Return to the
comp.compilers page.
Search the
comp.compilers archives again.
| Related articles |
|---|
| Re: Mixing languages stephens@mcs.com (1995-06-03) |
| Re: Mixing languages pardo@cs.washington.edu (1995-06-05) |
| Re: Mixing languages ok@cs.rmit.edu.au (1995-06-24) |
| Re: Mixing languages gbaker@rp.CSIRO.AU (1995-06-27) |
| Re: Mixing languages jdean@pysht.cs.washington.edu (1995-06-28) |
| Newsgroups: | comp.lang.scheme,comp.compilers |
| From: | pardo@cs.washington.edu (David Keppel) |
| Keywords: | design, interpreter |
| Organization: | Computer Science & Engineering, U. of Washington, Seattle |
| References: | <3pf727$qn2@goanna.cs.rmit.edu.au> 95-06-013 |
| Date: | Mon, 5 Jun 1995 02:20:53 GMT |
| Status: | RO |
>Richard A. O'Keefe writes
>>[Sun/Java compile to a portable IR and demand compile to native code.]
In article 95-06-013 stephens@mcs.com writes:
>[Why invent another virtual machine? Use x86 and write one virtual
> machine for all other platforms.]
There's at least two reasons.
First, a VM emulator is only half the problem, the other half is the
system call interface (ABI). For DOS, that's a Very Big half.
Second, Java is used to move untrustable programs between systems, and
Java makes such movement safe by guaranteeing that the execution engine
(interpreter or cross-compiler) will implement certain checking.
First, that checking isn't required for arbitrary x86 execution.
Second, Java uses security mechanisms that would be hard to implement
at the machine-code level.
;-D on ( My Chine Code ) Pardo
--
Return to the
comp.compilers page.
Search the
comp.compilers archives again.