Authored by: ais523 on Thursday, May 24 2012 @ 09:39 AM EDT |
I think Oracle's point is that the list in the method header of thrown
exceptions could have been expressed in a different way, specifying only the
minimum set of types needed to cover all the possible exceptions without
introducing new ones. This is technically correct in an "it would work and
be compatible with existing programs" sense, but it'd break a whole load of
IDEs, because although the exception-handling code you've shown there would
still work, the IDEs would no longer have enough information to generate it.
In this trial, Oracle's been doing quite a lot of saying things that are
technically correct but misleading. (The way that symbolic reference question's
worded, for instance, Oracle could probably truthfully answer that numeric
references at one location in the code don't preclude the existence of symbolic
references at an entirely different unrelated location in the code. Which is
correct, but irrelevant.)[ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Thursday, May 24 2012 @ 09:42 AM EDT |
Oracle is correct. That code would still work and the throws portion of the
method signature would probably not need to changed if throwing super type
exceptions. If the method threw SomeSignatureException that inherited from
SignatureException, everything would be fine.
Of course, it's irrelevant as the super type would have the same SSO plus
extensions which would violate the copyright on the SSO of
SignatureException if Oracle had their way.[ Reply to This | Parent | # ]
|
|
Authored by: Ian Al on Thursday, May 24 2012 @ 11:48 AM EDT |
If the exceptions are functional rather than creative expression and if the
exceptions are short phrases and if the exceptions are ideas rather than
creative expression then they are uncopyrightable.
Just in case the court does not understand that just because a program compiles
using a modified API does not mean that it is still compatible with the program
that had the original API.
Finally, what is the whole work in which the exception creative expression is
fixed? Where is the registration for that whole work?
---
Regards
Ian Al
Software Patents: It's the disclosed functions in the patent, stupid![ Reply to This | Parent | # ]
|
|
Authored by: Anonymous on Thursday, May 24 2012 @ 04:49 PM EDT |
it is worse than this. Client code would no longer compile if Android used
supertypes in the throw clauses, since the catch clauses would no longer be
broad enough. Oracle is simply being dishonest--having the same throws
clause is necessary for compatibility.
On the other hand, the order technically doesn't matter for if code compiles,
but
keeping it the same improves the level of compatibility when it comes to tool
support. It is important that your IDE experience doesn't change when using
two versions of the platform. [ Reply to This | Parent | # ]
|
|