Hi David,

We're nearly at completion of the first phase our movie site using your API which has been really great to work with.

Thanks!

Will the clients be expected to do the queries for the data so as to spread over many ips? (key exposed)

This really just has to do with how you guys designed your app. Our rate limiting is purely IP based so if it's more of a server (hosted) app, this would be nearly impossible to do. If it's something like an iOS app then having each client do the requesting is perfectly fine and would cause very likely no trouble with the rate limiting.

Is the requesting meant for just the server but with extensive caching?

I'm not entirely sure I understand the question but I think I kind of answered it above. If it is a server backend, yes, we encourage you to cache the data to make less requests over time as the data is cached.

Thanks Travis, I'll share the link with shortly once we're up and running, we've created a recommender system and a new type of rating so look forward to seeing what you think.

We have designed it server based to be responsive to various devices so will look to cache the data to keep calls down to a minimum, do many of the server-based api users handle heavy traffic still staying under the limit?

Hey David,

It depends how long you're keeping the local cache for. I know a lot of our users end bringing the data in house mapped to their proper DB schema so that they can build the features they need both easier and properly. Then they just "fill" the DB as needed. You can then run periodic updates based on whatever criteria makes sense for you.

Thanks again for coming back so quick, did they fill their databases just from requests over time or receive a data dump? Of course, we would welcome this, or if it's ok, store your data down for our app's purposes from requests over time and use the change api to update it with fewer calls all round.

Thanks for all your help around this.

Did they fill their databases just from requests over time or receive a data dump?

We don't provide any dumps, so it would probably be a combination of over time and grabbing the popular and/or top rated items in advance. It would be pretty easy to pre-fill the top 1000 movies and then let everything else fill in over time. I'm just giving examples of course ;)