A brief look at IBM's counterclaims [LWN.net]

Your editor has had the opportunity to look at IBM's latest filing in the SCO case. As is the norm in this case, it seems, IBM pulls no punches. IBM's counterclaims are strong, and will certainly cause a degree of discomfort in SCO land.

The first 16 pages of IBM's filing consist of a paragraph by paragraph response to SCO's complaints. IBM denies almost everything that SCO claims, and states flat out that SCO is not entitled to any sort of relief. At the end of this section, IBM asks for a favorable judgement and repayment of attorney's fees.

The real fun begins with the counterclaims. In the intro IBM states:

SCO has misused, and is misusing, its purported rights to the Unix operating system developed originally by Bell Laboratories, then a research and development arm of AT&T Corp., to threaten destruction of the competing operating systems known as AIX and Linux, and to extract windfall profits for its unjust enrichment.

What follows is several pages describing IBM's AIX and Linux business, and a section on the GPL. IBM notes that SCO continues to make Linux code available via its site, and also that SCO, by modifying and distributing Linux, has agreed to the terms of the GPL.

Then there is a section on the failure of SCO's business; IBM points out that SCO has never had a profitable quarter until it began its attack on IBM and Linux. "With apparently no other prospects, SCO shifted its business model to litigation."

There is a lengthy description of SCO's suit and threats against IBM, and, in particular, the alleged revocation of IBM's license for AIX. There is an interesting claim: it is said that, when Novell sold its various rights to Unix, it retained a sort of veto power over license revocation. Novell apparently exercised that power in June, telling SCO to cease its attempts to revoke IBM's license. Of course, IBM denies that SCO ever had the ability to revoke this license, but it makes a big point of the fact that SCO has continued to talk publicly about the revocation when it knows that it is unable to yank IBM's ability to use Unix.

There is a section on SCO's refusal to back up its claims.

Moreover, key developers and influence leaders in the open-source community, including leaders of Linux kernel development, have stated publicly that they are prepared immediately to remove any allegedly offending material from the Linux kernel. Rather than permit remediation or mitigation of its alleged injuries (which are non-existent), SCO has declined to reveal the particulars of the alleged violations in order to artificially and improperly inflate the price of its stock.

IBM devotes several paragraphs to the damage that have been done by SCO's attacks. Then the specific counterclaims are laid out:

Breach of contract, by trying to deprive IBM of its ability to use its Unix license.
Lanham act violation, in the form of saying false things about IBM's products and services.
Unfair competition through false ownership claims and attempts to deprive IBM of the value of its Linux products and services.
Interference with prospective economic relations.
Unfair and deceptive trade practices, again through false statements.
Violation of the GPL, by claiming ownership rights over Linux; trying to collect license fees; copying, modifying, and distributing Linux ("including IBM contributions") on terms other than those set out in the GPL; and by seeking to impose additional restrictions on recipients of GPL code. IBM states that SCO's rights with regard to the Linux code under the GPL have been terminated.
Infringement of patent 4,814,746 ("data compression method").
Infringement of patent 4,821,211 ("method of navigating among program menus using a graphical menu tree").
Infringement of patent 4,953,209 ("self-verifying receipt and acceptance system for electronically delivered data objects").
Infringement of patent 5,805,785 ("method for monitoring and recovery of subsystems in a distributed/clustered system").

IBM finished up by asking for injunctive relief and damages. No amounts for damages are proposed.

to post comments

Happy, but uneasy

Posted Aug 7, 2003 18:57 UTC (Thu) by rjamestaylor (guest, #339) [Link] (3 responses)

I'm happy to see IBM respond with vigor, but am a little worried about the introduction of patents into the mix. As Bruce Perens recently warned, the patent portfolios of the friends of Linux pose a latent threat to Linux developers and distributors as well.

Happy, but uneasy

Posted Aug 7, 2003 23:26 UTC (Thu) by JoeBuck (subscriber, #2330) [Link]

IBM has explicitly licensed some of its patents for use in GPL code. For example, GCC is now able to use register allocation techniques that were formerly blocked by IBM patents.

I would prefer to see IBM take this much further and put it on a formal basis (especially if GNU/Linux distributors could be accused of violating the same patents that they are charging SCO with violating).

Happy, but uneasy

Posted Aug 7, 2003 23:57 UTC (Thu) by dkite (guest, #4577) [Link] (1 responses)

Unfortunately it's the way it works. And the idea of having an idea, patenting it
and living rich the rest of your life is so ingrained in the north american
culture, I don't see it changing any time soon.

