|
|
| Sandeep Gupta's Declaration of July 2004 - Redacted - SCO v. IBM |
 |
|
Saturday, July 16 2005 @ 02:21 PM EDT
|
We finally get to see the previously sealed Sandeep Gupta Declaration in Support of SCO's Opposition to IBM's Cross-Motion for Partial Summary Judgment [PDF], dated July 7, 2004, redacted by SCO.
Heavily redacted, I might add. All the accusations remain, but very little evidence. That surprises you?
It was submitted in opposition to an IBM motion that was denied, but without prejudice to refile or renew after discovery is done, and in case you are not a tech person and so you start to think it sounds convincing, you might wish to read the recently unsealed declaration by Dr. Brian Kernighan, who pointed out that Mr. Gupta failed to first remove unprotectable elements before doing his study, or to analyze whether any similarities were substantial, and that if he had done so, there would be nothing left: 3. In summary, I find fundamental errors in Mr. Gupta's conclusions. His conclusions of substantial similarity are flawed because he fails to exclude from comparison unprotectable elements of the allegedly copyrighted code, and he uses an indefensible standard for what qualifies as "substantially similar" code.
4. If unprotectable elements are excluded from the comparison and an appropriate standard of similarity is applied, there is no similarity between the parts of Linux identified by Mr. Gupta and the allegedly copyrighted works. In it, poor Mr. Gupta was given the unenviable task of trying to persuade the court that ELF is copyrighted by SCO and no one can use it without infringing their rights, among other things he argues. You will notice, on page 24, in paragraph 77, Gupta says that he will describe ELF code "that is copied either identically or substantially similarly in Linux." But it's all redacted after that, all the tables, all the specifics, pages and pages that say nothing but REDACTED. You may think it's because they read Groklaw and realized that we'd make fun of them for claiming ownership of something Novell put into the public domain years ago and that isn't protectable in the first place. But no, I think it's because they plan to swing around and argue this point some more when IBM refiles or resubmits their motion or at trial, if it makes it to trial. Or maybe both.
He also details the SCO claims regarding RCU, which Dr. Kernighan also answered in detail, as IBM explained in their memorandum in support of their motion: For example, Mr. Gupta opines that "Linux RCU is substantially similar to UNIX RCU". (Gupta Decl. ¶ 10.) Mr. Gupta's entire analysis, however, is focused on unprotectable ideas that must be filtered during any assessment of "substantial similarity". See Gates Rubber, 9 F.3d at 836 ("One of the fundamental tenets of copyright law is that protection extends only to the author's original expression and not to the ideas embodied in that expression."). Mr. Gupta himself describes the allegedly similar material he identifies in his declaration as "routine[s]" (¶¶ 3, 5, 10) and "methods" (¶¶ 6, 7) that "perform the same five acts" (¶ 11). This material is plainly unprotectable. See Gates Rubber, 9 F.3d at 836-37 (noting that "the main purpose or function of a program will always be an unprotectable idea" and that "the expression adopted by the programmer is the copyrightable element in a computer program . . . the actual processes or methods embodied in the program are not"). Moreover, when the actual expression--i.e., the code--in what Mr. Gupta calls "Linux RCU" and "UNIX RCU" is compared side-by-side, as in Mr. Gupta's own Exhibit A (in columns 1 and 4), even the untrained reviewer can determine that they are completely different and not even close to being "similar". You may know of other points you'd argue regarding RCU or other things Mr. Gupta claims. So enjoy. Speaking of fun, can anyone do an OCR or HTML of this declaration? Thanks, if you can. Just copy and paste REDACTED over and over, and you'll be mostly done. And if you wish to go back and reread Mr. Gupta's declaration submitted in opposition to another IBM summary judgment motion -- the 10th counterclaim -- and review the whole flap about whether he was testifying as an expert or a layman (he said as a layman), IBM's denied motion to strike (see also its reply memo in support), and how MIT's Dr. Randall Davis testified in his declaration that despite Mr. Gupta's efforts as a layman to prove similarity, he, testifying as an expert, found no such similarity, here you go. Davis concluded: Despite an extensive review, I could find no source code in any of the IBM Code that incorporates any portion of the source code contained in the Unix System V Code or is in any other manner similar to such source code. Accordingly, the IBM Code cannot be said, in my opinion, to be a modification or a derivative work based on Unix System V Code. Remember all those motions in July of 2004? No wonder the judge finally told them to stop until after discovery. What I ask myself is, how do they get Mr. Gupta to do these declarations? And now that two world-renowned experts have said he is mistaken, will we see any more? And a question to those who think once SCO is in front of a jury in Utah, it will be a slam dunk, imagine these two experts on the stand and then Mr. Gupta saying the things he does. Pretend you are a juror. Who would you believe?
