- From: Paul Leach <paulle@windows.microsoft.com>
- Date: Tue, 2 Jan 2007 14:19:29 -0800
- To: "Travis Snoozy (Volt)" <a-travis@microsoft.com>, Larry Masinter <LMM@acm.org>, Mark Nottingham <mnot@mnot.net>
- CC: <ietf-http-wg@w3.org>
I'd also point out that 14.46 doesn't seem to make the same distinction about deleting 1xx warn-codes and not deleting 2xx warn-codes -- it seems to treat them pretty much the same. To wit, from 14.46: "However, if a cache successfully validates a cache entry, it SHOULD remove any Warning headers previously attached to that entry except as specified for specific Warning codes." No distinction between 1xx and 2xx. There's an awful lot of redundancy between 13.1.1-2 and 14.46. -----Original Message----- From: ietf-http-wg-request@w3.org [mailto:ietf-http-wg-request@w3.org] On Behalf Of Travis Snoozy (Volt) Sent: Tuesday, January 02, 2007 1:49 PM To: Larry Masinter; 'Mark Nottingham' Cc: ietf-http-wg@w3.org Subject: RE: NEW ISSUE: 13.1.2's Definition of 1xx Warn-Codes Larry Masinter said: > > The modified proposal (after discussion) is ... > > "A cache MUST NOT generate 1xx warn-codes for any messages > > except cache entries, and MUST NOT generate 1xx warn-codes > > for a cache entry except in response to a validation attempt > > for that entry. 1xx warn-codes MUST NOT be generated in > > Request messages." > > I think this rewrite is worse than the text it > proposes to replace, as far as being misleading. > The text is part of a description of the differences > between 1xx warnings and 2xx warnings, and the > 'right' rewrite is to make the descriptions more > parallel. > > The actual conditions for when a 1xx warning > may be generated (and MUST NOT) be generated > are contained in section 13.1.1. 13.1.1 specifies when Warning headers need to be generated; it defers to 14.46 when it comes to the actual warn-codes that need to be included. > Probably the right thing to do is to tighten up the > language in 13.1.1 so that it is clearly normative, > and then chanage the 3.1.2 Warnings section so that > it doesn't attempt to summarize them more succinctly > than they can be. I'd suggest: > > 1xx Warnings that describe the freshness or revalidation status of > the response. These warnings are generally deleted after > successful validation (the rules for when a cache MUST or > MUST NOT include or delete a warning response are in section 13.1.1.) > > 2xx Warnings that describe some aspect of the entity body or entity > headers that is not rectified by a revalidation (for example, a > lossy compression of the entity bodies). 2xx MUST NOT be > deleted after a successful revalidation. > Works for me, but I'd move "the rules for when a cache MUST or MUST NOT..." bit to precede the table (since it applies to both 1xx and 2xx codes), OR make it explicitly reference 1xx warn-codes and section 14.46 (since 3.1.1 does not actually talk about 1xx warn-codes at all). Thanks, -- Travis
Received on Tuesday, 2 January 2007 22:20:09 UTC