Microsoft Expert Witness Stumbles

parking_god writes "MIT prof Stuart Madnick, testifying on MS's behalf, was caught out twice when a government attorney asked him to name an OS (other than one made by Microsoft) where the browser couldn't be removed. Madnick also faltered on several other questions." Basically he doesn't understand what GNOME and KDE are, and since we're all holier-than-thou know-it-alls around here, we might as well laugh at Microsoft's expense ;)

Comments

Coincidental Slashdot Fortune

[DLWormwood]

He missed an invaluable opportunity to hold his tongue.
-- Andrew Lang

Talk about perfect timing for a random draw from the fortune file...

don't be too polemic

[stew77]

Yes, this guy obviously doesn't have a clue what an operating system is. However, it's true that any KDE-based distro is in the same situation as Windows is: Sure you can remove the browser, but that will kill certain other programs that need to be replaced as well (e.g. the file browser) and other programs using the browser functionality will also lose freatures (e.g. no more HTML help in your IDE).

Re:don't be too polemic

[Rogerborg]

• it's true that any KDE-based distro is in the same situation as Windows is

It's 100% true... less a few tiny differences.

1. KDE isn't an operating system.
2. KDE isn't a monopoly operating system.
3. KDE isn't an illegaly leveraged monopoly operating system.

So apart from it being completely, utterly different to anyone but an uninformed hermit, it's exactly the same.

Re:don't be too polemic

[sheldon]

This is sort of like arguing Ford wasn't responsible for installing Firestone tires on the Explorer because tires aren't a car.

Yeah, didn't make sense to me either.

Re:don't be too polemic

[walt-sjc]

Um, you seem to be forgetting something. MS WAS tried and found guilty of abusing their monopoly power. This was upheld in the court of appeals. MS broke the law.

We are now in the remedy phase where indeed the government CAN tell MS what to do, just like in any civil case where the guilty party can have all sorts of penalties such as fines, requirments to change contracts, etc., etc. etc.

Our laws are here to protect us from companies that behave like MS, and allow for penalties to prohibit them from continuing illegal behavior.

Your analogy is also like a rapist defending his right to rape. "Why should the government be allowed to tell me what to do?" Well, maybe to protect society from the people breaking the law.

If you don't like it, write your congressmen. Tell him that you don't like having the government penalize people for breaking the law and see how far that gets you...

Huh??

[adam613]

If someone who is a CS prof at MIT doesn't understand what a window manager is, I fear for the future of CS research. I have friends who are English majors and could explain that KDE, Gnome, and XFree86 are all prograams that may or may not be installed on a particular Linux system.

Although I have to wonder what sort of deal did Microsoft offer him to forget the difference between Windoze and KDE? :)

wait a second...

[ACK!!]

Gnome does not necessarily even have a built-in browser for its desktop. Galeon gives you the option of being the default browser but does not have to reside on the same system with the rest of the desktop. Nautilus is the same way. If you still use GMC you have no built-in browser sucking up space.

I thought with KDE you did not HAVE to have Konquerer though it is by default the file manager/browser for KDE. There are other file managers that can be used with KDE that do not have built-in browsers I think.

I understand fully that KDE and GNOME are desktop environments for the Linux OS. Even so, even if the desktop could be considered the OS, his examples still do not apply.

Am I wrong on this or is this guy just the clueless MIT professor ever?

This is not a Troll I would actually like to know if I am wrong.

________________________________________________ __

Re:wait a second...

[tb3]

This guy is clearly out of his depth. Here's his homepage at M.I.T. He seems to be more of a management expert than anything else (kinda like a graduate-level PHB).

However, he is the auther of the classic textbook "Operating Systems". So classic that it was written in 1974, and has been long out of print. What the hell was MS thinking? This guy wouldn't know a GUI if it bit him!

Re:IE is just a shell

[radja]

THE shell? would that be bash, ksh or tcsh?

//rdj

Know-It-Alls

[colmore]

Basically he doesn't understand what GNOME and KDE are, and since we're all holier-than-thou know-it-alls around here, we might as well laugh at Microsoft's expense ;) Well given that this man is supposed to be an "expert witness" *some* knowledge of major competing OSes might be expected. The vast majority of Microsoft's business tactics are legal yet unsavory. I respect that. This is capitolism after all. What bothers me about Microsoft is their monolithic view of their role in computing. The honestly believe that without them, no innovation would have occured between 1985 and now, and so we should just let them walk over consumers and competitors out of gratefulness. I know it won't happen, but what I'd like to see come out of this trial would be a Microsoft not split up, shackled, or fined out of existance, but a Microsoft scared into respecting other's place in the industry. In all honesty they've done a better job than anyone else at creating a useable desktop OS good for a wide range of activities on a large variety of hardware. I'm not quite sure how they've been so successful in the server market, though. Advertising, I guess. And for my money, they still make a damn good mouse.

Re:Know-It-Alls

[royalblue_tom]

You don't just get scared. You are scared because you fear something happening that you don't want to.

What don't they want to happen? Microsoft split up, or shackled.

They would actually like to be fined massively (as a final, no other restrictions remedy) - out of existence is almost impossible given how much cash they have, and without the shackles, they'll just tack it onto the cost of the next version of windows and office.

So if Microsoft know that the situations that they fear are not going to happen, they're not going to be scared, are they ...

Of course the expert doesn't understand the difference between an application and an OS. The concept that there is a difference is alien to the entire Microsoft argument at this point ...

Re:Know-It-Alls

[stinky wizzleteats]

I'm not quite sure how they've been so successful in the server market, though.

The answer lies in your analysis of their success in the desktop OS segment. Here's how it works:

• You are a business with 50,000 users. 99% of those users use some flavor of Windows.
• Microsoft shows up at your door one day, and suggests that you change all your servers to NT. If bribing the CIO into forcing the change down IT's throat doesn't work, and/or this suggestion is resisted...
• Microsoft threatens to do a license audit of all your PCs. You can either:
  1. Find 50,000 license certificates spread among 15 campuses, 10,000 of which are remote laptop users, and 1,000 of those are overseas, all within the two week preparation period Microsoft gives you before the audit
  2. Swallow the blue pill and become a 100% Microsoft shop.

Cisco employs similar tactics, but since they don't have the license audit leverage, they engage in character assassination of IT people who resist Cisco implementations. Isn't capitalism fun?

Madnick is not an MIT computer science professor

[Anonymous Coward]

He is an MIT Sloan School (business school, department 15, management) professor. Many of us from course 6 (EECS) are happy to disavow him.

Generally pathetic witnesses for Microsoft

[burgburgburg]

I've been submitting stories for awhile now (all rejected) on the ineffective witnesses that Microsoft has been using during this phase of the trial.

They've had several industry witnesses who were forced to admit that they'd never read the settlement or the states proposals. The economist who testified for Microsoft had to admit that all of his research in this area had been funded by Microsoft, the Autodesk exec who after defending Microsoft had to relate how screwed over he felt by them excluding Java from Windows XP (needed for some Autodesk software). The most fun was the former Microsoftie, now head of his own company, who testified that the states plan would lead to the "balkanization" of Windows. On cross, he admitted that the Microsoft lawyers wrote the first draft of his testimony, and that he hadn't even know what balkanization meant.

