- From: Koen Holtman <koen@win.tue.nl>
- Date: Mon, 17 Feb 1997 21:47:12 +0100 (MET)
- To: hanna@world.std.com
- Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com, ietf-calendar@imc.org
Steve Hanna: > >I just posted a new Internet Draft titled "Calendaring Interoperability >over HTTP (CIH)". This document describes (in fairly vague terms for >now) how calendaring and scheduling systems could communicate with each >other using HTTP. A URL for the Internet-Draft is: >ftp://ds.internic.net/internet-drafts/draft-ietf-calsch-cih-00.txt I read the draft, and the idea to use HTTP for calendaring looks sound to me. If you are worried about the overhead of implementing HTTP/1.1, you may want to consider allowing both HTTP/1.0 and HTTP/1.1 to be used, while requiring that all 1.1-based implementations are downwards compatible with 1.0. 1.0 can be implemented easily, and there is a lot of free code around that does it. 1.1 can be used when high performance is needed. It is trivial for an 1.1 implementation to be downwards compatible with 1.0. If you go this way, I suggest changing the CIH URL encoding to http://cih.world.std.com/cih/hanna ^^^^^ this will make it easier to set up CIH servers by using existing HTTP/1.0 software. (You can basically let your regular web server act as the CIH server if you set up cih.world.std.com as a DNS alias for www.world.std.com, and install a CGI script under /cih). A problem is that HTTP/1.0 is not an official IETF standard, this may make it difficult to use HTTP/1.0 process-wise. If CIH is a long-term project (say planned for the end of this year), you may want to forget about using HTTP/1.0 altogether. By that time, a sufficient amount of free HTTP/1.1 code will be available. Some 1.1 code exists already, but as far as I know it is mostly unix-only. >Steve Hanna Koen.
Received on Monday, 17 February 1997 12:58:35 UTC