Microsoft/Novell Partnership

[Peter P.]

John Macdonald <john at ...> writes:

> On Fri, Nov 03, 2006 at 08:36:10PM +0000, Peter P. wrote:
> > It is extremely unlikely that any part of 'Linux' contains any Microsoft IP
> > since it has BSD Unix and U.C.Berkeley etc ancestry. Also anything GPL 
> > embraced
> > and extended and republished with another license is going to place the GNU
> > legal wing into battle position.
> 
> You've fallen into the trap of thinking that IP is a single
> coherent concept.  It's not - it's a conglomerate of three
> drastically different concepts.

I don't think that I have fallen into that trap. GPL is a license and I referred
to that license and especially to its effect wrt. 'aggregate works' (which
covers the extending part of 'embracing and extending' for most practical
purposes imho).

> There is trademark, which gives the some control over how a
> company's identity can be used.  (It's more complicated than
> that but trademark is not really involved in the Novell-MS
> agreement.)  Linspire has proven that Linux companies *are*
> subject to being attcked with trademark - but the fact that
> they just changed their name from Lindows and didn't get much
> additional penalty shows that the attack is not especially
> worrisome.

First of all Windows is a proof of the fact that dictionary words in common use
can be trademarked, even if they were already used in the sense implied by the
trademark when the trademark was applied for and granted (the word was used by
Apple in their OS and by Xerox et al as such at the time when MS applied for the
trademark for Windows, and it was well-known in the 'art') (and it also proved
that a trademark that could never have been registered in an EC country can be
upheld in it [3]). Second, Lindows proved that variations on trademarked
dictionary words are enforceable although the trademark is on 'MS Windows' and
Lindows does not contain the word 'MS' or the word 'Microsoft'. Third, Linspire
proved what you said in a backward sort of way (Microsoft settled at a loss
[2]), but it also exposed the fact that the depth of the pockets of the accuser
defines how far and for how long the defendant can be chased, and that the
barriers against leakage abroad (to the EC no less) from the currently flawed US
system are nonexistant (reference to [3]). Now, how does that rhyme with
equality in front of the law ?

It is interesting that Linspire did/does exactly what the Ballmer/Novell duet
promises to do in the future (and what Xenix, Microsoft's aborted Unix version,
developed afaik by SCO, did in the past for DOS and Unix interoperability,
before Xenix being dropped by Microsoft and following Microsoft using SCO as a
failed Corax on IBM and Novell).

Older but good articles describe this better (2002, before MS lost/won anyway):

  http://news.zdnet.com/2100-9595_22-893895.html
  [2] http://en.wikipedia.org/wiki/Linspire @Litigation...
  [3] http://www.theregister.co.uk/2004/01/30/microsoft_wins_lindows_fight/

> Finally, there is patent, which gives the owner control over
> an idea; even if others have the same idea but are unable (or
> can't afford the court costs) to prove that it was a well-known
> or obvious idea before the patent was applied for.  There is

In most places on this planet it is not reasonable to assume that a public
domain object, or something that has been in use for some time and is well-known
in the 'art' can be patented because it is 'novel' or has a 'novel application',
regardless of how much money the applicant has or how many times (and in how
many places) he tries to apply (and regardless if he tries to apply directly or
through lobby organisations, like 'copyright protection leagues' financed by
record and movie companies). That applies for procedures (and algorythms), and
dictionary words (and their plurals) as well. Most electronics companies have
gotten the message and changed their names accordingly. Unsurprisingly, those
are no longer dictionary words, nor common names, even if theor developer's and
founder's names had to be dropped (perhaps this is related to the fact that they
prefer to spend their money on R&D as opposed to tiying up the legal system with
lawsuits against John Does whose domain names resemble their own too much). ON
Semiconductor (maybe a less fortunate one), Renesas, Freescale, Agilent you name
them, they have all changed. 

The reason is that most sane people will laugh incontrollably at any attempts to
patent a microscope used to drive in nails as a 'novel method for using
microscopes and nails', and refer the applicant to www.halfbakery.com, and the
same thing applies for double clicks and business administration methods per se
(as opposed to implementations thereof - f.ex. a novelty product representing a
specially reinforced microscope designed to withstand driving nails in with
would be patentable, and so would be a particular mouse implementation (the
device) that uses double clicks for a specific purpose, as opposed to the
well-known principle of single, double (and multiple) activations of a control
element for the purpose of conveying different signals with the same element -
be they buying signals, or something else). Even 'Spam' was declared a common
word recently, although that was definitely not a dictionary word. There are
only a few dinosaurs left, resisting this. Windows is one of the larger ones.
Imho, the writing is on the wall. And, by the way, 'Linux', the word, could go
the same way if some packaging efforts are not made. That could allow the
Ballmer/Novell duet, and others, to use the name Linux anyway, even if
infringing the GPL. Does anybody want that ? I do not. So that's why names are
important and using the right name is important (this note is for Lennart - it
is not at all important what RMS thinks, it is important what the judge will
think when he will face that in this context). 

All one can hope is, that sanity will prevail eventually. In a sense, the US is
very advanced in the IP/trademark/patent domain, so advanced in fact, that there
are many things in the system that have serious flaws and were never or
insufficiently tested by time (they are being tested now, by these lawsuits). In
a way the current trademark and IP wars are that testing, and the companies
which get burned in the process are the early worms which get eaten, as the
saying goes. 

I also think that these proceedings must be watched from a distance and allowed
to take their course, and that their conclusions should not be exported to other
places too soon (as in, years or decades), especially as most other countries
cannot afford to support disastrous year-long lawsuits and litigations that are
sometimes baseless and often lead nowhere. This alone is imho a very good reason
to limit the damage to the proving grounds, which is inside the 20-mile zone of
the country that has chosen to allow these trials to take place on its territory
(and is not even consistent on that, since legislation inside the US varies from
state to state), and pay the price for the inconvenience they cause (see
Blackberry blackout and forcibly abandoned Lindows users among others, to stay
on topic).

