bit of a shameless plug this, but i do have some technical questions
too. we've just released some
drop.io api powered code here:
http://www.sidepodcast.com/realtime/
and a description of what it is and how it works here:
http://www.sidepodcast.com/2009/03/08/powered-by-dropio/
in building the solution a couple of points came up that might be
worth sharing...
firstly, because the api works so well, for both uploading and
previewing media files, it might be possible that end users remain
oblivious to the existence of
drop.io. is there a requirement to
promote the name anywhere when using the api, and is there such a
thing as a "powered by
drop.io" style logo that we can use?
secondly, we're implementing the api via google app engine. it's
probably not necessary, but the reasons for using it were:
a) super fast response time
b) detailed information relating to throughput and bandwidth
re: point a. we're using memcached on appengine to cache the first
request for asset information, this speeds up subsequent calls, but i
was wondering if the location of the server also has a bearing on the
response time? does
drop.io use, or have any plans to use ec2
instances in europe, or even cloudfront?
re: point b. it wasn't clear to me as an api newbie, how many calls
could be made to the api in a small time period, and what happened
when we exceeded that number? i don't expect we'll ever hit the
limit, but the ability to track exactly what's going on with appengine
seems rather useful. are there any plans in the roadmap to provide a
dashboard-style front end to api developers?
anyhow, javascript -> python ->
drop.io and back again is the solution
we're currently using. it works very well and
drop.io is helping us
overcome a huge limitation within wordpress.
thank you
sidepodcast