|
Authored by: dio gratia on Monday, May 14 2012 @ 08:59 PM EDT |
Symbol table generically refers to structured data. There's more than one
way to organize symbol data (e.g. List, Self Organizing List, Hash Table, Binary
Search Tree). I'd be a little more careful and say if you are comparing the
symbol to a stored value you are doing a symbolic reference. Oracle seemed to
be trying to 'confuse a cat' by talking about numeric references, when names in
Java must begin with a character.
You'd think a compiler writer could also
demonstrate why the the class/method/object structure, sequence and organization
of Java is functional, in that names as not copyright eligible organized in path
names allows name reuse by limiting scope and visibility. The very reason why
names can't be copyright, they aren't unique, they aren't
original.
The SSO issue revolves around path names. See 1018, Final
Charge To the Jury (Phase One) and Special Verdict Form, at 10, instruction
20:
Now, let me tell you the law about names. The copyrights do not
cover the names, such as those given to files, packages, classes, and methods,
because under the law, names cannot be copyrighted. This applies to the name
“java” as well. Although “Java” has been registered as a trademark, there is no
trademark claim in this lawsuit. The name java cannot be copyrighted, nor can
any other name, whether one or two words or longer in length. While individual
names are not protectable on a standalone basis, names must necessarily be used
as part of the structure, sequence, and organization and are to that extent
protectable by copyright.
It's this SSO, seemingly hierarchical
names that allows names to be used locally. Java doesn't have a flat name
space, and all names don't have to be unique. A path name makes a non-local
name visible in the current scope without a local declaration or concern for
reuse.
(And this is done by using a symbol table, too). (An example path
name demonstrating SSO: java.io.File). [ Reply to This | Parent | # ]
|
|
Authored by: BitOBear on Tuesday, May 15 2012 @ 01:18 AM EDT |
"Bring me the Oracle v Google File" is a demand for symbolic lookup
because you, the "bringer" must deal with finding the file in question
inside the file cabinet or storage box or whatever. Neither gross location nor
position within that location are of any interest, or even necessarily known to,
the entity making the request. On the average day the boss doesn't -care- how
you file the files.
"Bring me that thing that is on top of that desk" is a demand for
non-symbolic lookup. You have been told exactly where the thing to be fetched is
and your job is to go and get it. No thinking required. You are a pair of feet
and a grasping hand.
"Bring me the third file from that box there" is a table lookup, which
is still not symbolic. You are told where to go, and where within that larger
location, to find the thing. Again you are a pair of feet and a grasping hand.
You will, also have to be able to count to three this time, and you will have to
multiply by "size of file" or some such. The amount of time to
"figure out" this request is fixed and you, the fetcher thing, have no
duty to bring "the correct file, by name or any such" since you are
relying on the assumption that the files are already in the correct order and
the asker knows what he is asking for by that placement and order.
So, as always, pretend computers do not exist. If I ask you to get something
"by name" you know that names are "symbolic references" to
things. "BitOBear" is a symbolic reference to me in this context. If I
am also "Groklaw member 88832585765383" then that number -could- be a
symbolic name for me as well. The "number-ness or not" is not the
point. But if that were truely "my number" it isn't likely because
there are trillions of unique groklaw members since there are only billions of
unique people on the planet.
To get "the 45th thing from a bin" you do not have any
"symbolism" to deal with.
The "expert" is -deliberately- trying to conflate "a arbitrary
number used as a name" (which is -NOT- what is happening in this case) with
the mathematical entity "the number-eth slot in the box/table/list at
hand".
One can only assume he is either lying or very bad at his job if he claims this
conflation is correct or useful.[ Reply to This | Parent | # ]
|
|
|
|
|