- From: Henrik Frystyk Nielsen <frystyk@w3.org>
- Date: Fri, 18 Dec 1998 21:35:44 -0500
- To: Chris Newman <Chris.Newman@innosoft.com>, discuss@apps.ietf.org
At 09:31 12/18/98 -0800, Chris Newman wrote: >I only skimmed the HTTP extensions draft and got the impression it was >unnecessarily complex. I won't have time to give a detailed critique >before going on vacation, so I hope someone else will. The following >point is important: Things have to be as simple as possible but no simpler. The choice between parameters vs. name spaces have been discussed for a long time and the consensus is to go with name spaces. Both solutions have pros and cons - none is significantly simpler than the other. >FYI, there was a long discussion in the USEFOR WG on header field prefixes >for headers with various characteristics. At the last IETF meeting of >USEFOR, the room reached the conclusion that adding such prefixes was >unnecessary complexity. The current model where all headers are optional >seems sufficient for extensibility. There was even a discussion of >labelling hop-to-hop headers in Netnews which is similar to the HTTP proxy >problem, and the same conclusion about unnecessary complexity was reached. What I believe you are saying is that HTTP is sufficient as is without an extension mechanism like the one proposed. I think the experience from the multiple ways HTTP is actually being extended clearly indicates that this is not the case. Henrik -- Henrik Frystyk Nielsen, World Wide Web Consortium http://www.w3.org/People/Frystyk
Received on Friday, 18 December 1998 21:36:09 UTC