In a computer system, a method for increasing the execution speed of virtual machine instructions, the method comprising:
- inputting virtual machine instructions for a function;
- compiling a portion of the function into at least one native machine instruction so that the function includes both virtual and native machine instruction; and
- overwriting a selected virtual machine instruction of the function with a new virtual machine instruction, the new virtual machine instruction specifying execution of at least one native machine instruction.
You think that sounds like an idea worthy of a patent? Probably depends if you're a programmer or a patent lawyer.
I'll leave you to judge the merits of the other 46 claims, the next of which are:
Sun bought several of these, especially the developers of Strongtalk, a compiled, statically typed dialect of Smalltalk. So if there is any prior art on this, chances are that all the related patents are now in Oracle's portfolio.
A patent isn't an idea. It's the specific method for implementing an idea. If there is a way to get the same end result using a completely different method, the patent doesn't apply.
A "method" which is not being acted out physically, but instead is in someone's head or written down on paper is a type of idea. So some ideas are patentable, others aren't. The mantra that ideas aren't patentable presumes some other category of thought completely apart from ideas. Besides, what could be more purely and precisely methodical than machine-readable source code, in this case Smalltalk source code?
How does that relate to my comment that the patent in question sounds like it could have come out of Sun's work on optimizing Smalltalk instead of their later Java implementation? (Note that even if Sun could have worked around their own patents, they would have had no reason to bother.)
In a computer system, a method for increasing the execution speed of virtual machine instructions, the method comprising:
- inputting virtual machine instructions for a function;
- compiling a portion of the function into at least one native machine instruction so that the function includes both virtual and native machine instruction; and
- overwriting a selected virtual machine instruction of the function with a new virtual machine instruction, the new virtual machine instruction specifying execution of at least one native machine instruction.
You think that sounds like an idea worthy of a patent? Probably depends if you're a programmer or a patent lawyer.
I'll leave you to judge the merits of the other 46 claims, the next of which are:
2. Claim 1 in a box
3. Claim 1 with a fox