Update: We also now have the Unsealed Exhibits to the Declaration in Support of SCO's Opposition to IBM's Cross Motion for Summary Judgment [PDF], a 1990 USL/Sequent licensing agreement.
|
|
|
|
| Authored by: Anonymous on Saturday, July 16 2005 @ 02:30 PM EDT |
| HTML, clickety, preview. All that, thanks. [ Reply to This | # ]
|
| |
| Authored by: Briareus on Saturday, July 16 2005 @ 02:31 PM EDT |
Corrections Here Please
---
scary times are never dull[ Reply to This | # ]
|
| |
| Authored by: MadScientist on Saturday, July 16 2005 @ 02:31 PM EDT |
| [ Reply to This | # ]
|
| |
| Authored by: dmomara on Saturday, July 16 2005 @ 02:32 PM EDT |
"LinuxRCU"? "UNIXRCU"? These people really aren't stupid
enough to think that a technique on which Sequent (currently IBM) holds patent
and the code for which was licensed to them during the Monterey project for
inclusion into their product range is theirs? [ Reply to This | # ]
|
| |
| Authored by: Briareus on Saturday, July 16 2005 @ 02:57 PM EDT |
The difference couldn't be more starkly demonstrated. IBM's expert witness
examines and dismisses the source-code claims that SCO's layman witness portrays
as being appropriated with the a supposed interchangability approaching that of
Lego DUPLO blocks. Unenviable task for Mr. Gupta indeed.
It's amazing, watching this story develop over the last two years. A lawsuit
conjured in greed, formulated against all reason and evidence, and cast from the
twilight zone, smearing the names and reputations of thousands of people.
Despite my respect for the legal system, it's hard not to be frustrated with
this whole turn of events. But considering the boost it has given open source
methodology, I do hope the exhaustive approach the Judge is taking will give SCO
all the rope it needs to hang itself, with prejudice.
Who knows, this may end up being looked back upon as an early and vivid example
of how enlightened community helped to end the age of frivolous litigation.
---
scary times are never dull[ Reply to This | # ]
|
| |
| Authored by: ak on Saturday, July 16 2005 @ 02:58 PM EDT |
Who would trust someone who is "employed by The SCO Group, Inc." - and
acts in the interest of the Yarro/McBride-gang ?!
I really love the formulation in point 16: "The first act, 'allocating a
new data structure of a certain size,' is expressed in UNIX RCU and Linux RCU by
a single line of nearly identical code. From a software programmers'
perspective, the UNIX RCU expression of the act of allocating a new data
structure has been identically copied into Linux RCU."
In fact from a writers' perspective, the word "code" has been been
identically copied into SCO statements from texts which I have written.
[ Reply to This | # ]
|
- ah, so its *you* who owns the "code" - Authored by: Briareus on Saturday, July 16 2005 @ 03:11 PM EDT
- lemme guess - Authored by: DebianUser on Saturday, July 16 2005 @ 03:33 PM EDT
- lemme guess - Authored by: Anonymous on Saturday, July 16 2005 @ 08:56 PM EDT
- lemme guess - Authored by: Anonymous on Saturday, July 16 2005 @ 10:07 PM EDT
- ISO C vs. C++ - Authored by: Anonymous on Saturday, July 16 2005 @ 11:39 PM EDT
- lemme guess - Authored by: Anonymous on Sunday, July 17 2005 @ 04:45 AM EDT
- lemme guess - Authored by: trs on Saturday, July 16 2005 @ 10:22 PM EDT
- Not in C++ - Authored by: Anonymous on Sunday, July 17 2005 @ 01:40 AM EDT
- "I am employed by The SCO Group, Inc." - Authored by: oldgreybeard on Saturday, July 16 2005 @ 05:09 PM EDT
- Feh, who's this "Brian Kernighan"? - Authored by: Anonymous on Sunday, July 17 2005 @ 04:10 AM EDT
| |
| Authored by: pcoady on Saturday, July 16 2005 @ 03:21 PM EDT |
PJ,
I'd like to feel as comfortable about our jury system as you; however, viewing
the outcomes of the corp. fraud trials in NYC vs. the one in
"hometown"
Birmingham, where the accused 'convinced' a hometown jury that he went
thru 6 CFOs in as many years but still knew nothing of his own corp.'s
finances, doesn't give me confidence.[ Reply to This | # ]
|
| |
| Authored by: sjgibbs on Saturday, July 16 2005 @ 04:10 PM EDT |
Use the preview button and please try to make links clickable. Check the
instructions in red below the comment box.[ Reply to This | # ]
|
| |
| Authored by: kawabago on Saturday, July 16 2005 @ 04:58 PM EDT |
Plain and simple. Just like McBride and all the directors at SCO he is
motivated and ruled solely by greed. It is their creed and it will also be
their undoing.
