|
Authored by: Ian Al on Sunday, April 29 2012 @ 05:58 AM EDT |
I read your comment about the requirement for registration after I wrote this.
Apologies. If I have you right, only proven damages are awarded for unregistered
copyrights. If Oracle register the asserted copyrights now, there is no
retroactive effect.
If you have problems with the other stuff, please say.
---
Regards
Ian Al
Software Patents: It's the disclosed functions in the patent, stupid![ Reply to This | Parent | # ]
|
|
Authored by: Ian Al on Sunday, April 29 2012 @ 11:41 AM EDT |
Ever since your comment I have had if - then declarations going around in
my head like dodgems.
Judge: You said you reviewed the 37 along
with others. Did you review them to
see if there was an affirmative statement
of ownership by Sun, or did you review
them to see if there an absence by
others?
scav: It is not allowed to have two classes
defined in the same
source file (a simplification but basically true). A
package
containing more than one class will be fixed in more than one
source
file, and comprise more than one .class file when
compiled.
A
Java package is not a single medium. It is not a document with creative
expression fixed in it. Each class file is a document and is copyright in its
own right. If I've got it right (finally?) both a collected work and a
compilation has to be a single work fixed in media and cannot be a fluid
'library' of 'documents' with their own individual copyrights. The whole work
has to be a single work, although it might be a registered CD with a lot of
individual files on it. I would have thought that such an aggregation would have
to be registered if it was not to be just a fluid concept in the aggregator's
mind.
The Oracle example of copied SSO was a tree diagram of the
directory structure and file names of some of the .class files in the
“java.nio.channels” package. That was from the actual directory in which the
library code is kept rather than a document specifying what the API
does.
For copyright infringement of the SSO it would have to be
possible to own the copyright on a number of files of the same name in a
directory structure of the same name as both the protected work and in the
infringing directory.
We know that the file contents are different
because the judge, himself, established that by questioning the expert witness
for Oracle, Dr. Mitchell.
It's the same for the Java SE API
Specification. The only thing we have seen is the odd document from a webserver
directory of html files which we are told by Oracle is the Specification. Is it
sufficient to meet the copyright law requirements as a single work covered by a
copyright?
If it is not individually registered, how is it not,
legally, a movable feast of files rather than a work with creative expression
fixated in it?
Since some entire packages are owned by third parties,
where is the evidence that the documentation in the Specification, created by
running javadoc on the authors' individual class code files, is creative
expression belonging to Oracle? That goes back to the question put by the judge.
The creative expression belongs to the author of the individual class file code.
How many of those carry the Oracle copyright mark?
I was under the
impression that a copyright complaint goes along the lines of 'you copied my
copyright work without a licence and this is the work you copied'. What is the
copyright work that Oracle assert that Google copied? Is it files in a directory
structure on the internet purporting to be an API Specification work?
Is it the thousands of .class files in a directory of 166 package
directories? That's where Oracle's expert 'found' the copied SSO. The filenames
are not mentioned on the 'Specification' website documents except as
declarations.
Only 37 directories for Oracle and Google share the same
directory and file names. The file contents are conceded to be different. What
is the whole work? Is it the 37 directory and filenames or is it the 166? What
did Oracle actually register?
Judge: You're not explaining
why you can't get Mr. Gonsalves here. You're avoiding
that.
PJ:It affects damages. If there is a copyright
but it's not registered, then you can only
sue for actual damages, and that
would be
mighty hard to prove.
But there's another issue --Does
Oracle have
a copyright at all in the items on its list?
Or do others? The
problem is they didn't list
the items individually, and Java the whole
thing
is copyrighted to Sun and IBM and others.
So which party did
what?
There is a further issue: a list of what and where is the
list? Is it a list of copyright documents in a compilation? Where is the
evidence of the compilation existing at the time that Google are alleged to have
copied it? Is the list a whole copyright work? Is it registered? I think the
phrase 'the list of APIs' is now seen to be meaningless and irrelevant in the
context of the case.
One final question: what happens when the dodgems
and the music stops?brbr --- Regards
Ian Al
Software Patents: It's the disclosed functions in the patent, stupid! [ Reply to This | Parent | # ]
|
|
|
|
|