|
Authored by: Anonymous on Tuesday, May 08 2012 @ 11:37 PM EDT |
I don't disagree with your description. But I think you misunderstood what
Oracle's lawyer is aiming at.
The source code he's talking about is the source code of the Dalvik VM, and its
related tools.
If that source code (the Dalvik/tool source) defines a process that operates as
Oracle's claimed 'invention' then that would be evidence of infringement.
So all Oracle has to do is put up some slides of the accused 'C'/ 'C++'(?)
source code and it will become immediately obvious to the jury what the
operational flow and consequences of that code are.
Maybe, for the benefit of the judge (who, being a lawyer, does not have the
jury's expertise in C/C++), Oracle could throw in some UML diagrams. But that
really is overdoing it. Any normal layman such as a juror has an intuitive
grasp of C/C++ source code, its so transparent and understandable.
Slam Dunk.
Now if the lawyer were actually talking about the source code for an
>application< to be compiled into bytecode to run on the Dalvik VM, you'd
be right; he's looney-tunes.
But remember PJ is right, we are now in dimension n+1 of an n-dimension
universe.
Not a lawyer
I was patented, but I've expired
JG
[ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Tuesday, May 08 2012 @ 11:47 PM EDT |
Not hard to put a table of variables into your machine code.
Start the program with generic code, then the tables, then
your program.
The generic code takes its current address and a known
offset to figure out the start of the variable block, stores
it, and then uses the known length of that block to jump
ahead to the actual program. Then you have known offsets
within that block (the compiler determines arrangement,
handles the offsetting, etc.) to address the variables, etc.
Not hard. Not something I want to ever do. Doesn't take care
of variable-length arrays and such higher data structures,
but I don't know when those started to be supported in
programs anyways.[ Reply to This | Parent | # ]
|
|
|
|
|