So, IBM and others, collect patents for the most arcane ideas. Over 2000 a
year. Sue IBM for any reason and they will put you out of business. That is
why IBM doesn't get sued very often.

This is why most patent suits are now initiated by companies who don't make
anything, essentially lawyer firms. They can't be countersued since they don't
do anything real.

Derek

Happy, but uneasy

Posted Aug 17, 2003 3:57 UTC (Sun) by md_maloney (guest, #9194) [Link]

I have a feeling that the patent points are more of a "Oh yeah? I'll teach you to sue IBM!! Take that!!" sort of thing. These points would never had been brought had SCO behaved.

--Mike

A brief look at IBM's counterclaims

Posted Aug 7, 2003 18:59 UTC (Thu) by sbergman27 (guest, #10767) [Link] (1 responses)

IBM's and RedHat's filings are certainly interesting. But here is a question I have not seen addressed:

Are we talking about something that could see quick legal action or is all this stuff that won't be heard in court until sometime in 2005 and may be resolved by sometime in 2007?

It makes a difference.

Anyone have any insights?

A brief look at IBM's counterclaims

Posted Aug 7, 2003 19:16 UTC (Thu) by daniel (guest, #3181) [Link]

"Are we talking about something that could see quick legal action or is all this stuff that won't be heard in court until sometime in 2005 and may be resolved by sometime in 2007? It makes a difference. Anyone have any insights?"

IBM will no doubt go for the jugular, with an application for a preliminary injunction.

The patents

Posted Aug 7, 2003 19:13 UTC (Thu) by jbh (guest, #494) [Link] (8 responses)

These are the patents and what sense I could make of them from a quick glance. They look kind of obvious to me except maybe the first one.

4,814,746 (Lempel-Ziv compression with LRU pruning of dictionary)

4,821,211 (Menu-tree thingy, couldn't see at a glance what makes it special)

4,953,209 (Click-through-license for executable programs where the program isn't executable before the license is accepted)

5,805,785 (Recovery of failed cluster members where the individual members are not themselves failsafe or -aware, using heartbeat and membership protocols)

The patents

Posted Aug 7, 2003 20:01 UTC (Thu) by rise (guest, #5045) [Link]

Looks like IBM decided to cherry-pick four from all over their portfolio (SCO has got to be violating hundreds) that are largely unrelated. This means SCO will have to fight four almost completely unrelated patent battles, and the LZ compression claim is probably so widespread in their code that they've got no hope of winning it.

The patents

Posted Aug 7, 2003 20:08 UTC (Thu) by sbergman27 (guest, #10767) [Link] (5 responses)

So why these particular patents? From their vast portfolio, this is the best they could do?

I'm rather familiar with OpenServer. Not so familiar with Unixware.

I don't really see where lempel-ziv is absolutely crucial to Open Server. It uses the older unix compress/uncompress.

The only place I can think of that the "Tree Thingy" would come into play is in the administration menus and while that *can* be done graphically it also has a text based mode which works about as badly^Wwell.

Can't think of anything that the click thru stuff would affect.

Nonstop clusters would be affected by the 4th patent, and that does not apply to Open Server.

i.e. Open Server (a significant revenue stream even if it is 1992 technology) seems largely unaffected.

Seems like they ought to have something more like:

Patent #31415926: A method by which 2 or more users can use a computer system simultaneously.

Not that that would be good news.

The patents

Posted Aug 7, 2003 22:30 UTC (Thu) by roelofs (guest, #2599) [Link]

I don't really see where lempel-ziv is absolutely crucial to Open Server. It uses the older unix compress/uncompress.

You have that backwards. compress(1) uses LZW, which is claim 7 of IBM's patent. Since Unisys also patented LZW and IBM apparently was content to let them reap the royalties for more than a decade, the fact that IBM's patent is a continuation of an application with a priority date three weeks earlier than Unisys's application apparently isn't relevant--i.e., it appears that claim 7 is unenforceable, or at least questionable enough that IBM has never chosen to press the issue. But there are a lot of other claims in there, and it's conceivable that SCO really is touching some of those. See the comp.compression FAQ for details. (Note that none of this has ever been tested in court, TTBOMK, so treat the FAQ as expert opinion rather than case law.)

Greg

The patents

Posted Aug 8, 2003 5:11 UTC (Fri) by mbp (subscriber, #2737) [Link] (1 responses)

> So why these particular patents? From their vast portfolio,
> this is the best they could do?

