[openstreetmap-website] Increase cachablility for individual OSM elements on the API (#129)

Andy Allan notifications at github.com
Fri Oct 12 01:00:42 GMT 2012

I think TomH's point on the redirects is valid - it's probably best not to implement the 303 (or similar) responses if that incurs extra round trips.

Given the optimistic locking on the api, we could probably have room to be more aggressive with caching, even on volatile results like normal way calls. However, as apmon alludes to, there's not many situations where we are going to get good cache hit ratios. The one situation where we have lots of people behind the same caching proxy asking for the same data is during (some) mapping parties, but then caching of volatile responses like map calls would work against us!

So we may as well add cache information to static calls (bearing in mind the redaction API, of course) but I can't think of many versioned calls that get frequent hits.

Reply to this email directly or view it on GitHub:
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/rails-dev/attachments/20121011/da127e12/attachment.html>

More information about the rails-dev mailing list