---
TTFN[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Saturday, July 16 2005 @ 05:12 PM EDT |
I was hoping that would never happen. I am beginning to lose hope a bit but it
seemed to me that the case could be settled as a matter of law and that a jury
would therefore be unnecessary. The Novell ruling suprised me because I was
sure it could have been settled as a matter of law and yet judge K's words
seemed to imply that it would go to a jury.
What are the odds of this case getting to a jury?
[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Saturday, July 16 2005 @ 05:42 PM EDT |
This is mostly irrelevant. He did not put himself forward as a expert witness
and the court did not take him as one: "It is astonishing that SCO has not
offered any competent evidence to create a disputed fact regarding whether IBM
has infringed SCO's alleged copyrights through IBM's Linux activities".
If it comes to trial SCO are going to have produce proper evidence. This was
just to prevent summary judgement (and rejected by the court even for that
purpose, even though summary judgement was not given).
[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Saturday, July 16 2005 @ 06:15 PM EDT |
The Gupta declaration and basically all of SCO's arguements are the flip side of
Davidson's email statement, "The project was a result of SCO's executive
management refusing to believe that it was possible for Linux and much of the
GNU software to have come into existance without *someone* *somewhere* having
copied pieces of proprietary UNIX source code to which SCO owned the
copyright."
10 out of 10 for Davidson calling that correctly and telling Darl so.
The flip side is that they WILL ACCEPT ANY STATEMENT that so much as SEEMS to
IMPLY anything irregular.
PROVING OTHERWISE IS SOMEONE ELSE'S PROBLEM.
This is SCO's "case".
---
Are you a bagel or a mous?[ Reply to This | # ]
|
| |
| Authored by: bmcmahon on Saturday, July 16 2005 @ 06:51 PM EDT |
Paragraph 53 asserts:
With regard to the organization of Linux
SysVIPC and UNIX System V IPC, both
consist of three mechanisms: message
queues, semaphore, and shared
memory. There is no reason for the organization
to be identical other than
the fact that Linux SysVIPC has been copied from
UNIX System V IPC.
So what other mechanisms of IPC are
there, anyway? Shared files, I
suppose, except those are really just a special
case of shared memory.
Challenge: Can anyone come up with an operating
system anywhere that
uses an IPC method that does not fall under one of
these three
headings?
In OpenVMS, for example, IPC is
handled by:
- Shared files (shared memory)
- Common event
flags (semaphor)
- Logical names (shared memory that happens to
be organized a
certain way)
- Mailboxes (message
queue)
- Global sections (shared memory)
- Lock
management system services (the trickiest to map, but
essentially the lock
acts as a semaphor)
This is the entire list from section 3.2
"Communication Between Processes" of
the most recent (January 2005) edition of
the
HP
OpenVMS Programming Concepts Manual.
Let's hope Mr. Gupta doesn't
get his hands on a
VMS
source CD next!
[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Saturday, July 16 2005 @ 07:23 PM EDT |
I think the most interesting thing is the best he can find (for RCU which seems
to have less redaction than the rest) is 5 lines/short-sections of code - in
entirely different places - that do roughly the same thing.
Emphasis
added:
14. UNIX RCU and Linux RCU perform the same five acts in
the same sequence, and UNIX RCU and Linux RCU have the same structure and
organization.
