LWN.net Logo

Linux's lucky lawsuit (National Post)

This National Post column says that the SCO case is a good thing for Linux. "Open-source advocates are outraged at the audacity of the lawsuit. They should instead be thankful. Linux must inoculate itself against the nasty legal toxins that are endemic in the corporate environment. And if we were to perversely pick a poison, the SCO suit has a lot going for it. SCO is strong enough to provoke a strengthening of Linux's defences but not so strong that it poses any real danger."
(Log in to post comments)

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 16:23 UTC (Sat) by chel (guest, #11544) [Link]

First they ignore you,
then they laugh at you,
then they fight you, <--- We are here.
then you win.
-Ghandi

Linux's lucky lawsuit (National Post)

Posted Aug 11, 2003 17:50 UTC (Mon) by davidw (subscriber, #947) [Link]

As an aside, as of Monday the 11th, SCOX seems to be sinking like a rock again. Anyone know why? I haven't seen any new news.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 17:15 UTC (Sat) by jdthood (subscriber, #4157) [Link]

Good article. The inoculation metaphor is apt. (I hope that
the one of the toad and the steamroller will prove to be so,
too.) However, the author gives no detail about how GNU/Linux
can build up defenses against patent assaults.

> So far the Linux community has taken a passive stance
> toward patent rights. It is relying on the legal firepower
> of its corporate friends to shield it from attack. The more
> expensive alternative is to set up a stand-alone body that
> actively pursues patent protection for Linux. This has the
> advantage of helping Linux maintain a certain independence.

Where is this stand-alone body going to come from and what
form will it take?

Linux's lucky lawsuit (National Post)

Posted Aug 11, 2003 14:20 UTC (Mon) by djabsolut (guest, #12799) [Link]

Not exactly independent, but OSAIA and OSDL are a stab in the right direction.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 17:32 UTC (Sat) by lees (subscriber, #13854) [Link]

I have to admit, I mostly agree with this article and have been waiting a while to read something like it. The SCO lawsuit is like a flu-shot or other innoculation, it hurts, and you might get a bit of a cold, but in the end it leaves you stronger and better protected.

So far, there has been a great gnashing of teeth but development hasn't been impeded, nor have deployments and sales been hurt. (Apparently, though I don't think we have heard much from SuSE or Red Hat about sales, and considering they have both started legal action against SCO, they must be a bit worried. Or, at least feel the need to be able to point out to their customers that they are fighting back against SCO in some way.)

How will this leave Linux stronger? I think it will create a nice precedent (if not legal, than cultural) that will answer a lot of doubts that people have about free software. There is a too-good-to-be-true aspect to free software, and once this lawsuit is settled (in Linux's favor, whatever that means) Linux evangelizers and Linux company sales-forces will be able to say to prospective customers: "You don't have to take our word for it, you really do have all these rights, SCO tried to take them away, and look what happened."

And I have no doubt that when push comes to shove SCO is going to have the shit slapped out of it in a court room. All of their claims are confusing, cannot be publicly shown, and their story keeps changing, always signs of a weak, desperate case. Think about how quickly the record companies get injunctions against people when they detect copyright violations. If SCO had anything the least bit meaningful they would immediately ask the courts to stop kernel.org from allowing free, unrestricted downloads of its intellectual property, and I'm sure court would oblige SCO, especially in today's atmosphere. The mere fact that this has never happened is enough to convince me that SCO is just playing a game, hoping to parlay this lawsuit into some money before it is laughed out of court.

That said, and as long as I am in wild-speculation-about-stuff-I-hardly-know-about mode, it seems like the worst thing that could happen at this point is for IBM to buy SCO and make all this go away. That won't make anything go away. That would just suggest that SCO's claims have merit, and moreover leave the Linux community (and the rest of the free software community as well, because at the moment it seems that whither goeth Linux, whither goeth free software, at least in much of the public's eye) tremendously in IBM's debt and somewhat at its mercy.

What would you rather be fighting: a floundering SCO with fluffed up, bogus IP claims, or a few years down the road, an arrogant and powerful IBM, looking to lock up Linux? I know we think now that IBM "gets it", but all that has to happen for that to change is a few quarters of declining revenues. No thanks. IBM might be a good friend to have, but I wouldn't want it to be anything more than that.

Linux's lucky lawsuit (National Post)

Posted Aug 11, 2003 1:21 UTC (Mon) by oloryn (guest, #7408) [Link]

If SCO had anything the least bit meaningful they would immediately ask the courts to stop kernel.org from allowing free, unrestricted downloads of its intellectual property, and I'm sure court would oblige SCO, especially in today's atmosphere. The mere fact that this has never happened is enough to convince me that SCO is just playing a game, hoping to parlay this lawsuit into some money before it is laughed out of court.

This assumes that SCO wants the infringement to stop. All indications are that they don't want it stopped, they want it to continue as the basis for claiming license fees on each Linux installation. Their problem is that their normal revenue is shrinking. Stopping downloads won't increase their revenue. Letting the downloads continue and then charging for the 'IP' the downloads supposedly contain (from their viewpoint) will. Hence their resistance to actually identifying the supposedly infringing code, as identifying it will allow us to replace it, thus cutting of this supposed revenue source.

Linux's lucky lawsuit (National Post)

Posted Aug 11, 2003 11:13 UTC (Mon) by Wol (guest, #4433) [Link]

SCO may not want it stopped, but IBM want SCO stopped. That's why they've sued for breach of the GPL.

And given that the RIAA and music industry get "cease and desist" orders quickly, shouldn't we be glad that IBM has gone for the same? Expect SCO to get kicked out of the linux industry by a clause 4 nuke in pretty short order.

Even if the case then takes years to get to court, it'll be very hard for them to sound credible when we've got a court order saying they've stolen our IP by breaking the GPL ...

Cheers,
Wol

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 17:36 UTC (Sat) by sandy_pond (guest, #9734) [Link]

"Tighter copyright control is needed on Linux software."

Although I disagree with the need for customer indemnification I do think that the copyright situation needs to be cleaned up or at least organized. Maybe OSDL should put a together website of the existing Linux copyright holders and what code they hold the copyright to. Then get everyone listed to claim their copyright or sign over their copyright to one of the active maintainers.

Then if licensing issues arise at least we have a chance of getting a consensus for license modification.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 19:31 UTC (Sat) by mattdm (subscriber, #18) [Link]

Obviously, the author has got some factual errors. If a kernel contributor dies, the status of his/her kernel code isn't uncertain at all -- you can't revoke the GPL.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 19:57 UTC (Sat) by coriordan (guest, #7544) [Link]

> If a kernel contributor dies, the status of his/her
> kernel code isn't uncertain at all

Unless he/she decided to "outsmart" FSF by releasing their work under the GPL v2 only (removing the "or any later version"). In this case, when GPL v3 comes out, there could be compatibility issues when the majority of the files are v3 but the dead persons files are still v2. The person that inherits their copyrights may not have any interest in software.

Ciaran O'Riordan

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 20:29 UTC (Sat) by steven (guest, #13639) [Link]

I believe there actually is a lot of code in linux that is GPLv2 only, because many people distrust the FSF. You never know that the FSF may come up with for v3 that a developer really dislikes, and developers have no control over future GPL versions. With the "or (at your option) any later version" you potentially give away more rights than you really wanted to.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 20:51 UTC (Sat) by sandy_pond (guest, #9734) [Link]

"I believe there actually is a lot of code in Linux that is GPLv2 only"

This is correct and also why I think we need better controls.

In the present state, if licensing issues arise with GPLv2, what probably would happen is that some organization would release the code under a GPLv3 or some such after getting a consensus from many of the copyright holders and then wait for specific copyright holders to complain. Then try to either address the complaints or replace the problem code.

Before this would happen though there would have to be a major problem identified with GPLv2.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 21:11 UTC (Sat) by coriordan (guest, #7544) [Link]

There will be GPLv3 drafts made available for public review when it's ready. The planned date for a GPLv3 draft is approx 12 months away, it's main goal is to plug the "web services loophole", clarify it's patents patents position, and make it compatible with more free software licenses.

Making these changes is more important than retaining backwards compatibility, so GPLv3 code may not be mixable with GPLv2 code, hence the standard distribution terms are "v2 or any later version".

