Original URL: http://www.theregister.co.uk/2007/01/23/microsoft_incompatibilities_then_and_now/

Microsoft: yes, we have no incompatibilities

From DR-DOS to DRM

By Andrew Orlowski

Posted in Operating Systems, 23rd January 2007 12:09 GMT

Analysis Savour this irony.

Last week, we learned that incompatibilities Microsoft hadn't written into its operating system posed a grave threat to users. Last week, we also learned that genuine incompatibilities Microsoft had deliberately written into its operating system posed no threat at all.

In the first instance Microsoft had primed a public relations campaign to warn of the dire consequences posed by these bogus incompatibilities. In the second instance, and times must be tough up there, Microsoft avoided using its public relations professionals to tell us that the genuine incompatibilities were harmless.

Are you still with us? If you're feeling bewildered, your confusion is understandable.

The first instance refers to incompatibilities between Microsoft's Windows 386, 3.x and Windows 95 products and DR-DOS. This was an operating system developed by Digital Research, and later acquired by Novell, which was 100 per cent compatible with Microsoft's MS-DOS. Thanks to citizens in Iowa, who are pursuing a consumer class action lawsuit against Microsoft, these ancient malpractices are being aired once again, and Microsoft executives have been on the stand in Des Moines defending the company's conduct.

Microsoft wanted users to believe that interoperability between DR-DOS and Windows was problematic.

The second instance refers to incompatibilities between Windows Vista and next-generation High Definition DVDs, BluRay and HD-DVD. The incompatibilities are deliberate, and part of the specification Microsoft gave hardware manufacturers so they could design Vista-compatible hardware. Just before Christmas, Peter Gutmann published a technical analysis of the Vista incompatibilities, listed some of the potential security and stability threats they posed, and some of the situations where they might cause real harm.

In this instance, while Microsoft has gone to great lengths to booby-trap its software to disable functionality when certain media discs are being played, and to degrade performance when it finds what it thinks is counterfeit media, or "unauthorized" copying — it wants us to believe this will not have serious consequences for users.

(Until Vista is tested in real world conditions, we won't know for sure if Gutmann's claims are alarmist and Microsoft is telling the truth, or not — or somewhere in between.)

We'll deal with the most topical first.

Gutmann analysed the hardware specifications and declared several problem areas. When "premium content" was being played some functionality was deliberately disabled, specifically video I/O. Vista uses "tilt bits" to detect fluctuations in voltage and severely degrade the operation of the computer. He also said the specification posed problems for programmers developing free software device drivers, and would make the Vista-compatible hardware more expensive than it should be. Finally, Gutmann described catastrophic consequences for users who discovered their driver had been "revoked".

Rather than address questions from Gutmann himself, or from the technical press and analysts, Dave Marsh, Microsoft's lead program manager for video chose his own questions to answer, and passed them along to a colleague, who posted them on his blog.

Naturally these include several answers to questions Gutmann didn't ask, but avoiding the press by selecting awestruck bloggers instead is Microsoft's preferred way of dodging hard questions these days: at CES this month, Gates would only be interviewed by bloggers.

"Gates seems really relaxed unlike in many other interviews I have seen," noted a blogger after watching one of these grillings, conducted by a former Microsoft marketing guy.

Gutmann hadn't asked whether Vista's "content protection requirements apply equally to the Consumer Electronics industry supplied player devices such as an HD-DVD or Blu-Ray player", but Marsh answers anyway. From then on, it's a mixed bag.

Marsh agrees that Vista's DRM taxes the CPU. He dodges the issue of Vista's specs making hardware more expensive by saying that integrating DRM onto the chip in volumes will eventually bring the price down. (That's a "yes", then).

He agrees that S/PDIF, component video and audio are degraded, but says they are already in Windows XP and invoked when requested - and he passes the blame onto Hollywood. He refutes Gutmann's claim that playing back protected content degrades the rest of Vista video output. (Gutmann cited the hypothetical case where medical images would be displayed in lower than optimal resolution when a protected High Definition DVD was being played at the same time - although if your radiographer is watching Porkys III Hi-Def Edition while looking at your scans, we suggest you find a new radiographer).

Marsh confirms that "tilt bits" will cause problems, but he ducks the question of what circumstances will cause tilt bits to be set, and throws the responsibility back on to the hardware vendors. He writes:

"It is pure speculation to say that things like voltage fluctuations might cause a driver to think it is under attack from a hacker. It is up to a graphics IHV to determine what they regard as an attack. Even if such an event did cause playback to stop, the user could just press 'play' again and carry on watching the movie (after the driver has re-initialized, which takes about a second)."

