Even with bot downtime, it shouldn't be a problem to build a
bot-downtime monitoring system. It would require two bots
participating in a wave:
1. 'Pong Bot' - which needs to be able to respond to any Wave event
with a timestamp written as a blip.
2. 'Ping Bot' - which upon a cron tick will attempt to modify a
wave(let). It should then listen to any change events and when the
other bot responds, update a 'last known good' timestamp.
This last-known-good timestamp should be publicly accessible and is a
good basis for any downtime monitoring system. Notification can be
done using appengine's own cron mechanism.
Cheers,
Tal
On 7 Aug 2009, at 00:02, Jason Salas <digitalpon...@gmail.com>
wrote: