|
Authored by: Anonymous on Friday, May 11 2012 @ 06:55 PM EDT |
I'm still wondering what that patent does, that isn't part of a decent course
in programming vms.[ Reply to This | Parent | # ]
|
|
Authored by: YurtGuppy on Friday, May 11 2012 @ 07:25 PM EDT |
which, being interpreted means: (what a stupid question!)
---
a small fish in an even smaller pond[ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Friday, May 11 2012 @ 07:26 PM EDT |
The whole difference of the symbolic referencing is what time does the process
occur. Java does it at "runtime" where runtime is defined to be when
the program is actually running.
Android does it at compile time and possibly while the program is loading.
Oracle is trying to expand the idea of "runtime" to be when the phone
itself is running. Notice all the talk of when the device is "powered
on". Google spent time emphasizing the technical meaning of runtime as used
by programmers. You can see this emphasis in the ReReDirect of Mr. McFadden.
That seems to me to be the critical difference between the two.
P.S. I have not read the patents and I'm not a lawyer. I am a
mathematics/compsci student. [ Reply to This | Parent | # ]
|
|
Authored by: bugstomper on Friday, May 11 2012 @ 07:26 PM EDT |
I was worried about two points in Oracle's case. 1) documentation describing
dexopt optimization that talks about some things that can only be done at
runtime; and 2) the documentation about dexopt as a "back door" into
the Dalvik VM when it is run as part of installation on an Android device.
Andrew McFadden blew away those two arguments completely in his testimony that
(1) is a paragraph describing what possible optimizations dexopt _doesn't_ do
because it doesn't do any dynamic optimizations, and (2) uses the term
"back door" to mean that dexopt accomplishes some things that the VM
does by using some common code without ever running the VM - Hence it is coming
in the "back door" by _not_ running the VM, while the VM running is
what he thinks of as the "front door".
[ Reply to This | Parent | # ]
|
- I agree, but... - Authored by: Anonymous on Friday, May 11 2012 @ 08:30 PM EDT
- I agree, but... - Authored by: Anonymous on Saturday, May 12 2012 @ 01:17 AM EDT
|
Authored by: Anonymous on Saturday, May 12 2012 @ 12:46 AM EDT |
David August was excellent in his testimony - clear, to the point and just
destroyed Oracle's case[ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Saturday, May 12 2012 @ 10:37 PM EDT |
From the court's claim construction order (according to
todays
transcript):
a symbolic reference is a reference that
identifies data by a name other than the numeric memory
location of the data,
and that is resolved dynamically
rather than statically
A
numerical index into a table of field definitions is
arguably not the numeric
memory location of the data itself,
thus under the courts definition it may be
a symbolic
reference, even though it is numeric and not
textual.
Furthermore the existence of the non-quickened
instruction
form in the dex "bytecode language" may indicate
that the Dalvik VM can execute
this directly in case it
isn't optimized to the offset form by dexopt, although
this
was not discussed in court today. If true, that would make
the numeric
field index reference a symbolic reference under
the courts
definition.
That little word game might ruin this for Google.
[ Reply to This | Parent | # ]
|
|
|
|
|