BBC APIs

[Barry Norton]

Bit of a late notification, but I was talking again last week to Dave Rogers about the APIs that the BBC build over the (OWLIM) triplestore, and I realised that not everyone might be aware of this:
http://www.bbc.co.uk/blogs/bbcinternet/2012/08/olympic_data_xml_latency.html

Essentially, in order to isolate the triplestore from any Tom, Dick or Harry that might want to mess with it (either with resource-hungry queries or dangerous updates), each API call wraps a scoped, parameterised and well-tested (SELECT and then) CONSTRUCT query. This makes them very amenable to the descriptions we make with graph patterns.

Not all of that is explained in the blog post, but it's nice to have something to cite.

Barry