- From: Koen Holtman <koen@win.tue.nl>
- Date: Mon, 4 Sep 1995 21:37:50 +0200 (MET DST)
- To: Roy Fielding <fielding@beach.w3.org>
- Cc: koen@win.tue.nl, http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
Roy Fielding: > >>But I must object to your use of `should'. This is the http-wg list, >>so I expect everyone to be using the language of the draft HTTP spec. >>The spec uses `should == must', as far as I can tell. (Lots of RFC's >>may not, but that is another matter.) > >No it doesn't. Must is a requirement for compliance, should is >a recommendation. I kind of expected you to say that. Let me try to fix my own sloppy use of terminology: I read the `should' in the spec to a bit weaker than `must', but not much weaker. On a scale of recommendedness: must > should > strongly recommended > recommended > may So I think that the `should' in the spec implies far less optionalness than Shel's `should' seems to do. In particular, I read the spec's `should' to take precedence over all efficiency considerations: gaining efficiency (saving bandwidth) is not not a good enough reason to violate a `should'. This is why I complain about caching `shoulds' that introduce very inefficient worst cases (like having to clear the entire cache memory if the cache comes up after having been down). > ....Roy T. Fielding Department of ICS, University of California, Irvine USA Koen.
Received on Monday, 4 September 1995 12:42:22 UTC