John Peacock wrote:
> "Craig A. Berry" wrote:
>>
>> >The docs above doesn't make it very clear, but my own (admittedly naive)
>> >understanding of fork is that it copies a snapshot of the parent's
>> >VM to the child process, and starts the child just after the point where
>> >the parent did called fork.
>>
>> Hmm, so it does want a genuine fork(). Rats.
> This is probably simplistic, so ignore it if it makes no sense. If VMS
> can start a subprocess with Perl running the same script, could you
> diddle the source code for the child to include a GOTO to the line
> after fork. For example, shove the script into the child interpreter
> via STDIN with the appropriate changes on the fly.
This way lies madness ;)
Just imagine the debugging nightmare.
I think it would be better to try and figure out how to really fork...the
VM copy one could do (a new kernel-mode system service, some sort of
"dump"/"reload" in Perl) but one has to somehow keep the i/o straight and
I suspect that would be hard.
--
Drexel University \V --Chuck Lane
======]---------->--------*------------<-------[===========
(215) 895-1545 _/ \ Particle Physics
FAX: (215) 895-5934 /\ /~~~~~~~~~~~ [EMAIL PROTECTED]
