No code yet, waiting for Massimo's opinion on that.
I'm not too keen on using sched, because FWICS that means you should
start a new thread within web2py, start sched.run() in the thread,
which then sleeps there until the desired event occurs. This is OK if
you use web2py directly or through mod_proxy but... How would this
setup handle, say, two WSGI web2py processes ? Running a cron task
twice can cause quite a bit of trouble, and synching processes that
get restarted on the webserver's whim is not that simple. Also, the
web requests would have no guarantee of execution until your cron task
finishes which can be an issue for some long-running cron tasks. The 0
argument only gives chances to other threads between to event calls,
not during the execution of the event itself. So, for hard/exact
timing cron, I'm still not sure what's the ideal cross-platform,
execution environment independent mechanism for this.