IBM probably have good grounds to suspect that SCO violated those patents. But it doesn't really matter. The technique is fairly similar to stoning to death: IBM can start a practically unbounded number of patent lawsuits if they wish, from their portfolio of thousands. Some of them will hit, but even if they all missed, SCO could not afford to defend them.

As Ian Lance Taylor said <http://www.linuxjournal.com/article.php?sid=6956>

> IBM is a past master of the IP extortion strategy. For example, see
> this Forbes article about IBM's shakedown of Sun in Sun's early days.
> For SCO to attack IBM using IP is somewhat like trying to eat a live
> tiger.

http://www.forbes.com/asap/2002/0624/044_print.html

IBM are fighting to preserve a multi-billion dollar Linux and AIX business. The minor expenditure of bringing four, or forty, or four hundred other lawsuits against SCO is well worth while.

I am not really happy that the legal system works this way, but in this case it's for the good.

The patents

Posted Aug 8, 2003 5:28 UTC (Fri) by jonabbey (guest, #2736) [Link]

I am not really happy that the legal system works this way, but in this case it's for the good.

Yeah, it is. It's a deal. We open source developers give corporations so damn much value that they can't bear to think of losing it, and we get them on our side when it comes to the minority interest whose ax is gored by UNIX on Intel being a zero-cost item.

Rah, rah economic self-interest, what?

The patents

Posted Aug 8, 2003 5:39 UTC (Fri) by error27 (guest, #8346) [Link]

So why these particular patents? From their vast portfolio, this is the best they could do?

Obviously that's the first thing anyone familiar with software patents will think. Why four and not hundreds?

It seems that these four were chosen because they cover SCO's entire product line.

I think another reason was these were chosen was to send a message. "Don't sue IBM if you work with computers because IBM owns the patent on everything including the start menu. Literally."

To programmers, the four patents probably seem stupid. Frankly the whole idea of software patents seems stupid and I personally believe that software patents are unconstitutional based on the first amendment. However, SCO has to take the patents seriously. Unfortunately for them if they can prove prior art for all four patents, IBM will simply find four more.

The patents

Posted Aug 8, 2003 15:15 UTC (Fri) by ksmathers (guest, #2353) [Link]

So why these particular patents? From their vast portfolio, this is the best they could do?

I think they were chosen for tactical reasons. Each patent touches upon exactly one SCO product, chosen to represent the most valuable and proprietary assets SCO owns. This is a strike to SCO's heart -- a little like targeting a MIRV at Berlin, Paris, Madrid, and London. I don't think it will cow SCO since they were obviously already betting their company on this, but it may make some investors rethink their strategies. Or then again, maybe not.

The patents

Posted Aug 12, 2003 9:50 UTC (Tue) by NAR (subscriber, #1313) [Link]

Am I right that patent #4,821,211 covers (for example) the Start Menu editing functionality in Windows '95 and later or the KDE Control Center? It's really scary that someone tries to enforce a patent like this, even if it is done for the "right cause"... I've just read an article that Microsoft has to pay about 520 million USD to a Eolas and UCLA, because they use one of Eolas's patents in Internet Explorer. I wonder when will the big companies wake up to the fact, that software patents are bad - even for them - and lobby to abolish this stupid practice.
Bye,NAR

A brief look at IBM's counterclaims

Posted Aug 7, 2003 19:26 UTC (Thu) by jdthood (guest, #4157) [Link] (3 responses)

IBM and SCO never cross-licensed their patent portfolios? Remarkable.

A brief look at IBM's counterclaims

Posted Aug 7, 2003 19:34 UTC (Thu) by Dabuk (subscriber, #1507) [Link] (2 responses)

I don't think SCO has many (any?) patents. They were never an especially innovative company. Not that that's required to get a patent these days.

Three patents possibly held by The SCO Group

Posted Aug 7, 2003 20:08 UTC (Thu) by StevenCole (guest, #3068) [Link] (1 responses)

These three are the only I've found so far:

6,529,784 Method and apparatus for monitoring computer systems and alerting users of actual or potential system errors.

6,362,836 Universal application server for providing applications on a variety of client devices in a client/server network.

6,104,392 Method of displaying an application on a variety of client devices in a client/server network.

The first is assigned to Caldera Systems, Inc. and the last two are assigned to The Santa Cruz Operation, Inc.

