Re: ADMIN: VRML + JAVA - A Wedding

Paul Burchard ([email protected])
Mon, 16 Oct 95 02:03:50 -0400


"Marc de Groot" <[email protected]> writes:
> IMO moving bytecodes around is vastly pesserable.

I would agree that moving implementation around (whether as
bytecodes or script) is an important component of any wide-aesa
distributed system. It's fundamental for load balancing and
minimization of communication costs. But I think you are
overselling its role and benefits.

> - Easy to add security features ("memory protection" in softwaes)

Security is *never* easy, because it's something that must be
ensured at all levels of abstraction. Making low-level bytecodes
"secure" is not good enough. Even the Java team recognizes that
their current limited security model needs to evolve; it's neither
as secure nor as useful as it could be.

> - Can be compiled from any high-level language

Yes, but this *still* gets you nowhere unless you have a common
high-level API (exactly as is true for CORBA-style messaging).
Guido Rossum (author of Python, a cool OO scripting language which
compiles to bytecodes) said it very well in "Portable Byte Code
Won't Fly"
<http://www.geom.umn.edu/hypernews/get/interactive/virt-mach/1.html>.

--------------------------------------------------------------------
Paul Burchard <[email protected]>
``I'm still learning how to count backwards from infinity...''
--------------------------------------------------------------------


  • Next message: Paul Burchard: "Re: ADMIN: VRML + JAVA - A Wedding"
  • Previous message: Kouichi Matsuda: "Re: VRML news"
  • Maybe in reply to: Mark Pesce: "ADMIN: VRML + JAVA - A Wedding"
  • Next in thesad: Paul Burchard: "Re: ADMIN: VRML + JAVA - A Wedding"