Torvalds has asked the Linux hackers to use GPLv2-only. Presumably, if he likes the v3, he'll then ask the hackers to move to v3-only. Code written by anyone that has fallen out of communication since the move to GPLv2-only will remain GPLv2-only and will cause a problem.

It would have been better if he held of on his GPLv2-only declaration until he had seen the terms of GPLv3. (and I highly doubt FSF would stick anything contraversial into what is possibly the most used software license in the world.)

Ciaran O'Riordan

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 22:22 UTC (Sat) by dmantione (guest, #4640) [Link]

Maybe, but giving third parties is the right to license your software is plain stupid. Really
stupid. Even if it's the free software foundation.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 23:01 UTC (Sat) by coriordan (guest, #7544) [Link]

For projects to share code, they have to use compatible licenses. For the problems with the GPLv2 to be ironned out, GPLv3 has to be incompatible with GPLv2.

If all software packages have to be manually converted to GPLv3, adoption will be slowed, previous releases of packages may never be updated, and current releases of packages where the author is no longer interested will never be updated. A lot of code will become unusable for the community.

People can always switch to "GPLv2-only" after looking at GPLv3. And if GPLv3 is obviously bad, most people will switch, thus FSF would have wasted their time.

We don't live in a perfect world. Problems exist, we have to solve them. FSF aren't going to screw us.

Ciaran O'Riordan

Linux's lucky lawsuit (National Post)

Posted Aug 10, 2003 13:12 UTC (Sun) by evgeny (guest, #774) [Link]

> Torvalds has asked the Linux hackers to use GPLv2-only. Presumably, if he
> likes the v3, he'll then ask the hackers to move to v3-only. Code written
> by anyone that has fallen out of communication since the move to GPLv2-only
> will remain GPLv2-only and will cause a problem.

anyone that has fallen out of communication => unmaintained code => the piece of code will be dropped/rewritten anyway, as kernel history shows. So I'd expect any v2/v3 compatibility problems would be resolved "automatically" in a one-year or so time interval.

Linux's lucky lawsuit (National Post)

Posted Aug 10, 2003 13:56 UTC (Sun) by coriordan (guest, #7544) [Link]

> fallen out of communication => unmaintained code

Code can be maintained by someone other than the original author.

The original author would still be the copyright holder of his work. The maintainer would be the copyright holder of his changes. Strictly, it would be illegal for the maintainer to change the license of the files.

So you'd have two options:
1) quietly change the license anyway (this could haunt Linux)
2) rewrite the code, despite that it is maintained.

Linus made his annoucment on sep-8th-2000, the possible release of a GPLv3 draft is May-ish-2004. That's nearly four years for people to fall out of communication, making this problem quite realistic.

Ciaran O'Riordan
--
"Every line of code I wrote is under the GPLv2 or later"
-Alan Cox

Linux's lucky lawsuit (National Post)

Posted Aug 10, 2003 16:52 UTC (Sun) by evgeny (guest, #774) [Link]

> Linus made his annoucment on sep-8th-2000, the possible release of a GPLv3
> draft is May-ish-2004. That's nearly four years for people to fall out of
> communication, making this problem quite realistic.

Still, the following criteria should be met for a problem to arise:
1. The original author is certain about the v2 stickiness - in spite of the fact that Linus (and supposedly a great majority of kernel hackers) have adopted v3 by that time.
2. The code is in the core of the kernel; a driver under a different license isn't a big problem.
3. In spite of being a core kernel module, it still (after a couple of years, i.e. having survived 2.2/2.4/2.6/...) contains a significant non-trivial part of the original "v2-only" code.

I'd consider such a probability quite close to zero, but we'll see.

PS. I personally license my code under v2 with the "or later" clause, but understand those prefering first to see what's inside v3...

GPL v2 vs later

Posted Aug 11, 2003 1:42 UTC (Mon) by vonbrand (subscriber, #4458) [Link]

AFAIU (IANAL and so on) the fact that something was rewritten so it has no original code anymore doesn't make the result any less of a derivative work.

Anyway, such a switch would include a horrible amount of work. You'd have to track down everybody who contributed even small patches (no, their names aren't recorded except (possibly!) on LKML), and get them to agree.