And... then what? Wait for another tilt bit reset, we guess, from speculative causes.

That sure sounds like a fun evening in!

And we throw Marsh's reply to the F/OSS drivers issue open to you. Marsh asks,

"Do things such as HFS (Hardware Functionality Scan) affect the ability of the open-source community to write a driver?" And Marsh answers... "No. HFS uses additional chip characteristics other than those needed to write a driver. HFS requirements should not prevent the disclosure of all the information needed to write drivers." Gutmann, who isn't named in Marsh's ventriloquist routine, isn't impressed.

"Saying 'we were only following orders' has historically proven not to be a very good excuse," he told the BBC News Online website. "If you have got the protection measures there, the impulse is to use the most stringent ones at your disposal."

We'll deal with events taking place in freezing Iowa in a more detail in a follow-up tomorrow, but the basic facts are as follows. The case has been a replay of Caldera vs Microsoft, with evidence brought in from other investigations. Caldera had inherited DR-DOS from Novell. Suit was filed in July 1996, and discovery continued throughout 1998 and 1999, with a serious of unfavourable judgements against Microsoft, one of which expanded the scope of the lawsuit. Microsoft settled just a week before it was due to go to court in January, paying Caldera $275m in damages. We covered the trial in detail at the time (list of links here, juiciest quotes here.)

As with Internet Explorer, and Windows XP, Microsoft had failed to add new features to MS-DOS for several years. Microsoft adopted several tactics to destroy DR-DOS, the most damaging of which was tying PC makers into secret per-processor license agreements, which meant that they paid for Microsoft's MS-DOS whether they shipped it with the PC or not, foreclosing the most important route to market.

But as DR-DOS matured, and Novell developed an alternative retail channel for the product, Microsoft adopted a campaign of disinformation. With the growing popularity of Microsoft's Windows 3, which ran on top of either DOS, Microsoft wanted users to think that performance would degrade if using Windows with Novell's rival product.

The formidable talents of Waggener Edstrom were enlisted. Microsoft's DOS product manager Richard Freedman took the campaign to the press, vowing to "FUD DR DOS with every editorial contact made," and to "develop key DR DOS FUD points for all press tours".

"We’ll basically be covering all the key editors ... We recommend that we ‘informally’ plant the bug of FUD in their ears. ‘Have you heard about problems with DR DOS?’ ‘That security feature is a neat idea and, gosh, such a feature would be great, but it’s just too easily circumvented.’ ‘Gee, it’s unfortunate that DR DOS can’t be loaded high all the time. MS-DOS 5.0 can.’ We’ll do this very tactfully. ‘If Digital Research came to Microsoft for help making DR DOS work with Windows, would Microsoft help them? Maybe not?’"

On Friday the Des Moines court heard this piece of testimony. It's a video from an FTC hearing from 1993, and in the dock was Phil Barrett from Microsoft. It makes for an interesting comparison with the offering from Dave Marsh, above.

Question: Mr. Barrett, you were just asked if you had any knowledge of any Microsoft effort to produce any incompatibility between OS/2 or DR-DOS and Microsoft Windows. How do you define incompatibility within that context? What was your understanding of what you meant by that?

Answer: To prevent the products from working together.

Question: Would you consider an incompatibility something that popped up in, say, a nonfatal error message when there was no error that was being detected by that software?

Answer: No, I would not call that incompatibility.

Question: How would you make the distinction between the two?

Answer: Well, there was nothing done explicitly to prevent Windows from running on that operating system.

Question: Mr. Barrett, you were just asked if you had any knowledge of any Microsoft effort to produce any incompatibility between OS/2 or DR-DOS and Microsoft Windows. How do you define incompatibility within that context? What was your understanding of what you meant by that?

Answer: To prevent the products from working together.

Question: Would you consider an incompatibility something that popped up in, say, a nonfatal error message when there was no error that was being detected by that software?

Answer: No, I would not call that incompatibility.

Question: How would you make the distinction between the two?

Answer: Well, there was nothing done explicitly to prevent Windows from running on that operating system. That's what is meant by incompatibility. It's simply a message. If we played a tune, that wouldn't be an incompatibility. That's what is meant by incompatibility. It's simply a message. If we played a tune, that wouldn't be an incompatibility.

Just fancy that! ®