Three patents possibly held by The SCO Group

Posted Aug 7, 2003 21:00 UTC (Thu) by sbergman27 (guest, #10767) [Link]

Hmmm, seems like the last two would (should?) probably have been transferred to Tarantella, Inc.

Come to think of it. Didn't the original SCO change it's name to Tarantella while the operating systems and services divisions of the company were transferred to Caldera (now The SCO Group)?

So anything not expressly transferred would still reside with Tarantella, Inc?

A brief look at IBM's counterclaims

Posted Aug 7, 2003 19:54 UTC (Thu) by leandro (guest, #1460) [Link] (6 responses)

Now the US$1M question is: when is Linus going to cease-and-desist SCO from distributing Linux as per their GNU GPL violations?

Or rather, when will the FSF, as seemingly IBM has assigned them their Linux work copyrights?

A brief look at IBM's counterclaims

Posted Aug 7, 2003 21:50 UTC (Thu) by coriordan (guest, #7544) [Link] (4 responses)

> as seemingly IBM has assigned [FSF] their Linux work copyrights

I heard that IBM assigned it's copyrights for the s390 port to FSF, but a quick grep of linux-2.6.0-test2 says that the copyrights for most files in there are held by IBM and the remainder are held by Linus Torvalds.

FSF do have copyrights to one header file in the asm-s390 directory, and one s390 documentation file though.

Can someone clear this up?

A brief look at IBM's counterclaims

Posted Aug 8, 2003 5:00 UTC (Fri) by mbp (subscriber, #2737) [Link] (3 responses)

The header file to which you refer was borrowed from glibc, and therefore (C) the FSF.

I don't know why anyone would assign copyrights to the FSF for kernel work, since the FSF is not particularly involved with the Linux kernel. I suppose they wouldn't say no if you wanted to...

IBM has contributed code both under corporate and personal copyrights. (Try grepping for ibm.com addresses.)

A brief look at IBM's counterclaims

Posted Aug 10, 2003 1:33 UTC (Sun) by coriordan (guest, #7544) [Link] (2 responses)

> IBM has contributed code both under corporate and personal copyrights

but the question is:
where's all this code that IBM gave FSF the copyrights of?
(some say FSF own all of IBMs contributions, I heard FSF owns the s390 port, the Linux source doesn't confirm any of this.)

If FSF owns the copyright, the file has to say so.

A brief look at IBM's counterclaims

Posted Aug 10, 2003 23:59 UTC (Sun) by garloff (subscriber, #319) [Link] (1 responses)

> but the question is:
> where's all this code that IBM gave FSF the copyrights of?
> (some say FSF own all of IBMs contributions, I heard FSF owns the s390
> port, the Linux source doesn't confirm any of this.)

For a port you need gcc and glibc support as well.
Contributions to these projects obviously need to be done by assigning
the copyright to FSF.

A brief look at IBM's counterclaims

Posted Aug 11, 2003 0:24 UTC (Mon) by coriordan (guest, #7544) [Link]

ah, yes. A quick grep of the glibc sources shows that the contents of the s390 folders are copyright FSF.

thanks garloff.

Ciaran O'Riordan

A brief look at IBM's counterclaims

Posted Aug 8, 2003 4:58 UTC (Fri) by mbp (subscriber, #2737) [Link]

> when is Linus going to cease-and-desist SCO from distributing Linux as per
> their GNU GPL violations?

I doubt if Linus is going to do that himself; it would only increase the media feeding frenzy and presumably he has more useful things to do with his time.

However it doesn't have to be Linus: unlike Apache or GNU projects, copyright in the kernel is widely held. Anyone with copyrighted code in the kernel can potentially tell SCO to stop distributing it. Indeed, a German hacker (currently unnamed) did so a few months ago.

Unfortunately SCO, in their current death-agony mode, are unlikely to respond to a cease-and-desist letter. I suspect you would need enough legal firepower to get an injunction against them, and that doesn't go cheap. The organizations with enough at stake to fund it (RedHat, IBM, OSDL, etc) are already acting. Note that OSDL includes many other major companies.

Perhaps some kernel hacker who got rich off LNUX shares will decide to put a few thousands into suing SCO... unfortunately by the time IBM are done with them, there probably won't be enough left for anyone else to recover their costs. (Think rotten.com gunshot wounds....)

A brief look at IBM's counterclaims

Posted Aug 7, 2003 20:59 UTC (Thu) by nix (subscriber, #2304) [Link] (1 responses)

That's `Lanham Act', I think, not `Lenham Act'.

A brief look at IBM's counterclaims

Posted Aug 7, 2003 21:02 UTC (Thu) by corbet (editor, #1) [Link]

In general, typos will get fixed more quickly and reliably if you mail a note to lwn@lwn.net.

This one is fixed, sorry. I was in a hurry.

A brief look at IBM's counterclaims

Posted Aug 7, 2003 21:08 UTC (Thu) by euvitudo (guest, #98) [Link] (3 responses)

My question in this mess is: When will Microsoft sue SCO for selling them a useless license?

My conspiracy theory is that the only thing MS wants is domination, and that means killing the competition. They are also struggling with their image regarding FLOSS. Hence, they buy a SCO license, let the FLOSS community hate SCO, wait for IBM to countersue, then sue SCO for a useless license. MS would love to look like it's finally agreeing with the FLOSS community, so they go after SCO to obtain the admiration of the FLOSS community. Then they look for other ways to destroy the competition (read: especially now that everyone is balking about indemnification, and IBM is enforcing patent rights).

Any thoughts?

Cheers!!

A brief look at IBM's counterclaims

Posted Aug 7, 2003 23:28 UTC (Thu) by JoeBuck (subscriber, #2330) [Link] (2 responses)

Microsoft knows that the license is bogus; they paid the cash because it helps their anti-Linux FUD campaign and helps keep SCO alive.

A brief look at IBM's counterclaims

Posted Aug 8, 2003 3:32 UTC (Fri) by euvitudo (guest, #98) [Link] (1 responses)

My argument is that it's all part of the game they play. I agree that they are most likely aware of the uselessness of the license. However, they can claim that they didn't know and most of their cohorts and stock-holders will believe them, and support them in helping the rest of the world finish SCO, making MS look like they are "on our side", and not out to get us.

My first reaction is to be very skeptical at every move that these big corporations make. There is always an ulterior motive, and I believe you as well as many others share that opinion.

So, where will they put their resources next?

But then again, shouldn't we really be working with the hardware makers et al., and help them see the benefit of designing their hardware with Linux in mind. I think this is a much more worthy goal than to try to catch up with the competition, and allow them to influence the hardware makers and friends before we can. (Of course, I am also guilty of following every move the competition makes--something that I should change.)

A brief look at IBM's counterclaims

Posted Aug 8, 2003 7:13 UTC (Fri) by anselm (subscriber, #2796) [Link]

Microsoft can well afford to pay SCO ten million dollars on the off chance. They have
enough money to burn.

photo of SCO CEO Darl McBride on Yahoo News

Posted Aug 8, 2003 0:15 UTC (Fri) by cpeterso (guest, #305) [Link] (4 responses)

For you curious types, SCO CEO Darl McBride's photo is featured on Yahoo News. :-)

http://story.news.yahoo.com/news?tmpl=story2&u=/030807/168/4wymv.html&e=3&ncid=1216

photo of SCO CEO Darl McBride on Yahoo News

Posted Aug 8, 2003 4:07 UTC (Fri) by yodermk (guest, #3803) [Link] (3 responses)

Aaaiiieeee... please, please, PLEASE don't post long URLs in comments! This distorted the display and I had to scroll left and right when reading each line of the article and other comments. Not too much fun!

If you must post long urls, you can run them through http://tinyurl.com/ and paste the resulting short URL in the comment. Or, and I *think* LWN allows this, just put them in standard HTML "a href" notation.

Thanks!

photo of SCO CEO Darl McBride on Yahoo News

Posted Aug 8, 2003 18:10 UTC (Fri) by cpeterso (guest, #305) [Link] (2 responses)

are you using a 640x480 screen or something? <:-) It doesn't seem unreasonably wide to me, but your suggestion is noted. thanks.

(OT) Width

Posted Aug 9, 2003 12:16 UTC (Sat) by hummassa (guest, #307) [Link]

I use big fonts to browse, and I suffer from the same problem. Big URIs or other
stuff without spaces gets horizontally bigger than my 1152x896 screen.

photo of SCO CEO Darl McBride on Yahoo News

Posted Aug 14, 2003 10:57 UTC (Thu) by Wol (subscriber, #4433) [Link]

I browse in a WINDOW!

I *HATE* full-screen mode!

The whole point of html is that the USER should be able to CHOOSE how stuff displays. So long urls that won't wrap and screw up the display show total disregard for the principles behind html.

Cheers,
Wol