- From: Larry Masinter <masinter@parc.xerox.com>
- Date: Sat, 9 Dec 1995 08:00:18 PST
- To: koen@win.tue.nl
- Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Coming to agreement on a solution to a problem requires agreement on two things: 1: the problem, 2: the solution. We have to agree that there's a problem, and the nature of the problem, and we have to agree on the solution, that it really solves the problem, that if we have more than one potential solution to the problem, that we've picked the best one. I think in the case of state-info there's agreement that there's a problem, and there's mainly agreement that draft-kristol-etc probably solves the problem, but less agreement that it is a better solution than Netscape's cookies. It's easy to confuse "agreement that there's a problem" with "agreement on the solution", and it appears you have done so here. > One other thing: Roy Fielding recently said on this list: > | "Rough consensus" can only be tested by polling the list. > |To date, the only piece of HTTP that has ever gained "rough consensus" > |for includion in HTTP/1.1 is the Host header field and those features > |which already exist in HTTP/1.0 (by definition). > Would you (the WG chairs) please either confirm or deny this? I deny both uses of the word 'only'. I dislike spending http-wg bandwidth on 'process' questions; if you have further comments on process, please mail me directly. Since we've constituted subgroups, let's try to let them work.
Received on Saturday, 9 December 1995 08:03:24 UTC