due to the IE6 installer (can you even download it (legally) any more?) doesn't expect certain specific versions of Windows and refuse to run if the version string doesn't match.

Ha, you clearly have never investigated how Windows works. Internet Explorer is actually tiny, if you look in Program Files and examine the contents, you'll find there is really next to nothing in there — just a GUI shell and some support crap. The core, the crux of how the whole thing works is in System32, Windows' HTTP fetch, Trident HTML rendering and JavaScript engine are all Internet Explorer components. Replacing those with IE6 versions will break half the system and some third party apps as well.