This is the mail archive of the guile@cygnus.com mailing list for the guile project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Another use for Guile/Scheme/Lisp


> But this is also why Java is not secure after all, as I understand
> it. 
>
> The Java security comes from the type system, but the bytecodes do not
> carry as much information as the source code.

Not individually, but the verifier does some non-trivial data flow
analysis, so effectively, your statement is false.

(I have written such a verifier, and it is available on the
sourceware.cygnus.com site.)

> So what can be verified
> to hold of some lump of java source code cannot necessarily be
> verified to hold of some other lump of bytecodes.

I have no idea what is supposed to mean.

> The Secure Internet Group at Princeton (or some such) has done quite a
> bit research into this area, and their conclusion was that fundamental
> changes to java are necessary if it is going to become a secure model
> of computation, again in my somewhat limited and utterly biased
> understanding.

Perhaps, but I suspect they are not talking about type safety.

Can we wind this discussion down?  It is rather off-topic.

	--Per Bothner
Cygnus Solutions     bothner@cygnus.com     http://www.cygnus.com/~bothner