5 ms·

> Interpreting functions utilizing a hybrid of virtual and native machine instructions Are they serious? Unless I'm missing something this is a fantastic joke

> Interpreting functions utilizing a hybrid of virtual and native machine instructions Are they serious? Unless I'm missing something this is a fantastic joke, JITs far predate Java.

tptacek 16y ago
Careful. The name of a title and its actual claims are two very different things.
- patrickaljord 16y ago
  Careful. Google's implementation and the patent itself are two very different things.
- shasta 16y ago
  Well, here's claim 1: 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
  - btilly 16y ago
    It sounds like the sort of stuff that Smalltalk implementations were doing before Java existed. See http://en.wikipedia.org/wiki/Just-in-time_compilation#History http://en.wikipedia.org/wiki/Just-in-time_compilation#Histor... for more. (I wonder whether Sun's patents were filed for Self?)
    - fhars 16y ago
      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.
    - gaius 16y ago
      sigh 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.
      - btilly 16y ago
        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.)
      - koenigdavidmj 16y ago
        Yes! And the first few claims generally paint way over the lines. It's in the later claims that you talk about the unique stuff that you did.
      - kj12345 16y ago
        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?
  - Confusion 16y ago
    Again, as tptacek already said: careful. The summary isn't the actual claim either. The devil is in the details.
    - shasta 16y ago
      The above is the actual first claim from the patent. Of course it's subject to interpretation in context, but that is the text of the first claim.
nl 16y ago
I'm guessing, but I suspect working around this & other patents was why a JIT took so long to ship in Android.
- bmastenbrook 16y ago
  You can't write a good JIT overnight. You can't adapt an existing JIT for a JVM-like VM to Dalvik either. I don't think working around patents had much to do with it, especially since they got sued anyway.
js4all 16y ago
Careful. Hotspot compiling was introduced with the JVM.
- bad_user 16y ago
  There is prior art in Smaltalk and Self ... the Hotspot compiler itself is an old Smaltalk implementation that Sun bought, open-sourced recently by SUN btw ... http://www.strongtalk.org/ http://www.strongtalk.org/
  - loewenskind 16y ago
    I'm pretty sure the prior art was just with Self. Smalltalk had polymorphic cache tables, but it was Self that was doing really aggressive in-lining of those caches (e.g. in-lining the method and all code the method called right into the cache). Strongtalk was a Smalltalk implementation of the result of the Self research (plus optional H&M type checking) but before they released anything they got bought by Sun to get the expertise for their up coming HotSpot VM. So bottom line: Oracle may actually own all prior art via Sun (Self research was done at Sun).

> Interpreting functions utilizing a hybrid of virtual and native machine instructions Are they serious? Unless I'm missing something this is a fantastic joke, JITs far predate Java.

tptacek 16y ago

Careful. The name of a title and its actual claims are two very different things.

patrickaljord 16y ago
Careful. Google's implementation and the patent itself are two very different things.
shasta 16y ago
Well, here's claim 1: 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
- btilly 16y ago
  It sounds like the sort of stuff that Smalltalk implementations were doing before Java existed. See http://en.wikipedia.org/wiki/Just-in-time_compilation#History http://en.wikipedia.org/wiki/Just-in-time_compilation#Histor... for more. (I wonder whether Sun's patents were filed for Self?)
  - fhars 16y ago
    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.
  - gaius 16y ago
    sigh 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.
    - btilly 16y ago
      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.)
    - koenigdavidmj 16y ago
      Yes! And the first few claims generally paint way over the lines. It's in the later claims that you talk about the unique stuff that you did.
    - kj12345 16y ago
      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?
- Confusion 16y ago
  Again, as tptacek already said: careful. The summary isn't the actual claim either. The devil is in the details.
  - shasta 16y ago
    The above is the actual first claim from the patent. Of course it's subject to interpretation in context, but that is the text of the first claim.

nl 16y ago

I'm guessing, but I suspect working around this & other patents was why a JIT took so long to ship in Android.

bmastenbrook 16y ago
You can't write a good JIT overnight. You can't adapt an existing JIT for a JVM-like VM to Dalvik either. I don't think working around patents had much to do with it, especially since they got sued anyway.

js4all 16y ago

Careful. Hotspot compiling was introduced with the JVM.

bad_user 16y ago
There is prior art in Smaltalk and Self ... the Hotspot compiler itself is an old Smaltalk implementation that Sun bought, open-sourced recently by SUN btw ... http://www.strongtalk.org/ http://www.strongtalk.org/
- loewenskind 16y ago
  I'm pretty sure the prior art was just with Self. Smalltalk had polymorphic cache tables, but it was Self that was doing really aggressive in-lining of those caches (e.g. in-lining the method and all code the method called right into the cache). Strongtalk was a Smalltalk implementation of the result of the Self research (plus optional H&M type checking) but before they released anything they got bought by Sun to get the expertise for their up coming HotSpot VM. So bottom line: Oracle may actually own all prior art via Sun (Self research was done at Sun).