This is a way in which the FSF is right: Having all (c) turned over to one entity clears this up. OTOH, if a hostile takeover should ever happen...

GPL v2 vs later

Posted Aug 11, 2003 18:30 UTC (Mon) by jeroen (subscriber, #12372) [Link]

...then nothing happens in the case of the FSF, because the copyright assignment contract has clauses which says that they can't make your code proprietary.

Linux's lucky lawsuit (National Post)

Posted Aug 11, 2003 11:22 UTC (Mon) by Wol (guest, #4433) [Link]

As I understand it, Linus has NOT asked other people to "v2 only". He has just publicly stated that *his* code is "v2 only", which means that the kernel AS A WHOLE cannot be relicenced without his EXplicit permission.

If he decides he likes v3, he just has to change his position statement, and the kernel licence will change near-enough on the spot.

Anyway, as I understand copyright law, the GPL isn't a problem for copyright HOLDERS because you have dual licencing. There's nothing stopping me mixing MY code (with really nasty restrictive licencing) and a load of GPL code, and *distributing* *it*! provided my code is dual-licenced with GPL being one of the options.

Copyright law is dead easy to understand, once you get your brain round the fact that there is a difference between licencOR and licencEE, and that most EULAs are a CONTRACT while the GPL isn't.

Cheers,
Wol

Linux's lucky lawsuit (National Post)

Posted Aug 11, 2003 16:47 UTC (Mon) by rgmoore (✭ supporter ✭, #75) [Link]

It would have been better if he held of on his GPLv2-only declaration until he had seen the terms of GPLv3.

I think that Linus has this one right. It makes more sense to have everything licensed under a single, specific license rather than under v2 or later. This is because multiple licensing leaves the program as vulnerable as the worst license under which it's available. If, for instance, you licence under v2 or later and then discover that there's a problem with v3, it's too late; the code has leaked out under the (hypothetically) undesirable v3 and relicencing back to v2 only can't stop the damage. Similarly, if a flaw is discovered in v2, licensing under v2 or later can't stop anyone from exploiting the problem in that version and you'll have to relicence to v3 exclusively anyway.

Linux's lucky lawsuit (National Post)

Posted Aug 12, 2003 6:11 UTC (Tue) by MLKahnt (guest, #6642) [Link]

You do need to keep in mind, however, the distinction between Copyright (resting with the author, or the employer of the author if done as an employed task for the employer) and a license such as the GPL. Copyright continues to rest with the copyright holder, in most domains, for 50 years after the death of the copyright holder, or 50 years from creation date in the case of an employer (longer in the case of the United States, but Copyright protection has been traditionally weaker there.) After expiration of the copyright term, the item enters the public domain.

The GPL allows copyright material that can be accessed under the GPL to be re-copyrighted by someone making changes, so long as they license this varied edition under the GPL for whoever may be interested. Linus Torvalds, technically, can still claim copyright to the collective code on his branch of the kernel by "re-using" the code of submitted patches under the GPL, and applying his copyright to the collective work.

So assuming that Linus has another 100 years +/- some additional number of years, Linux would fall into the public domain in 150 +/- the appropriate point if it is copyrighted under Finnish law (iirc) - an interesting consideration is that a public domain Linux would permit its distribution under non-GPL terms, although IANAIPL (although I have studied it, including reading the appropriate treaties.)

Just remember, IP law is a pain to understand, and the GPL deliberately turns it on its head to accomplish something not in keeping with the way the law and treaties thought they were taking things.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 18:32 UTC (Sat) by rjamestaylor (guest, #339) [Link]

Saw this article on news.google.com this morning. Good to see this take outside the pro-open-source, non-technical media. I'd say "mainstream" but it is a Canadian outlet.


