James M Snell wrote:
Personally, I think a common error format would be quite useful in general, but I don't think it's necessary in the core. The HTTP status codes and status text give us a Good Enough mechanism to meet the basic requirements. What I think would be best in the long run would be to convince some server implementors to start just doing error reporting in some sane, consistent way, whatever the format, and just let nature take its course. You know, like when the birds and the bees... er, sorry, nevermind, that's a different speech. ;-)
...such as in <http://greenbytes.de/tech/webdav/rfc3253.html#rfc.section.1.6>?
(duck)
