- From: <Harald.T.Alvestrand@uninett.no>
- Date: Mon, 25 Mar 1996 13:00:56 +0100
- To: "Roy T. Fielding" <fielding@avron.ICS.UCI.EDU>
- Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Just clearing out the tail end of the chain here..
"Roy T. Fielding":
> > - It STILL loses the method information.
>
> I don't understand this one -- the method is still there. The only thing
> that is lost is the original URL scheme and that information is inherent
> in the connection to the origin server or gateway.
My point: I think that it was a design mistake to link together:
- The thing is identified by a "name-like" thing that starts with HTTP,
and has a last resort access method using the HTTP protocol
- The fact that we are currently using the HTTP protocol to access it.
Even on the link to the origin server.
But if we all accept that:
- HTTP/1.1 is a stopgap until multiplexed HTTP (2.0) comes along
- HTTP/2.0 will have full URLs and no Host: header
then I don't see any reason to keep on beating this dead horse at this moment.
> > - It STILL gives us no path to where it seems everyone wants to be, namely
> > with full URLs, until we throw out HTTP/1.x altogether
>
> We have a path. HTTP/1.0 now --> HTTP/1.1 on May 1 --> HTTP/2.0 as soon
> as multiplexing + PEP + cookies have a solid, agreed-upon syntax.
> HTTP/1.1 contains the mechanisms necessary to make deployment of HTTP/2.0
> possible.
>
Is it possible to get this strategy written down in a 4-page "strategy"
document?
It might be nice to have the words down on phosphor that we agree upon.
(Or I might just have missed the URL of the document, of course....)
Thanks for caring!
Harald A
Received on Monday, 25 March 1996 04:05:06 UTC