So, it looks like we'll finally get a court
to weigh in on whether newly-created code
(LVM and JFS, that were created in about
1989 through 1992), is somehow "owned" by the
owner of a piece of code it has to interact
with.
The block device driver interface (used by the
LVM) was documented publicly multiple times
(Egan & Texeira in the mid-80s, Sun Device
Driver books, the System V docs themselves, etc)
It seems that unless the contracts specifically
state that these are considered "modifications"
of the original AT&T code and not new works that
interact with the AT&T code, then the LVM claim
is junk.
The JFS claims are a little murkier. It was not
possible to create a UNIX filesystem circa 1990
without the source code to UNIX.
And my recollection of AIX 3 and JFS was that
it didn't just "plug in" to a well-defined interface, it needed lots
of intimacy with the VM system and other bits
of the kernel.
While the LVM was big, I don't think it got
close to weighing in at a million lines...
Slashdot Mirror
So, it looks like we'll finally get a court to weigh in on whether newly-created code (LVM and JFS, that were created in about 1989 through 1992), is somehow "owned" by the owner of a piece of code it has to interact with.
The block device driver interface (used by the LVM) was documented publicly multiple times (Egan & Texeira in the mid-80s, Sun Device Driver books, the System V docs themselves, etc)
It seems that unless the contracts specifically state that these are considered "modifications" of the original AT&T code and not new works that interact with the AT&T code, then the LVM claim is junk.
The JFS claims are a little murkier. It was not possible to create a UNIX filesystem circa 1990 without the source code to UNIX.
And my recollection of AIX 3 and JFS was that it didn't just "plug in" to a well-defined interface, it needed lots of intimacy with the VM system and other bits of the kernel.
While the LVM was big, I don't think it got close to weighing in at a million lines...