|
Authored by: Anonymous on Thursday, May 17 2012 @ 09:09 PM EDT |
If you can show that someone's paid expert gave provably false testimony, can
the verdict be thrown out?[ Reply to This | Parent | # ]
|
|
Authored by: bugstomper on Thursday, May 17 2012 @ 10:29 PM EDT |
I disagree that Dr. Mitchell told lies about "symbolic reference". His
definition is closer to the one that most programmers would come up with if
someone talked about machine code instructions that contain a symbolic
reference. It would most likely mean that however instructions on that machine
are used to reference data, those instructions use that method to reference a
name that is to be used to look up the data that corresponds to that name. Dr.
August relied on the Court's construction that made the definition of
"symbolic reference" depend on "dynamic", something that has
nothing to do with the computer science typical usage of "symbolic
reference".
Google's lawyer, on the other hand, did something sleazier with that, and did
get Dr. August to acquiesce, I think, at least a few times. Google made an issue
of the literal interpretation of "instruction contains a symbolic
reference" as precluding a numeric reference to a struct that contains a
numeric reference to a name in string data. Dr. August did not argue with that
being implied with the series of questions "Is that a numeric reference? Is
that a numeric reference" ... until finally getting to the symbol. But Dr.
August never actually said "therefore that is not a symbolic reference in
the instruction" instead going for "it is not dynamic therefore by the
Court's construction it can't be called a symbolic reference".
On the other hand, Oracle did something at least as sleazy if not much worse
with the word "dynamic". They argued that "instruction containing
a symbolic reference" should be construed according to the standard
computer science meaning that Dr. Mitchell ascribed to it, instead of the
literal wording of the claim. But when it came to "dynamic" they want
to stretch the usage in the patent to include something that it clearly did not
mean by treating it as a plain English word and not a technical computer science
word.
That's where I dislike Mitchell for his testimony. Dr. August made an attempt to
not parrot Google's lawyer's attempt to paint the iget instruction as not having
a symbolic reference because it was numeric. Instead he relied on the strict
claim construction with the "dynamic" limitation, which at least is
literally honest. Dr. Mitchell joined right in on the misuse of
"dynamic" and in that he was intellectually dishonest. And he was
equally intellectually dishonest trying to characterize pattern recognition as a
kind of simulation. I don't believe he would ever have come to that conclusion
if he had come across these programs without being paid a huge amount by the
side that has a vested interest in his conclusions coming out the way they
did.
[ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Thursday, May 17 2012 @ 10:46 PM EDT |
I don't know if he lied, but I do think he betrayed his status as an expert.
It seem to me that he uncritically accepted a lawyer's interpretation of the
patent's claim construction. I think he disregarded the normal interpretation
that a person skilled in the art would make.
Yes, philosophically, numbers are symbols. Twelve, 12, XII, and x0C are all
symbols representing the same number. However, he was testifying as an expert in
the art - not as a philosopher. An expert in the art should distinguish between
numbers being used as symbols (e.g., dates or ID numbers) and numbers in binary
form used as indices.
I think that to a person skilled in the art, looking up a symbol in a table
implies that it is a symbol table. It implies that the symbols appear in the
table and that you have to search the table to find the symbol to retrieve the
corresponding value. The fact that the claim didn't explicitly say 'symbol
table' should be irrelevant.
Symbols can be substituted without changing anything else. For example, the
judge asked a witness to use "Y" instead of "FUN". Since the
symbol "Y" wasn't already in use, all that had to be changed were the
instances of the symbol FUN, including both in the code and in the symbol table.
To change the index 12 to (previously unused) 14 you also have to move the
contents of location 12 to location 14 and change the status of location 12 to
unused and location 14 to used.
He represented himself as an expert witness qualified in the art. IMHO he didn't
testify as such.[ Reply to This | Parent | # ]
|
|
|
|
|