How much are these Microsoft lawyers getting if this is the level of their trial prep?

Re:Generally pathetic witnesses for Microsoft

[drew_kime]

From news.com.com:

Jerry Sanders, chief executive of computer chipmaker Advanced Micro Devices, also conceded he had not read the states' proposed sanctions ...

"You've never checked to this day whether what Mr. Gates told you...was true in the remedies," Gutman challenged. Sanders agreed he had not read the states' proposals.

From the Register:

And in written testimony to the court, Sanders quotes from AMD's annual report: "If we fail to retain the support and certifications of Microsoft, our ability to market our processors could be materially adversely affected." (Remember, this is a witness for Microsoft.)

Back to news.com.com:

Gates' appearance next week would be his first in-person appearance at the trial. In the main portion of the trial, Gates appeared in a videotaped deposition. In portions of that videotape, Gates repeatedly answered questions with "I don't know" and "I don't recall." His statements were frequently contradicted by e-mails he had sent and received, and he frequently claimed no recollection of the messages.

Even Business Week, in a generally flattering review of Gates' testimony, leads with:

Is Bill Gates Doing Himself Any Favors?
He's poised and confident on the stand, a far cry from his deposition during the antitrust trial. But maybe he shouldn't be there at all

Good old news.com.com again:

During cross-examination Wednesday, states' attorney Steven Kuney brought up the issue of Windows XP Embedded, a version of Windows made for gas pumps and other machines that contains the core elements of Windows but doesn't necessarily contain browsers or messaging software, depending on how it is configured.

Kuney asked Gates if Windows XP Embedded could be installed on PCs. Gates responded, "You could configure it for that."

But Gates said he didn't know of anyone who had done such a thing, later acknowledging that one reason is because Microsoft doesn't license XP Embedded for that purpose.

Back to the Register:

One of the exhibits in the previous stages of the Microsoft antitrust trial included an email from one Chris Jones, recommending to Bill Gates that the binding of IE into Windows should be such that users would find running rival browsers "a jolting experience." At the time many people, not least of them the Department of Justice, seemed to think that this and other associated exhibits were all about the anticompetitive tying of IE into Windows in order to destroy Netscape. But apparently not - MS Windows exec Chris, taking the stand yesterday, put forward an explanation of almost patentable novelty.

What he meant, he said, was that the experience would be jolting for good reasons if it occurred because of the "great innovations" that integration of IE brought to Windows. So presumably you could think about the new versions of IE Microsoft was designing as being truly wondrous, and that users would therefore find use of the comparatively stone age rival products truly unpleasant.
...
Another interesting point was brought up by States' attorney Kevin Hodges, who established that the proposed MS-DoJ settlement had less teeth to it than appears at first glance. Under this deal PC manufacturers will have the right to install rival companies' software, but it's still feasible for Microsoft to bar them from running Netscape when the computer is first turned on. Jones seems to have argued that as IE was a part of Windows, Microsoft didn't have to give OEMs the right to run Netscape. (So much for Microsoft allowing competition on the desktop.)

Now from Wired:

Several companies, as well as the nine states, argue that Microsoft adopted open technology standards only to make them proprietary later, forcing many to use Microsoft products. Sutherland said he did not study any records of Microsoft's conduct.

"You did not take into consideration Microsoft's past conduct in these proceedings?" Schmidtlein asked.

"Only as background," Sutherland said, adding that he didn't find it relevant.

And again from news.com.com:

But under questioning from the states' lawyer, Sutherland acknowledged that he knew little about Microsoft's past anti-competitive conduct and had no experience with the kind of Web-based services at issue in the case.
...
Sutherland said any company that wants to compete in the telecommunications business must make its technologies work seamlessly with other companies' services.
...
Under questioning from states' attorney John Schmidtlein, however, Sutherland conceded he had no direct experience with Web-based messaging and was only part of a small group at Qwest that is studying the possibility of getting into the business of Web-based messaging.

He also admitted the group was formed less than a month ago--nearly two months after Microsoft named him as a witness in the antitrust case.

"My intention is to offer the court an understanding of how the communications world works," Sutherland told the judge. "My testimony is not specific to Microsoft's behavior on the Windows desktop."

As someone said, if this represents the level and quality of Microsoft's legal team's trial prep, you have to wonder how much they're getting paid.

Re:Generally pathetic witnesses for Microsoft

[sg3000]

Okay, look, just because Microsoft's witnesses have proved that they didn't read the settlement, got favors from Microsoft to testify, or generally can't tell between an operating system and an old shoe doesn't mean that Microsoft is doing poorly in the case.

I mean, you're acting like illegally abusing your monopoly to shut out competition and reduce consumer choice with hopes of cornering the market on desktop operating systems, office productivity software, media, and the Internet is a crime or something. What you're failing to take into account is I got Solitare for free. That's right, FOR FREE! Is that the sign of a heartless monopoly? No, of course not. And do I see you offering me free, cheapo quality shovelware? No, all I get from you are a bunch of "facts" and "news". Well, you can keep your "facts", Mister-- I've got a game of Klondike I must win!

Simpsons

[hotsauce]

*Bully from Simpsons voice* HA-ha!

The most popular prof...

[Merk]

So lemme get this straight -- this guy is a CS prof at MIT, home of the FSF. He voluntarily agreed to testify on Microsoft's behalf, and then didn't know the difference between an operating system and a desktop environment?

Man, this guy's courses must be popular! I bet you really have to fight to get in to: "Introduction to flicking on the power switch thingy 101" and "How to click on the start menu 304"

Re:The most popular prof...

[billnapier]

DOS is an OS? Really? Maybe just barely...

The Master was explaining the nature of Tao to one of his novices.

"The Tao is embodied in all software -- regardless of how insignificant,"
said the Master.

"Is the Tao in a hand-held calculator?" asked the novice.

"It is," came the reply.

"Is the Tao in a video game?" asked the novice.

"It is even in a video game," said the Master.

"Is the Tao in the DOS for a personal computer?" asked the novice.

The Master coughed and shifted his position slightly. "The lesson is over
for today," he said.

Re:The most popular prof...

[oever]

Konqueror has many functionalities, all modularized in socalled parts.
If you want to remove the browser from KDE, you can remove the KHTML part. The rest of the functionality will remain intact.

Re:The most popular prof...

[ocelotbob]

More importantly, you can change the browser and have no loss of functionality. KDE has well-defined hooks so that one can use whatever browser they choose - you can download kmozilla and have konqueror and the rest of KDE use the Gecko engine to render HTML instead of using KHTML

Re:The most popular prof...

[Odinson]

Mouth agape...

Then how did Microsoft postulate that he was a qualified expert on the removability factors of IE? Wasn't technical feasability a more primary point of contention than fiscal feasabilty? If not why not?

I realize Project Manager has the word manager in it but it has ALOT more to do with CS than traditional management. The wide availibity of more qualified persons at MIT must speak volumes about his relationship with MSFT.