:)

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 20:15 UTC (Sat) by josh_stern (guest, #4868) [Link]

The National Post column is deficient in a number of areas:

1) It makes too many assumptions about the factual basis of
SCO's case. For instance, it assumes there really is SCO
proprietary code that IBM "copied" into the kernel. In fact,
all we know is that a) SCO claims this, b) under a strict NDA
they showed some mysterious small amount of code ("80 lines")
that was near identical, c) they claim the "80 lines" originated
in their codebase, and d) they told some NDA participants,
including a Deutsche Bank analyst, that IBM did not contribute
the "80 line" code to Linux, and that IBM's violations were
contributing the general technologies of RCU, NUMA, etc.
http://asia.internet.com/asia-news/article/0,3916,161_2245911,00.html
Previous lwn stories have discussed this issue and found it
unlikely that SCO is the owner of any patents or trade secrets
related to RCU, NUMA, etc., and as Linus Torvalds and others have
correctly observed, whether or not they can make a contract case
against IBM for helping Linux with these technologies depends
mainly on the details of the non-public contract between SCO
and IBM and has no legal import for Linux distributors or end-users.
As someone who has followed most all of the stories and legal
filings on this case (typical LWN reader :), I find SCO repeatedly
contradicting itself and making claims that don't sound plausible
on either factual or legal grounds, and generally acting in a
way that seems illegitimate and intended to deceive. Therefore
I am not willing to suppose, until proven otherwise, that *any*
of their claims are true. I regard an article that does so, like
the National Post column, as very speculative (even though
they believe SCO doesn't have a winnable case anyway).

2) The column says that Linux distributors should "indemnify their
customers from any intellectual property issues". This is
probably a bad and impractical suggestion. It means, for instance,
that if any portion of "Linux" were found to violate some patent
claim or copyright claim, and, as a result, some functionality
was no longer available to Linux users, then the distributor
would be liable for any resulting business loss. The related
possibilities are so open-ended, and the due-diligence so hard
to do, that no Linux distributor would likely be able to afford
the premiums on the reinsurance policy that would make this
possible. And it would be very hard to find a company willing to
sell such a reinsurance policy. Moreover, it is not a known
practice among proprietary software vendors to offer anything
like this, for the same reasons. I couldn't assign any tangible
meaning to the later remarks in the column that suggests a
"stand alone body that actively pursues patent protection for
Linux". Does it mean that perhaps a big gang of IP lawyers
would volunteer non-billable time to search for potential
patent violations?



Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 21:02 UTC (Sat) by sandy_pond (guest, #9734) [Link]

<i>The column says that Linux distributors should "indemnify their
customers from any intellectual property issues". This is
probably a bad and impractical suggestion.</i>
<p>One thing you miss is that an indemnification clause is incompatible with the GPL.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 22:29 UTC (Sat) by dmantione (guest, #4640) [Link]

Nonsense, the GPL says there is no warranty. In other words, based on the GPL you
have no right to claim warranty. That does not mean that a company that is willing to
take the risks cannot provide warranty services.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 22:47 UTC (Sat) by sandy_pond (guest, #9734) [Link]

Well .. IANAL but given the conditions laid out in the GPL a warranty sub-license seems unworkable to me.

Linux's lucky lawsuit (National Post)

Posted Aug 10, 2003 1:28 UTC (Sun) by coriordan (guest, #7544) [Link]

Anyone can offer support for free or for a cost, it's got nothing to do with the softwares license.

Linux's lucky lawsuit (National Post)

Posted Aug 11, 2003 9:07 UTC (Mon) by forthy (guest, #1525) [Link]

You can't fully indemnify against IP right violations due to section 7 of the GPL: If you are
faced circumstances that contradict the nature of the license (e.g. you have to pay patent
license fees or some antidot license fees on a *per-CPU* basis), you can't do that. All you
possibly can is to buy unlimited licenses - if offered.

Linux's lucky lawsuit (National Post)

Posted Aug 10, 2003 9:55 UTC (Sun) by jeroen (subscriber, #12372) [Link]

To quote the GPL:

"You may charge a fee for the physical act of transferring a copy,
and you may at your option offer warranty protection in exchange
for a fee."

Linux's lucky lawsuit (National Post)

Posted Aug 11, 2003 16:52 UTC (Mon) by rgmoore (✭ supporter ✭, #75) [Link]

One thing you miss is that an indemnification clause is incompatible with the GPL.

Not at all. Just read the GPL yourself. Right there in section 1 it says:

You may charge a fee for the physical act of transferring a copy, and you may at your option offer warranty protection in exchange for a fee.

That makes it extremely clear that offering a warranty or other indemnification in exchange for a fee is specifically allowed by the GPL.

