>>>>> "DS" == Dan Sugalski <[EMAIL PROTECTED]> writes:
DS> It works pretty well, all things considered. Not the easiest way to
DS> program, but async code's always a big pain in the neck.
ahh, back to the days of rt-11 which had optional callbacks for all of
its i/o calls including files. no wonder i like the callback style so
much, i have been using it for so long. :)
too bad all modern OSs don't all have simple to use async i/o. it would
make life (and perl6) SO much nicer. the few that have async, all have
very different api's which makes writing the perl6 aio stuff a study in
porting. and add to that emulating async i/o on boxes without it. you
have to do a blocking i/o call and then call the callback.
this brings up a thought. we should specify the async i/o internal api
soon. this will be the way the rest of perl6 guts will see async i/o and
we port that api to each platform as needed. as i see it, the async i/o
api will be related to or part of the event api. we can spec all this
out now as we know the required features of these subsystems. any
thoughts on tackling these api's now since you are starting on the i/o
vtables?
uri
--
Uri Guttman --------- [EMAIL PROTECTED] ---------- http://www.sysarch.com
SYStems ARCHitecture, Software Engineering, Perl, Internet, UNIX Consulting
The Perl Books Page ----------- http://www.sysarch.com/cgi-bin/perl_books
The Best Search Engine on the Net ---------- http://www.northernlight.com