Business press doesn't see it that way

[sphealey]

I have been following the trial in the general business press as well as the IT trade press and of course the Linux-centric sites. Although the IT trade press is reporting that Microsoft's witnesses are doing a mixed job and are taking some significant hits in cross-examination, the general business press is taking the line that Microsoft's legal team has everything under control this time and is crushing the States.

My guess is that the judge's viewpoint is going to be closer to the general business press than the IT world (much less Slashdot), so I am not holding out much hope for a meaningful order here.

sPh

Re: Business press doesn't see it that way

[Black Parrot]

> the general business press is taking the line that Microsoft's legal team has everything under control this time and is crushing the States.

Let's not forget that the business press exists for the sole purpose of keeping stock prices high. It's hard to imagine that they would say anything different no matter what was going on.

But of course, they have the DoJ's desire to throw the game to give them confidence that they're going to be right this time anyway.

Re:IE is just a shell

[BusterB]

Is Internet Explorer any less a part of Windows than the shell is a part of Unix? Where exactly do you draw the line? Discuss.

Does Unix require one type of shell over another? You could write init scripts that used csh, ksh, bash, tcsh, or something else entirely. You could use python interactively, or make emacs the default shell. There is no requirement of one over another fundamentally.

He is not part of (EE)CS

[Anonymous Coward]

He is affiliated with the Sloan School of Management (Course 15), and not the EECS department (Course 6). Hence, the lack of knowledge about the OS itself. He's probably trying to get some more funding from Microsoft for the i-Campus initiative Here's his personal home page, FYI.

Re:He is not part of (EE)CS

[phasic]

Trying to associate him with only sloan does no good. Look at his qualifications:

Dr. Madnick has degrees in Electrical Engineering (B.S. and M.S.), Management (M.S.), and Computer Science (Ph.D.) from MIT. He has been a Visiting Professor at Harvard University, Nanyang Technological University (Singapore), University of Newcastle (England), and Technion (Israel).

C'mon, a doctorate in comp sci from MIT, with just one management degree compared to the 3 EE/comp sci degrees. He must know something about the subject, if not to the specific degree slashdot would like, but maybe we're not getting the whole story.

Re:He is not part of (EE)CS

[Fastball]

Perhaps his migration to the dark side of management has tainted the Jedi mind tricks he learned from the comp sci department.

Control, you must learn control!

[CptLogic]

>>Madnick testified that Microsoft (MSFT: Research, Estimates) probably would not be able to develop and market a workable version of Windows under the terms proposed by the dissenting states. He believes the requirements -- such as building Windows in such a way that computer manufacturers could alter it -- are not technically feasible.

And he's right, it's not technically feasible because Microsoft will not relinquish control of the necessary source, preferring to keep everything black boxed, the hell away from people who could alter the product that carries their name. It's about controlling how you can use the product that they are associated with, because, "hell, a third party could screw Windows up and Microsoft could get a bad rep."

We know Microsoft are control freaks, there's no way they'd allow Windows to be opened up like that, and without that unlocking of the black box, it *is* not technically feasible for a computer manufacturer to alter Windows, and the reason for this is "technically" MSFT are not legally bound to release their source, and "technically" could charge for any SDK they may choose to never release that would allow that access.

"technically" this poor bastard who's been set up to fail, trying to defend the indefensible, is correct. In an "I did not have sexual relations..." kind of way.

Chris.

The

how hard could it be to remove the brower, anyway?

[cheesyfru]

This has been bugging me forever. Nobody is saying that they need to remove the browser from the OS, they just need to disable it. How hard is it to remove the icons for it, and disable the "internet http browser" aspect until the user voluntarily downloads a tiny piece of plug-in code which enables the browser to work with internet protocols? If the world's largest and most powerful software company can't figure out how to do this, then how in the world are they getting big business to pay them millions of dollars to manage their mission critical software?

Microsoft maybe not as funny as you think

[rdelsambuco]

Microsoft has lost little in this whole process by "not holding its tongue." Say what you will about MS but the organization as a whole has done a good job of finding spokespeople (like the professor) who appear "respectable" -- and that's all they need.

The spin in Seattle on public radio was entirely positive onn this -- which was interesting.

Re:Microsoft maybe not as funny as you think

[fiber_halo]

the organization as a whole has done a good job of finding spokespeople (like the professor) who appear "respectable"

Yes, it never ceases to amaze me how many "respectable" people can be swayed by that payment for being a professional witness.

It would take an unimaginable sum of money for me to sell out and lose the respect of my peers.

Re:Microsoft maybe not as funny as you think

[FatRatBastard]

Not only that, but removing respective Browsers from KDE and Gnome doesn't kill either Windowing environment. He was wrong on so many accounts.

Re:Microsoft maybe not as funny as you think

[JabberWokky]

More than that, KHTML, while perfectly fine, can be replaced - you can run KMozilla instead, and the system works fine. View/View Mode lists all the HTML viewer plugins on your system in Konqueror (if you're viewing HTML), and KMozilla is pretty darn common.

Remove *all* the html viewers, and you still can do everything else with KDE - except view help and browse the web (both of which require something to view HTML). Or you can put in your own third party HTML viewer, and distribute KDE (after all, KDE dosen't even release binaries, the distros package them themselves). MS won't only not allow people to do that, they say it's impossible to do.

--
Evan

Re:Microsoft maybe not as funny as you think

[Thurn und Taxis]

the organization as a whole has done a good job of finding spokespeople (like the professor) who appear "respectable"

The particularly amusing thing about this statement is that this "computer and software expert" (from the article) is not a professor in the computer science department, but a professor in the business school. I guess Ron Rivest wasn't for sale. :-)

Re:Microsoft maybe not as funny as you think

[Malcontent]

"It would take an unimaginable sum of money for me to sell out and lose the respect of my peers."

Well lucky for you MS has unimaginable sums of money.

Product Interference Not So Uncommon

[Root Down]

Microsoft deliberately designs its products to interfere with technology made by other companies, forcing people to use Microsoft products...

This is not an uncommon practice. Don't believe me? Try installing Real Player (Real One) and watch the default installation - that which the majority of users would use - take over every media file in your system. This is directly interfering with the use of other media - now requiring extra steps to use anything but the default. Try unassociating - no obvious route exists. This is just one example.

Counterpoint: You are still able to use these alternative media, even though there is a "performance cost" involved in having to take extra steps. Don't like it? Don't be an idiot and use the default install.

Both are worth considering in the overall sense of programming specifically to exclude the competition and its prevalence in the computer industry - especially given the foreknowledge that the majority of your users will not consider themselves 'advanced' enough to select options in the non-default setup. It's another question of ethics that really has not been given a great deal of attention - though we've likely got more pressing issues to consider (e.g.: DMCA, etc).

Try This...

[ScumBiker]

If Internet Explorer is so tightly integrated into Windows, how come you can upgrade it? I just upgraded the browser on my NT workstation here at the office from Internet Explorer 5.5 to Internet Explorer 6.0. Does that mean I also upgraded my operating system? Do I get better performance reading large files? Can I crunch data faster? Is there better communication between my hard drive controller and my memory sub-system? Microsoft is SO full of shit.