Linux's lucky lawsuit (National Post)

Posted Aug 9, 2003 22:28 UTC (Sat) by mmarq (guest, #2332) [Link]

" Does it mean that perhaps a big gang of IP lawyers
would volunteer non-billable time to search for potential
patent violations? "

Why?... GPLed code is free to use, study, modification and distribution,... even if M$ use stollen code in their products, they are not really gaining ground but trying not to lose more... so in big majority of ways, lawyers are only needed for defense and not for attack.

Linux's lucky lawsuit (National Post)

Posted Aug 10, 2003 0:10 UTC (Sun) by josh_stern (guest, #4868) [Link]

I was talking about defense, which was the context of the original
article. In fact, I don't know exactly what the original author
had in mind and this is part of my criticism. The bit about the
lawyers was by way of reaching for an example of what it could
possible be. The idea is that they would go over Linux software
and the applicable patent databases with a fine tooth comb, looking
for areas of potential violation within Linux (which could then,
perhaps, be implemented in a different way).

Linux's lucky lawsuit (National Post)

Posted Aug 10, 2003 4:29 UTC (Sun) by neoprene (guest, #8520) [Link]

The statement: "For a start, the Linux community should move quickly on the following issues:- Anyone who packages Linux in their products should indemnify their customers from any intellectual property issues. "

forgets:

1.) that Linux is the _kernel_, which can be downloaded free from www.kernel.org, regardless of any association with a "package" and "product". It can be separated from a distribution, and installed by the "customer" through ftp via script whatever. [..from servers outside the US borders and treaties]

2.) "Linux" is not a commercial product.

3.) when you did not pay for something you are not a customer.

4.) again, the "Linux community" is not a corporation, they have no commercial agenda, no board meetings, no quaterly reports, no pension plan, no production and sales quotas. Use Linux "as is" or f**ck off.

These journalists, and most of the "suits", keep forgetting that Linux is not a commercial enterprise. The GPL is not just another EULA. If there was a chink in the GPL, M$, SUNX et al would have pierced it long time ago. All they can do now is lie about the GPL and promulgate idiotic "what-if" schemes.

The very idea of "indeminfication for customers" "from any intellectual property issues" is some stupid ass mantra being fathered at Redmond and is the buzz all over the web. With $45 Billion in cash M$ could afford to not only indemnify anything and everybody, three times a day, but also purchase several medium sized countries. Get off it.

I read in the XPee EULA that M$ gives your purchase price, or $5.00, back, whatever is more, if their stuff makes you unhappy.

If you don't like Linux and Open Source software, no problem, just plunk down the dough, buy that wonderful closed source OS, and be happy.

GNU/Linux is just the operating system, there's tons more money to be made from applications, support, and associated hardware sales. But if all you can do is an OS and a few chicken-shit webbrowsers, "office" packages et cetera, I'm sure you would be whining.
People and corporations ARE willing to spend big bucks on accounting-, plant management-, engineering-, CAD-, and graphics-software and associated support.
M$ has made too much money on low end commodity junk.


Linux's lucky lawsuit (National Post)

Posted Aug 11, 2003 10:27 UTC (Mon) by Duncan (guest, #6647) [Link]

As it now stands, perhaps the biggest thing holding up even faster adoption of
Software Libre (including Linux) is the lack of court case history. The GPL
remains unproven in court, and some corporations understandably have jitters
about putting their mission critical applications on something that could
conceivably be putting them in as sure a locked situation as proprietary-ware
does, when the company that developed and supported it goes belly up.

Thus, while the National Post article has some problems, already mentioned by
others, I too have been saying this SCO stuff is a good thing and can only
indicate the healthy maturing of the whole idea behind Software Libre. Even if we
lose the SCO court cases, unless it overturns the GPL entirely, it can't BUT be a
good thing, big picture, because it establishes some badly needed case history
where there is now none. Even if there are damages to pay, Software Libre in
general and Linux specifically will be better off, because it will now be a far more
legally defined quantity. If we win, as it looks like we have a good chance of
doing, so much the better! As they point out, it could have been worse. We are
lucky we got the case we did as a first one. It looks to be an easy one, not even
directly putting Linux and Software Libre in danger at this point, and the
precidents it sets will quite likely end up being very useful with the bigger ones
ahead. If it goes REALLY well, it might even cause many of those potential
bigger ones to disappear, and set us up well for the inevitable one or two
remaining. That setup may help them go good, leaving few willing to make the
challenge.

As for the further problems they suggest and actions they propose, not so good.

1) Indemnification. As others have pointed out, proprietary-ware doesn't even
attempt this, and asking open source to do it is not only foolish, but COULD easily
be deliberately setting us up to fall, which it seems SCO is now attempting to do.
(If it can't join the party, and is looking likely to be past history, it's going to take
what it can that it sees as doing it in down with it!)

As was pointed out in comments on another article, should some company be
stupid enough to actually fall for the indemnity suggestion and NOT go buy the
SCO license in ordered to do so, SCO will quite likely concentrate action on it,
forcing it to do so as the lower cost of that or fighting a court battle. The article
seems to think this should be a costless thing to do, as SCO doesn't have much
of a legal case anyway. That's NOT the case, as anyone foolish enough to do so
will likely end up either purchasing SCO licenses, or paying far more in court costs
than they'd cost to defend itself. Further, it would seem the licenses are
incompatible with the GPL anyway, for a distributor (not so for the end user, since
the GPL doesn't limit use rights, only copy and distribute rights), so any that buy
into that plan weaken the GPL in additon to directly supporting its enemy, even if
no one directly enforces the GPL on them forcing them to stop distributing.

2) A patent action team. Impractical. These cost big $$ (or big favors, if done
pro bono on the scale they seem to suggest), something in short supply when the
product at issue itself is available at no cost. The developers themselves can't do
it or they end up being tainted due to the research they've done, and therefore
less able to create totally "clean room" implementations, since it could be argued
that since they've seen the code, it's now a derivative work. That's in addition to
the time devoted to that that therefore isn't available for further open source
development. Few others are going to directly understand the issues involved
either on the legal or technical side without some serious study, and if they are
interested in the issue already, they are probably already on their way in either law
or open source development. If the former, costs, again, on the scale suggested,
if the latter, taking them away from the development they otherwise WOULD
have done in the software libre community. Thus, either case has its serious
costs.

Far better to take the issues as the come, since coding around them is likely to be
an option anyway, and that will again be far easier if the folks doing so don't have
inside knowledge of the code they are supposed to be developing an alternative
for.

OTOH, there's the approach of Red Hat, and possibly IBM (tho that remains to
be seen, ultimately, and even RH could pull a Caldara/SCO Jeckle/Hide switch).
Red Hat has been filing its own patents recently. It *CLAIMS* it will only use
these defensively, and those in the free source community remain free to use
them as desired without fear. Many IP infringment cases in the proprietary world
end up being settled with a cross-licensing agreement, since both sides often
have technology the other side wants to use. RH will be able to do the same with
its patents and in that way it's a good move for them AND free source.
Unfortunately, that leaves us simply trusting that they won't go the way of
Caldera and become another SCO.. Thus, this is a very controversial approach
as well and not without serious risks. Whether they are worth it depends on
whether you trust RH and fate, and we will have to wait on time to find out.

Duncan

Linux's lucky lawsuit (National Post)

Posted Aug 11, 2003 18:07 UTC (Mon) by Ross (subscriber, #4065) [Link]

Why are you so concerned (or do you believe businesses are so concerned)
about testing the GPL? Do they have concerns for any license or contract
that hasn't been tested? That would certainly make for a lot of lawsuits :)

IANAL, so don't take this as legal advice. I agree that the GPL is
different than a typical EULA-style software license, but it's primarily
in a way that makes it stronger.

A EULA-style license uses on-screen buttons or shrinkwrapping to show that
you "digitally signed" the license. The GPL does not do this -- it does
not try to restrict use of the software. You can run the software without
caring about the GPL at all.

It is only when you try to copy or to distribute derivative works that it
kicks in at all. So by distributing a GPL'd program you are showing that
you either:

a) got specific permission from the author(s)
b) got permission by agreeing to the GPL
c) are infringing on the copyright of the work

It seems pretty simple to me. I believe the GPL was reviewed by a judge
in the MySQL lawsuit, but a settlement was reached before the case went
very far.

Copyright © 2003, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds