|
Authored by: Anonymous on Saturday, May 12 2012 @ 11:17 AM EDT |
Message sends use a selector (a symbolic reference). You can send any
selector to any object, and it will look up the appropriate method
dynamically, using the selector. Objective C borrowed this mechanism
from Smalltalk.
But Java uses it for dynamic binding, not for dynamic dispatch. Java VMs
usually implement method invocation using vtables (the same as C++),
they just build the vtables at runtime instead of compile time.
In Java the symbols are generally "resolved on first use". (Different
from
the Smalltalk message sends, which would be "resolved on every use").[ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Saturday, May 12 2012 @ 05:16 PM EDT |
Hard to tell what the full implications will be, but honestly I'm pretty
encouraged that the judge actually spontaneously suggested a simple optimization
method; that's a definite indicator of a pretty critical base level of 'getting
it'.
Out here those with real understanding of computer science are screaming about
the true obviousness of purportedly novel solutions. Most people just don't get
what's being argued over well enough to judge, so they have to look for some
kind of objective criterion (like whether similar language to all the claim
elements exist in prior patents). Being able to spontaneously suggest an
optimization shows exactly the right kind of mindset for being able to recognize
the possibility of a solution being obvious once presented with the appropriate
problem.
This part of the case may well turn on the ability to establish a critical
meaning difference that destroys the applicability of the claim language, but it
certainly doesn't hurt to have it recognized that the space the claim has carved
out to claim novelty is fairly arbitrary and such novelty is probably
questionable.[ Reply to This | Parent | # ]
|
|
|
|
|