15. Each of the five acts of UNIX RCU -- and of the Linux
RCU -- routine is expressed in one or a few lines of code
At this
point I got tired of retyping, so if I could paraphrase my
understanding:
16. Act #1: Linux and UNIX both contain a single line of
"nearly identical" (i.e. DIFFERENT) code to "allocate a new data
structure of a certain size", i.e. I'm guessing a 1 line call to malloc or
similar
17. - 20. All redacted
21+22. Act #5: Linux RCU and
UNIX RCU, both achieve defered deletion. The code is apparently
DIFFERENT, Linux using a call-back and UNIX using some other mechanism...
but according to Mr Gupta, despite being different, despite using different
mechanisms, because they achieve the same general goal (deferred deletion) they
are somehow "substantially similar".
And here's me thinking that
copyright doesn't protect "methods of operations", "ideas",
etc.
Quatermass
IANAL IMHO etc[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Saturday, July 16 2005 @ 07:35 PM EDT |
| After reading what little remains of this document, I find myself amazed at the
audacity of Mr Gupta and his claims.
Lets jump right into it. Although the
RCU issue has been dealt with by many others, I want to look at some of
his/SCO's specific claims.
14: UNIX RCU and Linux RCU perform the
same five acts in the same sequence, and UNIX RCU and Linux RCU have the same
structure and organization.
I find it amazing that the esteemed
Mr. Gupta would find this an issue. The 'acts' as he calls them, are process
requirements for the functionality of the procedure in question. Any OS
implementing an RCU style process, would require these individual steps in the
same order.
15: Each of the five acts of the UNIX RCU -- and of the
Linux RCU -- routine is expressed in one or a few lines of code.
16: The
first act, "allocating a new data structure of a certain size," is expressed in
UNIX RCU and Linux RCU by a single line of nearly identical code. From a
software programmers perspective, the UNIX RCU expression of the act of
allocating a new data structure has been identically copied into Linux
RCU.
This statement borders on the ludicrous. There are a limited
number of methods for allocating a block of memory, and the most secure or
stable methods are usually the method of choice. By his line of reasoning then,
I myself have multiple levels of SCO IP infringement in my own code, as I often
allocate a data structure in a single line of code. He is implying that the mere
act of reserving a portion of memory for the process is copyright violation,
rather than an industry standard practice.
As a professional software
developer I can equivocably state that the act of allocating a new data
structure in C language will be almost identical to any other single line of
code written in C that performs the same function, even if the allocation has
little or nothing to do with this particular process.
In my opinion, he is
deliberately misleading the court by focusing on standardized methods of
implementation enforced on programmers by the structure of the C language, by
showing the similarity of independant lines of code that would be written in a
similar fashion no matter what the function of the overall process.
This
disengenuous rendering of minutia emphasises the similarities between disparate
entities, while blurring the distinctiveness of the algorithms in question. I
could equally argue the similarities of UNIX RCU data structure allocation and a
game app's data structure allocation, given the chance to look at the C code of
any application on the market.
22: Although the fifth act in Linux
RCU and in UNIX RCU are expressed differently, they both achieve deferred
deletion of old data structure.
Here again, we see the process as
copyright, rather than the expression. At least he's honest enough to point out
that the expression is different, but then lapses into cranial dissociation, and
basically states that it does not matter, as the code achieves the same result,
and thus is a violation of copyright.
23: In my opinion, Linux RCU
is substantailly similar to UNIX RCU, and appears to be derived from UNX
RCU.
24: As represented to me, contributors to Linux had access to UNIX
RCU.
I've noticed these two styles of statement in multiple
places, and while I disagree with his opinion, he is giving his opinion in a
recognized manner, and as a reputed expert in the field. Unfortunately, while I
see many opinions, I see very little fact. Computer Science is a harsh
mistress, and one usually has to give supporting facts for expressed opinion,
where applicable. In this instance, the applicability of fact is in a dominant
role, and he appears to be lacking in this regard.
As far as the statement
'As represented to me', I'm not sure what to make of this. Who
has represented any facts or opinions to him. Is this his
testimony/declaration, or was he just told what to type up? From the previous
Michael Davidson memo, we get the clear impression that SCO executives have
little regard for the 'expert opinion' of others where it does not jive with
their perception of reality at the time.
He uses similar argument of
process when dealing with the IPC code. Using this standard of argument would
make ALL operating systems (Yes, MS and Sun too) an infringement on their
alleged IP. His statement '..there is no reason for the organization to
be identical other than the fact that Linux SysV IPC has been copied from UNIX
System V IPC' is an amazing piece of syntactical ambiguity, where the
opinion begs the question, and relies on the conclusion to be accepted at face
value without clarification of why the organization would be identical across
the board.
Item 49 is not even opinion, but merely conjecture, and has no
place in a legal declaration. 'While working on changing the Linux ULS,
Mr. Lokier learned of the UNIX ULS implementation, may have had access to it,
and may have incorporated the infringing code into Linux.' Mr. Lokier
may have had bacon and eggs for breakfast as well.
SCO has not shown any
evidence of infringement, yet uses the term in it's declarations as a proven
fact, rather than an unsubstantiated allegation, and also uses this as further
weight to other allegations and fantasy ideals. Perhaps I am behind the curve
when it comes to legal matters, but it appears to me that if one is trying to
prove copyright infringement, then it is best to deal in fact, rather than use
the term 'infringement' to lend weight to specious claims, speculation and
obfuscatory tactics.
Then again, SCO has no evidence, and so is in the
position of finding a truly unique and workable 'Wookie Defense'. This, it
would seem, is the true calling of BS&F ( a rather apropos acronym).
[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Saturday, July 16 2005 @ 08:58 PM EDT |
Gupta has already been thoroughly discredited by experts (Dr. Brian Kernighan,
one of the inventors of UNIX, and by Dr. Randall Davis of MIT).
There is no need for people here to re-parse every single line of RCU et al. Any
such analyses fall in the category of "non-expert", just like Gupta.
[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Saturday, July 16 2005 @ 09:01 PM EDT |
| I've done pages 1-5. I'll proceed through 10. [ Reply to This | # ]
|
| |
| Authored by: Anonymous on Saturday, July 16 2005 @ 10:08 PM EDT |
"The claim that Linux 2.6 and their SysV have the same init makes
no allowance for distribution variances."
As noted in another
response, init is not part of the kernel and lives entirely in
userland.
"The slackware that sits in reserve on my media box is
kernel 2.6, and the init on that resembles the old BSD (e.g. Sun OS
4) setup
more than anything."
Actually it has only some resembalance to
BSD, and even more to SYSV Rel 2. BSD had just one single rc file. BSD had no
runlevels, while SYSV had all these runlevels and all the various
bootwait/respawn/etc commands for the inittab file (which BSD didn't even
have).
In fact, though I don't know what it is called now, when the SYSV
style
init package was first added to Linux, that is what it was called...
sysvinit. And there was a BSD style init too, which I don't think even exists
anymore???
My debian systems use S and K symbolic links up to
init.d, and for all I know that may be mandated by POSIX.
If I
remember right, SYSV R2 had the S and K plus a number to determine the order,
but the symlinks were added with R3 (R2 didn't have symlinks at
all).
My general impression is that someone proposes replacing the
init scripts with "something cleaner" about once a month. It sounds like a fun
thing to do, so who knows how the less popular distributions handle
it.
Actually it is a very educational thing to do!
Nearly 20 years ago I converted a SYSV R2 (ATT UNIX) box to the new SYSV R3
style init just to avoid having to deal with two different styles. Needless to
say, since then I have never hesitated to hack rc files... :-) [ Reply to This | # ]
|
| |
| Authored by: Anonymous on Sunday, July 17 2005 @ 04:19 AM EDT |
Now that the cat is out of the bag with respect to what infringement SCO claims,
I'm sure Linus will be all over this like a rash.
I'm not familiar with those areas of the kernel, but I'm sure someone here will
be. Any comments?[ Reply to This | # ]
|
| |
| Authored by: SilverWave on Sunday, July 17 2005 @ 05:46 AM EDT |
July 14, 2005
More Shoes
There's More to the Apple/Intel Deal Than
Even Bob Thought At First
By Robert X. Cringely
Link
http://ww
w.pbs.org/cringely/pulpit/pulpit20050714.html
---
"They [each] put in
one hour of work,
but because they share the end results
they get nine hours... for free"
Firstmonday 98 interview with Linus Torvalds [ Reply to This | # ]
|
|
|
|
|
