Re: New issue: Need for an HTTP request method registry

On tis, 2007-08-07 at 13:50 +0100, Jamie Lokier wrote:

> In 10 years time, won't we see exactly the same infrastructure problem
> repeated with something-over-HTTP, just moved up a level?

If the protocol isn't allowed to expand then yes, very likely, as people
will then define more and more application protocols on top of HTTP
which really should have been done within HTTP. Consider for example the
situation if WebDAV only had GET/POST/HEAD to work with and not allowed
to expand.. we would then have seen a standardization on POST body
content, and intermediaries wishing to enforce policies on "PostDAV"
would need to parse not only HTTP but also the "PostDAV POST application
format".

However, adjusting the protocol to deal with those abusing HTTP as
transport it in absense of a TCP/IP transport is not something we should
have on the agenda. There is already SOCKS covering that use case and
doing a very good job of it. But it's a sad truth that it has become
common practice to use HTTP for this purpose.

Regards
Hernik