« Future-Proofing REST Clients | Main | Two Sides of Data Science »

Tuesday, October 29, 2013


Feed You can follow this conversation by subscribing to the comment feed for this post.

Ruben Verborgh

While I also was in favor of OPTIONS, mnot points out some counterarguments: http://www.mnot.net/blog/2012/10/29/NO_OPTIONS
How about we use just HTTP linking to a description and then GET?


@RubenVerborgh, thanks for the comment. exposing the link hint/description via OPTIONS is just one way of doing it, and may not be so working so great in practice, given mark's reservations about cacheability. but if you simply replaced the OPTIONS approach with (and the current link descriptions draft actually does propose such a link relation http://tools.ietf.org/html/draft-wilde-link-desc-00#section-6.2), i think you end up with pretty much the same line of argument, but without the potential drawbacks of using OPTIONS. you could also decide to embed the link description in the representation you GET, but that might be very noisy and not what you want to do in practice. do you see anything wrong with that picture?

Ruben Verborgh

No, that's indeed a solution. Just a pity that it was decided to make OPTIONS non-cacheable—now we have the choice between an extra HTTP request or a noisy representation. Luckily, the automated discovery mechanism remains intact.

The comments to this entry are closed.