Ben forwards a link to this Byte article, SCO:
All your base are belong to us. His commentary:
One day I’ll have a blogtastic dalymount.com, but for the moment, have
you seen this priceless interview, in which SCO goes over the edge into
complete barking insanity?
‘We believe that UNIX System V provided the basic building blocks for
all subsequent computer operating systems, and that they all tend to
be derived from UNIX System V (and therefore are claimed as SCO’s
intellectual property).’
His emphasis. But let’s face it, he’s emphasising the right part ;)
So they now think they are owed money by every modern OS: that
includes FreeBSD, Windows, Apple, presumably QNX, etc. etc. Linux was
just the easiest one to start with, since the source is available and IBM
(with their deep pockets) are closely allied with it. MS have already
paid up for a SCO license, although many commentators see this as a means
to support SCO in their anti-UNIX lawsuits.
In more detail, SCO claim to have full IP rights to several major components
of any high-spec OS:
-
JFS (Journalling File System).
-
NUMA (Non Uniform Memory Access).
-
RCU (Read-Copy Update).
-
SMP (Symmetrical Multi-Processing).
Let’s pick one there: RCU in Linux seems to have originated (at a glance)
from code developed by Sequent for their DYNIX/ptx UNIX, which was an
AT&T UNIX System V-based OS. Sequent ran into trouble, and were
bought out by IBM. Later, patches to implement
RCU were submitted by IBM from Sequent’s code.
SCO now owns the AT&T UNIX System V IP; therefore SCO owns the RCU
code in Linux — even though Sequent developed it independently, on top
of the System V base, as far as I can see. Hey, that’s even more
‘viral’ than the GPL — at least the GPL tells you in advance what
mistakes you’d have to make for this to happen! ;)
In other words, it seems their POV is that, if any code came anywhere near
other code that may have been part of the original AT&T codebase, it’s
now tainted with SCO’s own ‘viral license’. Absolutely insane.
It’s unclear exactly how ‘all subsequent computer operating systems’ also
infringe this viral license, but SCO reckon they do.
In the meantime, they don’t seem to have realized that these kinds of
over-broad claims are not looked on favourably under EU law; while they
make cartooney threats in the US, they open themselves up to all sorts
of anti-trust-type claims elsewhere in the world. But then, at this
stage I don’t think they plan to actually offer any products, or operate
as a software company, so they probably don’t really care about that.
To really muddy the waters, an ex-SCO employee has recently
made allegations that SCO copied code from the GPL’d Linux kernel
into their UnixWare product.
Ah, fireworks. Anyway.
For a kinder, gentler form of total insanity, check out the guidelines for
forming ‘inexplicable
mobs’ in Manhattan — via bb.
Totally cool.
Patents: the SSLeay workaround
during this ongoing European software patents thing, I was reminded of a comment I heard a while back from a pro-patent guy.
He was around in the bad old days of SSLeay‘s patent woes. SSLeay, like many cryptographic products in the 80’s and 90’s before the RSA and other patents expired, was in a legal grey area due to patent issues. To quote the ‘Is This Legal?’ section of their FAQ:
Eventually, RSA relicensed their algorithms to be freely usable. Thankfully IDEA could be avoided by using alternative algorithms in the SSL transaction, so it wasn’t a biggie; most SSL users just switched it off. Finally, the RSA patent finally expired — so nowadays SSL is commonplace, and using SSL to protect security is a lot easier than it used to be.
Anyway, I’m diverging here… the relevance is this mail from Hartmut Pilch discussing the current euro-swpat proposal. He reckons even the SSLeay defense — saying ‘do not download this software in these countries unless you get these licenses’ — would not work with the current proposal: