Re: remove PATCH, COPY, MOVE, DELETE, etc.? Upgrade?

Roy T. Fielding (fielding@avron.ICS.UCI.EDU)
Thu, 29 Feb 1996 00:54:42 -0800 

>>State tracking has never been considered important to
>>HTTP/1.1 
> 
> Huh?  Roy, you must be aware that we have a state management subgroup,
> which did not at all conclude that state management was not important
> for HTTP/1.1.

Of course I am aware of it -- I'm in it.  What is important for
HTTP/1.1 are those issues which require a new protocol version to
indicate that the client and/or server supports that capability or
at least recognizes that feature of HTTP.  State management is
completeley (last time I checked) orthogonal to the HTTP version
number, and thus can be progressed as a separate specification
[the one being worked on by the state management subgroup].

State management is orthogonal to the HTTP version because we defined
a mechanism for identifying any given response header field as carrying
stateful (i.e., non-cachable or private) information.  The reason why
this was done (as opposed to just defining a single mechanism within
the standard) is because 

   a) at the time, there were two competing proposals with no resolution
      in sight [this is no longer true];

   b) there was no clear consensus that state management within the
      protocol headers of a stateless protocol was even a good idea,
      let alone one that could be approved within a fast-track standard.


 ...Roy T. Fielding
    Department of Information & Computer Science    (fielding@ics.uci.edu)
    University of California, Irvine, CA 92717-3425    fax:+1(714)824-4056
    http://www.ics.uci.edu/~fielding/