It would be interesting to offer different global memcache queues with
different policies.
In your case, it seems to me if you have things you want to cache
indefinitely, why don't you just load them in the current interpreter
into a global dict on first hit? Effectively run your own local
I think your second one may be working because using the string
substitution is causing an unexpected value to appear in the query
that still is accepted by the engine. You should generate a log entry
before the query to see what passing the id to %s is actually giving
you.
On Apr 21, 8:10 am,
The dev server appears (obviously) to be single threaded, as is the
django dev server.
The presents a problem if you want to post to your own server
location. An example would be posting to the url provided by
create_login_url, which in dev answers at the same location as your
app. The server
doesn't expose anything that can be
used to make thread-safe applications. Since the production version
carefully hides any thread-like behavior, this is unlikely to change.
On Apr 1, 6:06 am, Mark Wolgemuth fuma...@gmail.com wrote:
The dev server appears (obviously) to be single threaded