Well this guy was an MIT prof

[rblancarte]

But that is not the point. Fact is that they destroyed his credibility with these questions. Excellent work. I have a bit more faith that the DOJ actually has some brains about going after M$ this time.

But really - what does this mean? The layperson out and about won't hear about this. They are not informed about this and regardless of what happenes, short of M$ being broken up - people won't stop using their products. I like all of this news, but the masses just don't hear about these things (they don't read slashodot).

RonB

Re:Well this guy was an MIT prof

[dcgaber]

The DoJ has absolutely nothing to do with this phase of the trial, in fact they oppose it. They have a seperate remedy proposal ::cough:: sellout ::cough:: that they are trying to get the judge to aprove. It is far weaker remedies than what the non-settling states are proposing, and most all credible commentators believe it will do absolutely nothing.

The one to have faith in is the non-settling state AGs who are still pursuing the case. The DoJ has determined they don't care, they got enough of MS $$ to satisfy their needs.

Re:Well this guy was an MIT prof

[kelzer]

Windows and Microsoft shouldn't go out of its way to make it more difficult for non-MS apps to work well.

They don't, that's just anti-MS FUD.

Bullshit. You all can mod me down as flamebait, but I just can't let you get away with this lie. All one has to do is go back to the Caldera lawsuit, and read the transcripts. There was email after email between high-ranking MS execs coming up with a strategy to sabotage DR-DOS's perceived interoperability with Windows. They even went so far as to encrypt the code that displayed the "error" to make it harder to see that they were doing.

It ain't FUD if it's a matter of public record.

shells

[mikeee]

You mean there are shells other than emacs??

Re:shells

[egreB]

Nono. Well, emacs maybe a shell, Vi(m) is an whole beach.

Haha. Dunno if that worked in english, though.

Re:shells

[ruin]

Nono. Well, emacs maybe a shell, Vi(m) is an whole beach.

I think you misspelled "bitch."

Or possibly beast.

This takes the cake ....

[ProfMoriarty]

Madnick argued that perfect interoperability, which would allow products to be substituted for each other with no performance degradation, was a theoretical impossibility. "It would be surprising if two different products behaved exactly alike," he told the court Wednesday.

Ok then ... so what about the examples that you gave earlier ...

But KDE is a computer program designed to run on top of the Linux operating system, as Hodges pointed out. Madnick conceded that was true, and instead suggested GNOME as an example. But GNOME performs the same function as KDE on a computer equipped with the Linux operating system.

This is VERY funny ... on one hand, it's "theoretical impossibility" to have TWO INDEPENDENT systems that can "be substituted for each other with no performance degradation" ...

Yet he uses the PERFECT example of doing such ... KDE and GNOME.

This stuff is so funny, it writes itself ...

On a bit of a serious note, IS there any performance degradation between KDE and GNOME?

Re:Madnick is not an MIT computer science professo

[Rocky]

Yeah, but all three degrees of his are from Course 6, so someone there's to blame...

Of course, they're probably all dead now...

Madnick is *not* a CS prof

[Anonymous Coward]

Madnick is a has-been professor in Course 15, the Sloan business school, and he has nothing at all to do with Course 6-3, the computer science department at MIT. He teaches "MIS" style courses to accountants and economists.

He is most famous for co-authoring the book mostly called "Madnick and Donovan" which was some sort of IBM 360 OS bible back in the way-back days of punch cards.

BTW, it is might be interesting to note that Richard Schmalensee was the MIT professor who humiliated himself on the stand in the first phase of the trial, and he is also a professor of management in the same school at MIT. It's really not a bad school, they only look bad when they whore themselves for Microsoft money

STUART E. MADNICK is *NOT* a CS Professor

[Brian_Ellenberger]

I thought this was wierd, so I did some checking on this guy. I looked for him on MIT Electrical Engineering and Computer Science faculty list, but couldn't find him. So I looked him up in the people directory and found this:

name: Madnick, Stuart E email: smadnick@MIT.EDU phone: (617) 253-6671 address: E53-321 department: School Of Mgmt title: J N Maguire Prof Of Info Tech url: http://mit.edu/smadnick/www/home.html

His department is not EECS, it is the School of Management! His research is in areas such as Total Data Quality Management and Productivity From Information Technology. Here is a bio description from his web page:

http://mit.edu/smadnick/www/home.html Madnick finds ways to integrate information systems, giving organizations a more global view of their operations. He is leading a project that develops new technologies for gathering and analyzing information from many different sources, including conventional databases and the World Wide Web. He is also testing these new technologies in industries such as financial services, manufacturing, logistics, and transportation.

Microsoft basically found anyone from MIT they could because it is MIT. I'm surprised they didn't find a janitor from MIT to testify.

Brian Ellenberger

Re:STUART E. MADNICK is *NOT* a CS Professor

[Havokmon]

He joined the staff in 1972. Therefore, his degrees are from the 60's.

My mom was an RPG programmer in the 80's on 390's and PICK OS...

I suppose if you didn't know her, you'd think she could comment on the current state of programming and operating systems. I'll give you a hint - She hasn't kept up.

well, isn't he right?

[ceswiedler]

Konqi serves the same purpose in KDE/Linux as Internet/Windows Explorer (same thing these days) serves in Windows. It does file management, web browsing, help, and html email rendering. Both do all of this through a component architecture.

What would KDE be without Konq? Same thing as Windows. Not really usable the way it was intended.

The article ribs the witness for calling KDE an operating system. Well, no, KDE is a user interface / window manager / shell sitting on top of the Linux (or other) kernel. Same as Explorer, which is a user interface / window manager / shell on top of the Windows NT kernel (in NT/2000/XP anyway). Perhaps he should have said KDE/Linux, but do we really want to go there?

Re:well, isn't he right?

[klaun]

The article ribs the witness for calling KDE an operating system. Well, no, KDE is a user interface / window manager / shell sitting on top of the Linux (or other) kernel. Same as Explorer, which is a user interface / window manager / shell on top of the Windows NT kernel (in NT/2000/XP anyway). Perhaps he should have said KDE/Linux, but do we really want to go there?

Well, I'll go right out and buy a copy of just the Windows NT kernel with no MS window manager and install that other window manager for the Windows/NT OS from ... um ... Oh I guess there isn't another window manager for MS OS, and come to think of it Microsoft doesn't offer their OS without a Window Manager. Come to think of it they claim a Window Manager is an integral part of an OS. Strange that... guess all those systems with no console don't run an OS.

Re:well, isn't he right?

[ceswiedler]

Tell me, how do I replace the Explorer shell on an NT kernel box?

Well, isn't that Microsoft's point? The GUI shell (which includes IE) is tied to the OS. They are the same thing as far as the user is concerned. If you told the typical Windows user "you can run Windows, but with this other shell which makes it look and act nothing like Windows" they would stare at you blankly. "That's not Windows."

What is an OS? You guys are limiting the definition to the kernel. Explorer is not tied inextricably to the NT kernel. Explorer IS tied inextricably to Windows.

Great, Linux is better because I can replace the GUI shell. Linux is also better because I can replace the memory manager (via recompiling). Can Windows do that? No. Is it fair for the DOJ to force Microsoft to allow users to replace the MM?

Well, I guess that's the question, isn't it?

finger madnick@mit.edu :: NOT AN MIT CS PROF!

[CaptainAbstraction]

This has been said several times, but must be said
again and again. Madnick is not a computer
science professor at MIT!!!! I find this
frustrating, especially having graduated from MIT
in CS. I'm so sad that this guy is spoiling the
reputation of the MIT CS department.

He teaches management!!!!!!!!!!!!!!!!

[686 parkerlocal@waikiki Documentation]$ finger madnick@mit.edu
[mit.edu]
...

There was 1 match to your request.

name: Madnick, Stuart E
email: smadnick@MIT.EDU
phone: (617) 253-6671
address: E53-321
department: School Of Mgmt
title: J N Maguire Prof Of Info Tech
url: http://mit.edu/smadnick/www/home.html
alias: S-madnick

Re:how hard could it be to remove the brower, anyw

[Xzzy]

Apparently, it can be done with a 100k zip file, for free.

Re:IE is just a shell

[opkool]

Have you ever heard of Win98Lite?

It is a program that erradicates IExplorer from Windows. Windows is completely usable. You have to get another web browser (Netscape, Mozilla...) and a file amnager (the old on from Win95... or use Midnight Commander ontop a Cygwin bash shell).

I used back then, to get a P1666 no-mmx with 32MB of RAM to run Windows 98. Before Win98Lite, it was slow as.... as WinXP. After using Win98Lite, it was usable again, moderately fast... a cool college machine. Until SuSE 5.3 got its way into it, of course *grin*

More information can be found here

Feh

[Greyfox]

I would think that all you'd have to do to replace the browser would be to code your replacement to export the same APIs that IE does. If those APIs were documented, replacing the browser would be simple -- the Mozilla people probably would have already coded to those APIs if they were available. The desktop should just call an API to perform browsing functions. It should not care who wrote the program providing the API.

Seems to me all you'd have to do is force MS to publically document the API. Actually they should be forced to document APIs, file formats and protocols BEFORE their products are released, and they should be compelled to use only protocols and formats unencumbered by patents or copyrights (for things like XML DTDs.) The documentation should be unencumbered by any license and should be freely available on their web site for all to download.

He's NOT a CS professor

[Pachooka-san]

If you check Stuart Madnick's homepage, he's not a CS professor. He's a professor of management. Need I say more?

Re:IE is just a shell

[fanatic]

The shell in Unix underlies the Unix permissions system.

You haven't the slightest idea what you are talking about. Not even a little.

Permissions are applied to files by the filesystem code, typically a kernel module.

The shell is a program that runs on top of the OS, interprets user commands and runs programs (as well as providing a programming language, in many cases). It is totally separable from the OS.

Maybe you meant capabilities, but again, this would typically be imposed by kernel code, with the shell utilizing the information provided. In no way could the shell be said to underly any of this.

"House of cards"

[Keith Russell]

How's this for a ringing endorsement of Microsoft's products?

[Madnick] showed the judge a diagram that depicted Windows as a system made up of dozens of oddly shaped, interconnected pieces. Madnick said the diagram showed how Windows was like a "house of cards" that could collapse if any of the pieces were removed.

Emphasis mine. Source: ZDNet: Microsoft's MIT prof gets grilled by states

Mind you, this was a witness for Microsoft. Amazing. Microsoft is so arrogant, it can claim gross incompetence to avoid incrimination, and still look forward to getting away with it.

Apples and Oranges

[JPriest]

Windows is a GUI system with emulated DOS. Linux is a kernel, shell, X, then a windowing system. There are multiple layers involved where you are free to build on one as you please. Explorer is a larger more integrated part of what is known as windows. I, like BeOS really don't see the problem with MS making explorer part of windows because that's what works for performance. I have my gripes about the way MS does some things but this is not one of them. To me this is kind of like going after the mafia for tax evasion, if it's the only thing that holds water go with it.

Re:But what *is* an operating system?

[i0lanthe]

In this sense, Linux/KDE and Linux/GNOME both constitute OSes on which the web browser cannot be removed.

Except that someone else has already argued above that, in the case of KDE, it can. And probably by the time I post someone will say the same for Gnome.

I'm afraid I can't comment on which of you folks is actually right... since I use Linux/twm (fear my l33t "operating system" :-).

Re:Unbelievable

[cjpez]

Well, I don't think that Windoze is really dependant on the application Internet Explorer; more the libraries that come bundled with it. I've been forced to spend a lot of time in Windoze 2000 lately (need to access some PCAnywhere machines, unfortunately - I should see if there's an opensouce PCAnywhere client, but I doubt it), and I've noticed that IE's rendering stuff is everywhere. When you open up "help," when you're browsing around on your computer . . . The libraries that IE uses to grab webpages and show them on the screen have been re-used to hell and back again, which realy is an acceptable thing to do.

I don't know why Microsoft keeps on claiming that the application itself is nonremoveable. Just delete the IE binary; of course it's removeable. What they should do is have some kind of "Internet Services Pack" or whatever which is a basic, nonremoveable part of Windoze, and then just have IE be the shell that accesses those components. There, problem solved. I'm guessing it'd just be a matter of repackaging some things.

I'm guessing that MS is still claiming that IE the application can't be removed just because they want to keep everyone using it by default. Keepin' the resellers down and all. Or hell, I don't care if IE keeps on getting shipped with Windows, just let the poor OEM people install Mozilla by default! Anyway, yeah.

I could be totally wrong about all that, but that's how it seems to me.

Re:MS withdrawing witnesses

[dipfan]

There's a very good reason why Microsoft has withdrawn its witnesses: the states aren't allowed to introduce "new" evidence, and so were hoping to raise some 14 rather embarassing documents from Microsoft with MS employees. So MS withdrew the witnesses, and now the states can't raise the documents (it seems). Great tactic, it has made the states' lawyers look foolish.

There has been a couple of news items about this, there's one from the FT here - it says:

Lawyers for the nine litigating US states in the Microsoft antitrust remedy hearings yesterday appeared to have been comprehensively out-manoeuvered by their counterparts defending the software giant, after the Microsoft legal team decided to halve the number of defence witnesses it would call.

In particular, Microsoft's decision not to call Richard Fade, its executive in charge of relations with computer manufacturers, means the states' lawyers will probably not be able to enter critical evidence before the court.

This is the latest blow to the states' case. Earlier in the hearings, their lawyers misunderstood rules about how new witnesses should be called, leaving them without key testimony.

Great news, huh?

Re:Unbelievable

[Lonath]

That's so right. The really sad part is that I think the states would be happy if MS would just let the OEMs remove the IE shortcut from the desktop when they set up their customers' computers. It isn't even about removing IE, it's about not having IE staring you in the face and preventing any other browser from appearing anywhere. If MS wants to use IE for internal stuff so that it pops up when used automatically, who cares? Just let Dell and Gateway put Netscape on the desktop and remove IE from the desktop if they want. All of this bitching is over default icons on the desktop. :P

Re:IE is just a shell

[GreyPoopon]

Perhaps I'm taking you a bit too literally, but what is stopping people from installing and using another browser?

The issue is not the ability to install and use another browser. It's a tad more complicated than that.

1. Microsoft argues that they can't unbundle the browser, and it must be included with a Windows installation. This, of course, puts all potential competitors at a loss because a user is not likely to be motivated enough to try a competing product when they already have what seems to be a perfectly good tool.
2. Many people say that they should just remove the IE icons. That doesn't work either because too many other tools and applications on the system tie directly into the IE application. It will still come up anyway.
3. Footprint is the big one for me. If you don't unbundle the IE browser so that it can be replaced with another one, you're drastically increasing the memory and hard disk footprint needed for browsing. With today's monster hard drives, the storage space isn't all that much of an issue, but those blasted libraries that get loaded at bootup are already hogging memory. When I run Mozilla or Opera, they have to compete with IE because it is already loaded and taking up memory.

The truly surprising thing

[SwiftOne]

I'm not surprised that an MS witness didn't have perfect understanding of a different software paradigm.

I am surprised (pleasantly) that the lawyer recognized and was able to deal with the situation. I mean, sure, I have little doubt that the lawyers have been briefed, but this lawyer:

• Had to know that GNOME and KDE existed
• Had to know what they were, generally
• Had to understand that the answer was wrong
• Had to be able to articulate that the answer was wrong, with enough accuracy/confidence to have a witness with Comp Sci experience admit his error

I haven't been wowwed by this trial (I think MS has stiffled the industry, and I think the charges have focused on the wrong elements of MS behavior), but I am pleased to see that the legal staff has assumed an apparently comfortable amount of non-MS technical familiarity. This is a rare bit of good foreshadowing for future technical cases.

So what?

[The Silver Slurper]

I really can't believe that we're still arguing about browser bundling in Windows. This point may have been significant 5 years ago, but the battle has been over for awhile.

KDE and GNOME may not be "operating systems" in the strictest sense of the term, but for the end-user they form the most critical and recognizable part of the operating system: the desktop. What part of Linux is the OS anyway? Is it the underlying kernel that provides support for your hardware and devices? Is it the set of GNU system tools and utilities that you use to maintain your system? Is it the window manager and desktop shell?

Linux was designed to be more modular than Windows, but this additional freedom and flexibility come at a price. What parts of a bundled Linux distribution can be removed or replaced by other work-alike components? Almost everything, but when modern applications come to depend on the existence of other "operating system" components, the complexity of setting up a system can increase exponentially. The operating system itself, however, is not useful in the general sense; it is only necessary. For a computer to be useful, you need applications.

Microsoft has chosen the route of providing a consistent base of OS and applications which are always installed and, in some cases, cannot be easily removed. Consider this the lowest-common denominator approach that bundles every basic tool that the average computer user may need. This includes (in Windows XP): video and audio player/editor (Media Player, Sound Recorder, Movie Maker), basic text editor (Notepad and Wordpad), e-mail (Outlook Express), web browser (Internet Explorer), file manager (Explorer), image/photo viewer/editor (Picture Viewer and Paintbrush), and communications software(Hyperterminal and MSN Instant Messenger) among other things.

Out of all of these commonly bundled applications (after all what desktop OS distribution doesn't include one of these applications in some form or another), the web browser has assumed a unique and important role in the modern computing environment. It has transcended its role as a mere user application and has become a vital system component that other applications have come to rely on. Will your operating system work without a web browser? Yes but, as I stated earlier, the operating system *doesn't matter*.

People use computers to get work done. Work is done by using applications. Applications rely on the operating system to provide basic system services. HTML and HTTP have become basic system services for a large number of applications to provide online help systems, downloadable updates and enhancements, and even application user interface. Because a web browser is included as part of the operating system, Windows application vendors can rely on its existence to provide features to their own applications. Is this not, after all, the entire purpose of the operating system?

The states and the DOJ can force Microsoft's hand and make them remove Internet Explorer from the operating system, but does this really make any sense? Users have always had the ability to use another browser when they surf the web, but an integrated HTML rendering engine and HTTP protocol implementation that it guaranteed to be bundled with the OS makes so much damn sense I really, truly don't understand what all the fuss is about.

Did you see his Microsoft projects?

[Kris Warkentin]

Co-Director, PRoductivity From Information Technology (PROFIT) Program: http://mitsloan.mit.edu/research/profit/index.html

Co-Principal Inbestigator, COntext INtercharge (COIN) project: http://context.mit.edu/~coin/

PROFIT and COIN - yep. Must be a Microsoft shill.

standards

[mach-5]

"Madnick argued that perfect interoperability, which would allow products to be substituted for each other with no performance degradation, was a theoretical impossibility."

No, it is not a theoretical impossibility, it is called a "standard", and there are hundreds of groups out there working hard to create these.

Missed in the hubbub

[rgmoore]

It seems to me that one point that's been missed in the hubbub about whether KDE and GNOME are desktop environments or part of the operating system is that the witness was wrong about the web browsers' removability. It's quite possible to pull remove the web browser from either KDE or GNOME. If I decide that Konqueror is taking up valuable space that should be saved for Mozilla, I can just rpm -e kdeaddons-konqueror and it's gone. Similarly I can remove galeon with rpm -e galeon. I'll lose some functionality by doing so, true, but neither one is so deeply entwined into the system that it's unremovable.

The Dead Cow

[QuantumG]

You can read the whole transcript on the Microsoft web site:

Q. You mentioned in paragraph 20 TCP/IP. Could you tell us what is TCP/IP?
A. Well, the initials stand for transmission control protocol slash Internet protocol, and these are the two primary protocols used in the Internet for computers to communicate with each other.
Q. Is TCP/IP something that is part of the operating system or part of the Web browser?
A. In... I guess I would say part of the operating system in the sense as this section has illustrated, the functionality of operating systems have constantly increased over the past decades, and I believe almost every operating system, commercial operating system, I know of today provides TCP/IP whether or not, because -- if I can -- there are many other functions, such as FTP and others, that rely upon IP in order to do their job.
So there are many other functions besides browsing that operating systems rely upon these things, so therefore it would have to be part of the operating system.
Q. As part of the operating system in Windows 95, is that your testimony?
A. It was added, as I mentioned, over time. I don't -- I believe it was added into Windows 95. I forgot exactly which version it was added into.
Q. And in the current version of windows today, it's part of the operating system and not part of the Web browser. Is that your testimony?
A. As I said, as in many other -- most other commercial operating systems, I believe it is part of the key functions of the operating system.
Q. Let's turn if we could to paragraph 22 of your testimony, which is at page 11. Professor, at paragraph 22 you mention IBM's OS/2 Warp 3 operating system. Do you see that.
A. Yes, I do.
Q. And you say that IBM's OS/2 Warp operating system included Web browsing software. Do you see that?
A. Yes, I do.
Q. Was the OS/2 Web browser removable without impairing the functionality of the IBM operating system?
A. I do not know that. I did not study that aspect. My point in this section was to illustrate that these functionalities are included in operating systems in various ways.
Q. Since you don't know about OS/2, is there any other operating system you're aware of in which the Web browsing functionality is commingled with the operating system?
A. Yes, I do, if we take the view that the Web browsing functionality is also relied upon in other parts of the operating system.
Q. Which operating systems would those be?
A. Well, some examples, and there may be many others, would be the KDE user interface or GUI that exists on the Linux operating system.
Q. Now, KDE is not an operating system; correct?
A. I think I -- every definition in this court it would be middleware, in which case it would be a platform software.
Q. KDE is the graphic user interface, graphical user interface, for the Linux operating system; is that correct?
A. Yes. It's one of the interfaces available.
Q. It can be removed and replaced; correct?
A. Well, it can be -- if it is removed, of course, by -- if it's just removed, then the user will not be able to use the system. You could replace it by others and, in fact, most of the others I'm aware of likewise have, as you would call it, commingled Web browsing with their functionality.
Q. In Windows can you remove the graphical user interface?
THE COURT: Are you talking about now?
MR. HODGES: Today, correct.
A. As I understand -- I believe it's either yes or will soon be. I believe the provision that the Microsoft has agreed to as part of the settlement is that the end user would be able to remove access to the browser, if that was your question.
Q. My question is: Can the graphical user interface of Windows be removed?
A. I'm sorry. No, I do not believe so. It would no longer be Windows.
Q. Has it ever been the case that the graphical user interface of Windows could be removed?
A. I guess the answer might be yes in the sense, as I said again in this session, at one time operating systems had no graphical interface at all if you go back to essentially the original MS-DOS. So this is the examples of the kinds of functionality that operating systems have increasingly provided to users to enhance their effectiveness. So, yes, there was a point in time where it did not exist and there's a point in time where it was added to the operating system.
Q. If KDE is removed from the Linux operating system, then its Web browsing functionality is also removed; is that correct?
A. Well, the Web browsing that's provided through the interface is removed, yes.
Q. The Web browsing provided through KDE; correct?
A. That is correct.
Q. Now, you say that, in paragraph 24 -- it's actually on page 12, paragraph 24. I'll read this to you. "One cannot delete the Web browser from KDE without losing the ability to manage files on the user's own hard disk." Do you see that language?
A. Yes, I do.
Q. Now, isn't it the case that files can be managed by using standard UNIX command in the shell even if KDE is not installed?
A. That is correct. The assumption here was we are talking about the user using the system as a modern operating system which requires access to this kind of interface.
Q. We've talked about Windows and we've talked about the KDE interface, and my question is: Can you name any operating system, other than Windows, that commingles a Web browser with the operating system?
A. I have not attempted to identify all the others. As I indicate in this whole section, these are examples of the kinds of innovative features that vendors constantly add to the systems. Some have reached that stage of benefiting from the kinds of interactions possible, some have not. These are the ones I've identified as part of the study so far.
Q. Based on your experience as a computer scientist and as a Professor at the Massachusetts Institute of Technology, are you aware of any operating system, other than Windows, that binds the Web browser into the operating system? MR. LACOVARA: I'll object. We have now shifted from commingling to binding without a definition. It may have just been inadvertent on Mr. Hodges' part.
MR. HODGES: It was inadvertent, and I appreciate that clarification.
Q. If I change the word from "binding" to "commingling," let me ask you, are you aware of any operating system, other than Windows, that commingles a Web browser with the operating system?
MR. LACOVARA: I would object to that. I think it's the third time he's asked the question. Asked and answered. THE COURT: I'll let him to proceed. But this is the last time.
A. Okay. If I recall the question, I think I answered it in terms of identifying KDE and I believe GNOME, which is another interface on Linux, also has the Web browser functionality integrated. So those are two examples. And, once again, this was not an attempt to exhaustively study all the others or systems that are under development today.
Q. GNOME is a -- it's spelled G-N-O-M-E; correct?
A. That's correct, yes.
Q. It's not the way most people would pronounce that word. GNOME is also a graphical user interface for Linux; correct.
A. That is correct. It provides that kind of functionality.
Q. And it is also, like KDE, a removable graphical user interface for Linux; correct?
A. It's removable in the sense if you remove it you no longer have access to a graphical user interface.
Q. It's not an operating system; correct?
A. Well, it is part of what we described as middleware under the understanding of the terms being used, and we go from there.
Q. I'll try to stay in order, but I need to flip back to page 11 and paragraph 23 if I could. You say in the second sentence --
A. I'm sorry. What page?
Q. I'm sorry. It's page 11, paragraph 23. I can tell you, Professor, it also appears up there on the monitor in front of you, so whatever is easier for you is fine.
THE COURT: The small monitor has it, too.
THE WITNESS: It's sometimes helpful to see the context. That's why I like to look at the documents.
BY MR. HODGES:
Q. The second line of paragraph 23 -- the second sentence, I'm sorry -- you say that Windows, like all commercial operating systems of which I am aware, ships with a simple text editor, Notepad in the case of Windows, that is a relatively self-contained block of code that is easily removable. What's the basis for that statement?
A. It's a long sentence. Is there some particular part of it you're having a question about?
Q. Yes. I want to know what's the basis for your statement that the Notepad is easily removable?
A. The fact that there is a file -- I can't remember it's name, but it's probably something like Notepad.exe -- that in theory one could delete without having any other effects upon the operating system.
Q. Is this based on your review of the Windows XP source code?
A. Not specifically.
Q. Professor, have you had an opportunity to review the direct testimony of Robert Short of Microsoft?
A. I have seen it.
Q. Mr. Short is the vice president of Windows core
technologies. Does that sound right to you?
A. Yes, it does.
Q. Are you aware that Mr. Short testified that there are
cross-dependencies between the Notepad and Internet Explorer?
A. After I wrote my report, I believe I remembered hearing
that mentioned in his report or his testimony. Yes the answer
is.
Q. Do you disagree with Mr. Short?
A. I assume he knows much more about the internals of Windows
than I do. I believe my point may still be true, although I've
not consulted with him, in that I believe the removable of
Notepad does not impact any other part of the system.
I believe in his testimony -- I think he was trying to
illustrate that other parts -- using my earlier diagram of HTML
Renderer, for example, or Shell Doc Viewer -- that removal of
other parts of the middleware that might seem to be unrelated
might cause Notepad to fail.
Am I clear on the duality here or the differences? Am
I clear on the differences that removing Notepad may not cause
other parts of the system to fail, but that removing other
parts of the system that may appear to be file removed from
Notepad might cause Notepad to fail. I think that is two
different issues.
Q. Are you aware that Mr. Short used the term
cross-dependencies?
A. I don't recall what exact term he used.
Q. If there are cross-dependencies, doesn't that apply that
Notepad relies on Internet Explorer and Internet Explorer
relies on Notepad?
A. I can't speak for him.
Q. Is that what the term cross-dependencies means to you?
A. That would be a one interpretation, yes.
Q. And if there are cross-dependencies, wouldn't it be the
case that removing Notepad would affect other parts of the
Windows operating system product?
A. That might be true. I was only trying to give a simple
example here. If that one doesn't apply I'll have to find some
other example.
Q. I take it you were not aware of any cross-dependency involving the Notepad?
A. No, I was not.
Q. Is there any technical reason that there needs to be a cross-dependency between the Notepad and Internet Explorer?
A. As I said, this is not an area that I have studied. If you would like me to speculate or to try to conjure up a reason, I could try to do so, but it would be totally ad hoc thinking.
Q. I don't want you to speculate. I want to ask if you are aware of any technical reason that there needs to be a cross-dependency between the Notepad and Internet Explorer? And if you don't know, that's acceptable.
A. What might be helpful is the realization based upon my many years trying to understand all of the inter-dependencies that go on in a complex product is extremely difficult, and often I've been quite surprised myself to realize that one part of the system was able to make use of another part.
So, you know, with some careful thought it is possible I might find that there actually is a reason for cross- dependencies. But it was not something that immediately came to mind.
Q. So you could speculate that, but you don't know. Is that an accurate summary?
A. As I've said, I have not studied that issue.
Q. Now, you have reviewed the Windows XP source code; correct?
A. Yes, I have. Though I will not say I've looked at every 36 million or so lines of code carefully.
Q. Is it 36 million or 39 million?
A. As I said, a million here, a million there, it adds up after a while I guess.
Q. Pretty soon you're talking about real lines of code?
A. Exactly.
Q. What exactly have you done?
A. The main purpose of looking -- once again, given both the limited amount of time and the size, was really to better understand the way in which a system was modularized, the way in which it's broken up into individual routines and the types of interdepencies that exist, so it's more to get a feel for the overall structure of the system.
Q. Is there any way you can quantify how much effort was involved in your review of the Windows XP source code?
A. Physical amount of time, probably 8 or 10 hours.

