|
|
| Maciaszek Trial Testimony in SCO/Novell Contradicted - Updated: Wilson Confirmed |
 |
|
Monday, March 23 2009 @ 10:11 PM EDT
|
I found something eye-poppingly interesting. Do you remember at the SCO v. Novell trial, where John Maciaszek testified that there was never a charge for earlier versions of Unix/UnixWare? That going all the way back to AT&T days, the earlier products were thrown in free? I have gotten hold of a contract where it seems to me that the licensee was charged for earlier products. Yes. $400 a copy.
Update: And now I find on page two confirmation of Otis Wilson's testimony in SCO v. IBM that modifications were never claimed by AT&T.
It's an agreement between AT&T and the University of New South Wales, dated May 23, 1986. I have it in paper form, so I've scanned in the cover letter and the list of allowed software to show you. First, here's the cover letter [PDF].
Notice that it says that notwithstanding Section 4 of the Schedule to the contract, AT&T agreed that the University could also use "those UNIX operating systems delivered to you under prior educational software agreements". So they got *special permission* even to use older versions of the software they had already purchased. It's conceivable this letter was sent to correct a mistake, or clarify some confusion.
So now let's take a look at Section 4 [PDF].
Here's the eye-popping part. Under the subhead "Other Software" it lists products that could be used in the US and those that could be exported. Then it says: Versions of such products, except those marked with an asterisk (*), are available from AT&T for various types of CPUs at $400 a copy. Yet the testimony at trial went like this:
Q. Now, when you licensed UnixWare and at the time
that you were Novell, did Novell grant the customer any
license to any older versions of the System V products?
A. Well, the standard practice going back to AT&T days
was to grant the right to use prior products as part of the
new products.
Q. And you said that goes back to AT&T days?
A. Oh, yeah. Started well before I joined the
organization. Go back and look at the SVR1.1 prior, and there
were always prior products listed there....
Q. Okay. Did you have personal knowledge of that
licensing of prior products at the time that you joined UNIX
Systems Labs?
A. Oh, yes. I mean, I was involved in creating the
4.2 license, so I understand very well what the situation was.
Q. Were the customers who were given those prior
products asked to pay anything extra for the prior products?
A. Absolutely not.
Q. And is that a consistent practice throughout the
time that you have been at Novell and both its predecessor and
successor companies?
A. Absolutely. I can't remember or recall any
occasion where there would be any even thought given to charge
them for that.
Can you match those words up with the contract I just found? Me neither.
On page two, we find the grant of rights [PDF] section. Remember how SCO tried to claim that Otis Wilson's testimony in 2003 contradicted what he said in the BSDI case in 1992 on the subject of modifications and who they belonged to? This is, of course, an educational license. And note the language: AT&T-IS grants to LICENSEE a personal, nontransferable and nonexclusive right to use in the AUTHORIZED COUNTRY each SOFTWARE PRODUCT identified in the one or more Supplements hereto, solely for educational use and solely on or in conjunction with DESIGNATED CPUs for each SOFTWARE PRODUCT. Such right to use includes the right to modify such SOFTWARE PRODUCT and to prepare derivative works based on such SOFTWARE PRODUCT, provided that any such modification or derivative work that contains any part of a SOFTWARE PRODUCT subject to this Agreement is treated hereunder the same as such SOFTWARE PRODUCT. AT&T-IS claims no ownership interest in any portion of such modifications or derivative work that is not part of a SOFTWARE PRODUCT.
1. Educational use is limited to uses directly related to teaching and degree-granting programs and uses in noncommercial research by students and faculty members, including any uses made in connection with the development of enhancements or modifications to SOFTWARE PRODUCTS. Such uses are permitted only provided that i) neither the results of such research nor any enhancements or modifications so developed is intended primarily for the benefit of a third party, ii) such results, enhancements and modifications all to the extent that they do not include any portion of SOFTWARE PRODUCTS are made available to anyone including AT&T-IS and its corporate affiliates without restriction on use, copying or further distribution, notwithstanding any proprietary rights (such as copyright or patent right) that could be asserted by LICENSEE, its employees, students or faculty members and iii) any copy of any such result, enhancement or modification furnished by LICENSEE is furnished for no more than the cost of reproduction and shipping. Any such copy that includes any portion of a SOFTWARE PRODUCT shall be subject to the provisions of Section 7.05(b).
And that is exactly what Otis Wilson said. And a bit more. How do you like the part about sharing modifications freely, the "without restriction on use, copying or further distribution" language? Evidently in the early days, Unix was a lot more like the GPL, at least as far as modifications were concerned. And AT&T specifically disclaims any ownership interest in modifications, reserving its rights only to its own software.
Update 2: For contrast here's the language in the 1987 DaimlerChrysler license:
6. Other Software
The products listed below may be used in' the United States on DESIGNATED CPUs for UNIX System V? Release 3.0 as if they were that product. The products may also he used outside the United States on DESIGNATED CPUs for UNIX System V, Release 3.0, International Edition as if they were that product. Only those products marked with an "@" may be shipped outside the United States by AT&T.
All prior releases and versions of UNIX System V
@All prior releases and versions of UNIX System V, International Versions
UNIX System III
UNIX 32V Time-Sharing System, Version 1.0
UNIX Time-Sharing System, Seventh Edition
UNIX Time-Sharing System, Sixth Edition
UNIX :Programmer's Workbench System, Edition 1.0
UNIX Mini Time-Sharing System, Version 6
Same time period. No $400. All prior releases specifically allowed.
|
|
|
|
| Authored by: alisonken1 on Monday, March 23 2009 @ 10:23 PM EDT |
If you find any, post them in this thread
---
- Ken -
import std_disclaimer.py
Registered Linux user^W^WJohn Doe #296561
Slackin' since 1993
http://www.slackware.com
[ Reply to This | # ]
|
| |
| Authored by: alisonken1 on Monday, March 23 2009 @ 10:25 PM EDT |
Headline threads here.
Clickies post in HTML mode (see red notes below your entry box)
---
- Ken -
import std_disclaimer.py
Registered Linux user^W^WJohn Doe #296561
Slackin' since 1993
http://www.slackware.com
[ Reply to This | # ]
|
| |
| Authored by: alisonken1 on Monday, March 23 2009 @ 10:26 PM EDT |
Post 'em as you think of them.
Remember clickies as HTML mode. See notes below the entry box.
---
- Ken -
import std_disclaimer.py
Registered Linux user^W^WJohn Doe #296561
Slackin' since 1993
http://www.slackware.com
[ Reply to This | # ]
|
| |
| Authored by: Sparhawk on Monday, March 23 2009 @ 10:38 PM EDT |
The only problem I see as that the items marked with a * are not international
editions, and therefore not applicable to the sale (in Australia).
---
If Bill Gates had a cent for every time Windows crashed... Oh wait, he does.[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Monday, March 23 2009 @ 11:16 PM EDT |
SO????
What else did he forget? Everything else. Is this guy dreaming...? Should his
entire testimony be questioned now that this evidence has surfaced that he
obviously is not the "expert" that we were led to believe he was?
[ Reply to This | # ]
|
| |
| Authored by: IMANAL_TOO on Monday, March 23 2009 @ 11:32 PM EDT |
I wonder if the awards to Novell will need to be be recalculated?
.
---
______
IMANAL
.[ Reply to This | # ]
|
| |
| Authored by: crs17 on Tuesday, March 24 2009 @ 12:11 AM EDT |
I don't see this as such a big deal. As I read his testimony, he claims
personal knowledge on SysV version 4.2 and says that going back to 1.1, things
were okay. The cover letter in the pdf above refers to version 2.
As I interpret his testimony, he never claimed to have made a detailed study of
all the previous versions. He knew the current state (as of 4.2), probably had
read the original 1.1 licenses, but could very plausibly never have studied each
and every license in between, and missed this 2.0 license.
So he was misinformed but he never claimed authoritative knowledge on every
version. [ Reply to This | # ]
|
| |
| Authored by: Anonymous on Tuesday, March 24 2009 @ 12:45 AM EDT |
I seem to remember from the commercial source license in
the early 1980's that $400 was the standard charge for
making and shipping a tape containing the specific software
and was not an additional license fee.[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Tuesday, March 24 2009 @ 02:47 AM EDT |
If it was $400/cpu or $400/license, that would look like a fee, but $400/copy --
well, that sounds like the cost of the copy. Those 9-track tapes were expensive
-- I paid costs of about that for lots of "free" data from the
government in that era.
This idea that data can be actually free, as it is today, just doesn't have all
that long a history.[ Reply to This | # ]
|
| |
| Authored by: barbacana on Tuesday, March 24 2009 @ 03:06 AM EDT |
The article makes this sound like a big deal, verging on perjury, but it
isn't.
First off, the document says the software is "available" for
$400/copy. That doesn't mean that anyone actually paid
$400.
Second: Maciaszek was almost certainly telling the truth when
he testified "I can't remember or recall any
occasion where there would be
any even thought given to charge
them for that." What he says was always the
case, was certainly the normal case. Finding a few instances where it was not
true does not seem very important.
And finally, as a few people have
pointed out, the wording of the document - "$400 per copy", not
"$400 per license" - suggests that this is a copying/media/shipping fee, not
a license fee. Back in the 1980s, even the Free Software Foundation used to
charge you over $100 to put free software onto a magnetic tape and ship it
to you. [ Reply to This | # ]
|
| |
| Authored by: Steve Martin on Tuesday, March 24 2009 @ 07:01 AM EDT |
Everyone here is making comments over the Maciaszek testimony, and nobody
seems interested in the other part of this article, namely the Wilson testimony
re: ownership claims on derivatives. I personally think this is the bigger
issue.
As we know, The SCO Group's entire case against IBM hinges on
their interpretation of IBM's UNIX Software License Agreement, specifically
§ 2.01, and the confidentiality provisions therein. TSG has claimed that
the phrase "provided the resulting materials are treated hereunder as part of
the original SOFTWARE PRODUCT" required IBM to keep things like RCU confidential
rather than contributing them to Linux.
Of course, we all also know that
the AT&T "echo" newsletter made AT&T's position pretty clear on this
issue, but it was not part of the actual language of the Software Agreement, so
legally it might not have carried weight in this litigation.
Now, here
is an actual license document from AT&T to one of its licensees, with
language included specifically disavowing any such claim to non-System-V
parts of the licensees' derivatives. Granted that it is not part of the IBM
agreement, but it does demonstrate AT&T's intent. And since the IBM contract
is "reasonably susceptible" to that interpretation (sound familiar?), it seems
reasonable that this document would be admissible as extrinsic evidence to
support IBM's interpretation of the contract. It is therefore something that
Judge Kimball might be allowed to take judicial notice of, if it ever comes
up.
In short, IANAL, but this document would seem to put the final nail
in the coffin of TSG's outlandish reinterpretation of IBM's (24-year-old!)
contract with AT&T.
---
"When I say something, I put my name next
to it." -- Isaac Jaffe, "Sports Night" [ Reply to This | # ]
|
| |
| Authored by: Anonymous on Tuesday, March 24 2009 @ 07:26 AM EDT |
PJ, according to my reading of the section 4 PDF that you posted, The $400.00
dollars is the price to obtain a copy of the media for a particular version. It
doesn't say that they can only install such a copy onto one CPU only, nor does
it say anything about the price per CPU of the most current version.
The wording, in ALL CAPS, about "DESIGNATED CPUs", seems to be saying
that they have paid some amount per CPU for a license to run the latest version
on soem number of DESIGNATED CPUs, and that they could run any of the earlier
versions on the list on those DESIGNATED CPUs if they wished, or the latest
version, but that if they needed a copy of the media to install an earlier
version from, the media would cost $400 per copy.
Of course, I would have to show the rest of the contract to a lawyer to be sure,
but I don't think that it's saying that they would have to pay an additional
$400 per CPU to run an older version on a CPU that they had already paid to be
able to run the latest version, only thta if they didn't have a copy of the
older version, it would cost them $400 for a copy, and that they couold instal
lthat copy onto as many DESIGNATED CPUs as they wished.[ Reply to This | # ]
|
| |
| Authored by: JamesK on Tuesday, March 24 2009 @ 07:39 AM EDT |
"without restriction on use, copying or further distribution"
That might put a bit of a dent in SCOX's claims against IBM.
Given so much of SCOX's case goes against the evidence, shouldn't someone be
looking at fraud or other criminal behaviour on the part of SCOX?
---
There are 10 kinds of people in the world, those who understand binary and those
who don't.
[ Reply to This | # ]
|
| |
| Authored by: greybeard on Tuesday, March 24 2009 @ 08:36 AM EDT |
Aside from the various differences between educational and commercial licenses,
part of the issue with the various versions was that the commercial contracts (I
don't know about the educational ones) held that one could share modifications
with other licensees whose license was at the same or higher version number.
This was a real pain in the butt when sub-contracting development work. It also
tended to help make people like Oracle rich in that they demanded
recertification (that could cost as much as $1,000,000) for each and every
version.
There were just a whole lot of tidal forces pulling the old UNIXes apart and it
was difficult to reconcile them under the prevailing notions of how to build and
market a software product.
Re: John's testimony, there are several points of ambiguity.Is the 4.2 to which
he refers meant to be System V. Rel4.2 or is he alluding to the first versions
of System V that incorporated the 4.2bsd code?
I concur with the other posters that the $400 sounds like a media (shipping and
handling) charge. It is less than we charged folks who bought a source license
from us for our SV-based products. And it is not much when you understood that
the license itself was $75,000 and up. My experience was that ATT did not much
care what folks did with the older versions so long as they did not try to
"update" them by incorporating the modifications made to the latest
releases.
$0.02 from an old product manager.
---
-greybeard-[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Tuesday, March 24 2009 @ 08:40 AM EDT |
How do you like the part about sharing modifications freely, the
"without restriction on use, copying or further distribution" language?
Evidently in the early days, Unix was a lot more like the GPL, at least as far
as modifications were concerned. And AT&T specifically disclaims any
ownership interest in modifications, reserving its rights only to its own
software.
PJ - Please read the section on modifications again (my
emphasis added)
such results, enhancements and modifications all
to the extent that they do not include any portion of SOFTWARE PRODUCTS
are made available to anyone including AT&T-IS and its corporate
affiliates without restriction on use, copying or further distribution
This is not like the GPL. It is stating the obvious, which is
that you own the code that you write, which is exactly what IBM has been
arguing, that they can do with what they please with things that they create, as
long as the original UNIX code from ATT/Novell/SCO isn't included.
The
GPL says that the original code can be included with the modifications. This
contract says specifically the opposite, that the original code cannot be
redistributed, whether it contains your modifications or not.[ Reply to This | # ]
|
| |
| Authored by: sk43 on Tuesday, March 24 2009 @ 10:56 AM EDT |
The language about ownership under "grant of rights" in Section 2.01
is the same as that in the the DaimlerChrysler agreement and discussed here over
5 years ago:
http://www.groklaw.net/article.php?story=20040318012131886[ Reply to This | # ]
|
- Old News? - Authored by: PJ on Tuesday, March 24 2009 @ 01:46 PM EDT
- Old News? - Authored by: Anonymous on Tuesday, March 24 2009 @ 02:52 PM EDT
- Old News? - Authored by: Anonymous on Tuesday, March 24 2009 @ 04:22 PM EDT
- Old News? - Authored by: Anonymous on Wednesday, March 25 2009 @ 01:10 AM EDT
- Old News? - Authored by: sk43 on Tuesday, March 24 2009 @ 10:09 PM EDT
- Old News? - Authored by: PJ on Wednesday, March 25 2009 @ 01:37 AM EDT
| |
| Authored by: Anonymous on Tuesday, March 24 2009 @ 01:51 PM EDT |
Pretty much all I read was focused on the possibility the $400 was a cost
associated with producing the media. Here's an alternative:
The $400
cost was for those licensees who had not already previously licensed the
product.
The Diamler license, as an example, identified the previous
versions as no cost. Could this be an example of an "already
licensee"?
