[FoRK] overhead of RESTful stuff

Gregory Alan Bolcer greg at bolcer.org
Thu Mar 8 10:49:49 PST 2012


Lol wrt cache invalidation & coherence...A+REST. R+REST, EST+E. REST+D = 
  ARRESTED!

Greg

On 3/8/2012 10:44 AM, Lisa Dusseault wrote:
> The cache invalidation problem (and like you, I'd like to see more
> functionality to let people make use of this, but it is not commonly used)
> is why data/resource design is the most important performance work.  If
> there's any information in an app that can be publicly shared, then there's
> usually some way to cache that in a way that it can be re-used without
> getting too stale. Perfection is the enemy here, if data can be just a
> little bit stale, or if there are correction mechanisms for fixing stale
> public data if the user shows clear interest in a data point.  E.g. if "top
> ten user contributions by number of votes" is cached and gets ten minutes
> stale, and somebody looks at the item with 99 votes and clicks on it to see
> it really has a hundred, there's both a tolerance for staleness and a
> correction mechanism.

-- 
greg at bolcer.org, http://bolcer.org, c: +1.714.928.5476


More information about the FoRK mailing list