Re:The Dead Cow

[sheldon]

Interesting reading...

Clearly this witness was a lot more intelligent and knowledgeable than the slashbot responses suggest.

With regards to some of the final questioning there on the cross dependencies between Notepad and IE, I think the most obvious thing is if you were to delete Notepad from the system... View-Source would no longer function within IE. So yes, Mr. Short from Microsoft was technically correct.

Is there a technical need for the cross dependency? Well assuming we are talking specifically about View->Source. Then yes, in so much as the browser needs to have some way to display the source. The method must be well known, and exist at the time IE is installed to the system, or installed in conjunction with IE. Notepad is an obvious choice. Otherwise, the IE team would have to recreate this functionality within the program.

Can it be done? Yes. Is it a technical desirable solution? No... application modularity is very desirable and makes development more efficient instead of constantly recreating the wheel.

Focusing on the wrong issue

[g_bit]

Why do OEM's care so much about altering the desktop? It's M$'s product, the OEM's shouldn't be allowed to mess with it.

The fact that you can't be a licensed Windows PC Provider AND sell naked PC's or PC's with Linux or ANY OTHER OS on them when you sign the contract with Microsoft is the issue they should be looking at.

If I told you that you could sell PC's with Mandrake on them but if you signed up to do so were then legally inable to sell naked PC's or PC's with Windows on them you'd be pissed too.

Re:IE is just a shell

[walt-sjc]

Damn. Wish there was a "clueless" mod option. The shell has virtually NOTHING to do with permissions. It hardly even touches them. The kernel enforces permissions and other applications (such as chmod and ls) set and view them using the functions provided by the kernel. The shell is mainly just a user interface. In fact, when you "exec" a command from the shell, the shell is GONE and replaced by the application.

What you don't comprehend is that UNIX is modular. This allows us to replace virtually every single command/component in the system with a different one. It also is what allows us to fairly easily support multiple differnt file systems of which linux supports over a dozen (maybe almost 2 dozen), different terminal programs, window managers, desktop environments (kde/gnome/etc) and web browsers.

What MS is trying to do (and having a hard time with) is actually having a stand-alone component be integrated. If MS wants to be able to upgrade IE (and it history proves that it does), it MUST be modular. The very fact that IE is upgradable totally blows MS's argument that IE cannot be separated.

Now there is the argument that you would lose the shell. This is bull. MS already has a IE-free shell that was available in 95 and NT4. What they did since then was add hooks in the shell to call IE at various places.

The bottom line is that there is no reason that Mozilla / netscape / opera couldn't use the same APIs that windows uses to "integrate" with IE other than the fact that MS keeps those APIs secret.

A simple request

[Our Man In Redmond]

Would everyone who wishes to point out that Stuart Madnick is a business professor, and not a computer science professor, please check in at the desk, take a number, and wait in line over there along the far wall?

Thank you.

Re:yeah, and...

[markmoss]

shouldn't that headline read: "Witness for MS caught dissembling"?

Only by pretending to be an expert when the only OS he knows is Windows...

Re:IE is just a shell

[tshak]

Microsoft argues that they can't unbundle the browser, and it must be included with a Windows installation. This, of course, puts all potential competitors at a loss because a user is not likely to be motivated enough to try a competing product when they already have what seems to be a perfectly good tool.


What about all the competition with the calculator, Telnet clients, PPP clients (Who buys Trumpet Winsock?) etc. The Internet is now one of the most common uses of a computer. Of course MS is goign to bundle or even integrate a browser into their OS since that's what their customers want. Does a consumer even "know" what a PPP client is? Should these be "unbundled" from the OS so that there can be more competition? The competition is in the OS, and what the OS bundles. Maybe IE should be able to be removed (at least superficially, like the win98lite program does) by OEM's. Although, I've yet to see how this benefits the customer. They want something that works, not with 100 privacy settings.

Footprint is the big one for me. If you don't unbundle the IE browser so that it can be replaced with another one, you're drastically increasing the memory and hard disk footprint needed for browsing.

Huh? Opera takes an additional 4MB of space (on top of IE which is not that big either). While I use Opera, IE is not "browsing at the same time" and therefore is not "increasing the memory footprint" for browsing. Your browsing experience with Opera is not affected at all by the existence of another browser on your system. I fail to see any technical merit in your point.

Re:of course....

[_Sprocket_]

Nah ... Remedy phase designed to help Microsoft's competitors. Let's not sugar-coat the truth.

By definition, anything that hurts Microsoft will help their competitors. How does hurting Microsoft help the consumer?

The consumer is supposed to be served by a free market. A free market requires competitors. Yet, Microsoft has been rather skillfull in dismantling that free market - abusing their position to remove competition.

The remedy phase is designed to help consumers by restoring competition. Let's not obscure the truth with anti-market Microsoft cheering.

Link to the ORIGINAL Article...

[Lethyos]

This link was provided by someone who replied to my post. For those of you who haven't notice, CNN pulled the original article and replaced it with a more Microsoft-friendly one. Total bullshit. I am sure MS offered them a chunk of cash to keep this on the down-low.

Go grab it here: http://sage.che.pitt.edu/~harrold/tmp/73B9A1D4d01. html