On the other hand, the license for the University of New South
Wales identified a cost of $400. Could this be an example of a "first-time
licensee"?
Perhaps the clue to the cost is in the specific language of
the New South Wales University license:
Versions of such products,
except those marked with an asterisk (*), are available from AT&T for
various types of CPUs at $400 a copy.
Bolding mine. It is
a University after all and one would expect a University to have multiple
platforms available - such as Intel vs. IBM Mainframe vs. SPARC vs.
etc.
So perhaps the simple answer is:
You purchase a primary
license at a rate of per-cpu cost. To cover your alternate CPU's, we'll provide
a version of the software compiled for that platform at a cost of
$400.
Even if the University got the source code, I could see
it being beneficial to pay for a "working copy for SPARC" just in case they had
to modify anything in order to get it to work on the SPARC platform vs their
"main" platform. This would allow them to be up-and-running immediately while
they worked with the code.
So perhaps the conclusion the cost is mainly
for the media is correct.
RAS[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Tuesday, March 24 2009 @ 01:53 PM EDT |
Ok, so he lied. SCO has been quite consistent there.
But what does it mean?
SCO/Novell is final. My understanding is that means unless Novell contested it
at the time, it now stands.
So is this something Novell can re-open? What would they gain? Or are we
merely looking ahead to IBM/SCO? Does it matter if IBM points it out and the
record contradicts SCO/Novell?
Always good to have the truth, but what relevance does it have now?[ Reply to This | # ]
|
| |
| Authored by: mikeprotts on Tuesday, March 24 2009 @ 04:47 PM EDT |
"Evidently in the early days, Unix was a lot more like the GPL, at least as
far as modifications were concerned."
I would have though this was more BSD than GPL - a shame really because GPL is
much cleaner in the long run.
Cheers
Mike[ Reply to This | # ]
|
- GPL or BSD - Authored by: Anonymous on Wednesday, March 25 2009 @ 05:35 AM EDT
| |
| Authored by: leopardi on Tuesday, March 24 2009 @ 06:25 PM EDT |
| UNSW and various people at UNSW were involved in
development of various
flavours of UNIX from very early
on.
The work of John Lions and others
at UNSW prompted an
exchange of people between UNSW and the UNIX group at
Bell
Labs.
In a Salon
article on Lions' book, Dennis Ritchie explained:
"Probably the most
important contribution John made was to
start, at UNSW and indirectly at Sydney
Uni, a very strong
group of Unix people, many of whom have visited or stayed
here, and whom we have often visited".
The exchange of personnel
between UNSW and Bell Labs
included John Lions himself, who took sabbaticals at
Bell
Labs,
including one in 1978,
as recalled in an interview in The Amateur Computerist
(p. 1,
5-7).
In at
article
in the Melbourne Age, various ex-UNSW people
claim that their work on BSD
found its way into later
versions of UNIX.
Quote from the article:
' Steve
Jenkin, a veteran Unix sysadmin who is also
attending the AUUG conference and
was a student at the
UNSW in 1977, said a lot of the code which went into 2BSD
had come from the UNSW. "About 60 percent of the code is
from UNSW," he said.
"But when (the University of
California at) Berkeley started coordinating all
efforts
at development, the copyright notices were taken out and
replaced by
their notices." '
[ Reply to This | # ]
|
| |
| Authored by: Anonymous on Wednesday, March 25 2009 @ 07:48 AM EDT |
How curious to see this "psychological tic" in Otis Wilson:
.................................................
"Q. Were the customers who were given those prior
products asked to pay anything extra for the prior products?
A. Absolutely not.
Q. And is that a consistent practice throughout the
time that you have been at Novell and both its predecessor and
successor companies?
A. Absolutely. I can't remember or recall any
occasion where there would be any even thought given to charge
them for that.
.....................................................
This is not the first time that I've seen someone who is self-consciously lying,
and while under oath, do EXACTLY the same thing, to wit, use the word
"absolutely" each time they are about to tell a lie.
Poker players call that a "tell", I think.
LEXLAW
[ Reply to This | # ]
|
|
|
|
|