> certainly reason to expect that Linux is susceptible to attack
> through patents; just how badly is still open to question but
> if Microsoft has a patent on some especially critical concept,
> it could be crippling for Linux.  More likely are that MS has
> lots of nasty patents that , while they can be worked around,
> it can only be done with a bunch of work and end up with worse
> code that is less interoperable with any commercial code that
> makes use of the petent.  This is the core of the Novell-MS
> agreement.  Offhand, I'm hoping that Eben Moglen's statement
> that Novell might have put themselves into a position in which
> they cannot legally distribute GPL-software that is covered by
> the patents they have attained rights too, since they are unable
> to pass on the right to use the patents to the extent required
> by the GPL.  It will be interesting to see how that plays out.
> If it is proven to be true, it will provide fuel for those who
> claim that the GPL is a virus.  (They'll be wrong - you can
> have proprietary code, and you can have GPL code, but you can't
> merge the two and treat it a single entity.  The proprietary
> code is just a viral as the GPL code in that scenario.)

I am not qualified to judge what a good patent is, and I am not a lawyer, but as
recent (f.ex. Canadian, see RIM) history shows, the US type of patent and IP law
that currently exists is not good for nearly everyone else, and not even for
themselves, according to what some authors write. Also monster lawsuits like the
SCO/Novell/IBM one over Linux, which turned out to be mostly a hoax and lasted
three years, is not something that other countries can afford to tie up their
industries with, and be used not as a lawsuit, but as a political weapon,
spreading FUD and influencing and manipulating developers and investors, and
possibly destroying local industry for years and causing unemployment and
capital emigration and brain drain. 

It is reasonable to assume that the powers that be from elsewhere (for example
from Canada, where this mailing list is hosted, and from the EU, where a Dutch
judge muzzled Lindows a few years ago) are following the issues and that they
will not allow the flaws in that patent and IP system to propagate outside the
borders of the US, and especially not within their own borders if they can help
it. This probably includes most RIAA and MPAA-related 'media taxes', DRM, and
the like, which were taken over literally from that system without any impartial
testing, and caused and still cause endless scandals where they are applied. 

These are unproven technologies coming 'hot' and unproven from the melting
pot/proving grounds which are the US in this domain now, and any early adopters
are bound to discover the flaws rather sooner than later, exactly as they have
so far. 

That includes lawsuits brought against children, attempts at subpoenaing
computers belonging to research institutions and government departments for
'searching', and worse, outside the US, by US-led 'legal consortiums' for
suspicion of their being used to 'download stuff', and other embarassments, like
having any computer running Windows with auto-update (and any computer, not just
Windows, using auto-update for any software on it, such as anti-virus software
and web browsers, yes that includes Firefox and Mozilla on Linux and *BSD
incidentally, sorry about that) on being subject to be loaded with any software
the originating company designates as an 'update', where the originating company
is under US or other jurisdiction and the receiving computer is somewhere else,
and the updates are almost invaraibly binary and not transparent to the user. Or
the fact that this (automatic updating and registration) is mandatory for
survival on the internet, and will be more so in the future, or so the new EULA
says. 

Who is going to verify that a 140MB service pack does not contain a trojan that
transmits sensitive data to 'home base' using perhaps subliminal coding in
browser request headers for only 15 computers, exactly three times for each, at
a random time after the upgrade for each, the computers belonging to a certain
domain of another country. The patch (pardon, service pack) would be identical
for the remaining 100 million downloaders, yet it would do nothing in their
cases. And if it does not contain a trojan today, who is going to check the next
service pack ?
 
So one should watch with interest what is happening, and carry on as usual (and
lobby one's lawmakers whenever they are about to put on that particular yoke).
As far as I am concerned, Linux is a Finnish product developed in a certified
white room. The 16 years of subsequent developments are improvements are
aggregate works and must abide by the original (GPL) license. I have a right to
an opinion like everyone else has.

As to viral genome, I think that the GPL gene causes less damage, and often
out-of-court settlements within a shorter time than the current USPTO gene which
is proven to generate year-long billion-dollar border-crossing
industry-crippling lawsuits over next to nothing, and additionally ipset
customers, investors and developers. If there is to be a genetic selection, then
the GPL should prevail as the more benign of the two, and the more damaging
strain remain quarantined in the legal area where it broke out until it will
become recessive or be superseded by something more sane and less dangerous and
costly.

In the bigger picture, hardware linux aggregators have often been in the
position of GPL infringement. Equally often they have yielded and instead of
pulling products they have complied and went on to become best sellers (see
Linksys routers which have a special version that is still marketed specifically 
because it can run Linux, and others). Code inspection technology precludes
'hiding' extensions to GPL code and thus it is a question of when, not whether
plagiates are discovered. How these are handled is a legal issue which, for most
people who do not live in, or develop in or under US jurisdiction, shold not
depend on US jurisdiction, but on that of their own countries, hopefully with
such systems that rich bullies cannot be more equal than others.

Peter P.

PS: Sorry for the long post, it will only happen seldomly.

PS2: I have an idea for a Halloween trick that could be good for OSS developer
friends - a little late for this year though - Put on a blue shirt over your
clothes, daub some water on it and on your forehead so you appear to be sweaty,
then when the door opens and you get to do the trick, run three times in a
circle smiling extatically and shout 'Microsoft, Microsoft, Microsoft in a
hoarse voice'. Instant success/FUD should be assured.


--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists