- From: timeless <timeless@gmail.com>
- Date: Wed, 12 Aug 2009 12:43:51 +0300
Jonas Sicking wrote: > The only site where I can > remember seeing content negotiation actually used is on w3.org fwiw, MXR (and even LXR) uses some content negotiation, and it generally "magically works". OTOH it's transparent, so you shouldn't "see" it :). But yes, I'd say that content negotiation is a failure. Another failure in that area is Link: :). The problem with content types i think stems more from: 1. images generally having random file extensions because users didn't understand them 2. users who were publishing content in other formats not actually having enough easy control over their hosting providers to specify the content type, or knowing that it was necessary Partially, both of these stem from the internet mantra: be strict in what you send and graceful in what you receive. However, only computers can be strict in what they send. So "be ignorant of what you send" + "be graceful in what you receive" results in "no one can trust what is sent".
Received on Wednesday, 12 August 2009 02:43:51 UTC