W3C home > Mailing lists > Public > ietf-http-wg-old@w3.org > May to August 1996

Re: Minor fixes to draft 05

From: Jeffrey Mogul <mogul@pa.dec.com>
Date: Wed, 03 Jul 96 15:56:49 MDT
Message-Id: <9607032256.AA03593@acetes.pa.dec.com>
To: "Roy T. Fielding" <fielding@liege.ICS.UCI.EDU>
Cc: http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
    >   - and it is the most up-to-date response appropriate to the request the
    >   - cache has (see section 13.2.5, 13.2.6, and 13.12).
    
    > I'm not sure why you propose deleting those lines.
    
    Because, when I rewrote that section after draft 04, I put those lines
    in the first paragragh (i.e., look above it if you have the draft handy).
    It was just a paste error when Jim applied the changes (no worries).

Oops, I missed that.  Sorry.
    
    >  *** 4022,4029 **** [13.2.3 Age Calculations, as requested by Ben Laurie]
    
    >  In this discussion, we use the term "now" to mean "the current value of
    > ! the clock at the host performing the calculation." Internet hosts that
    > ! use HTTP, particularly those hosting origin servers and caches, should use
    > NTP [28] or some similar protocol to synchronize their clocks to a
    > globally accurate time standard.
    >       
    > If I understand your use of the word "Internet" (as a replacement for
    > the word "all"), this change now means that intranet hosts (i.e., those
    > not directly reachable from the Internet) are not expected to
    > have synchronized clocks.
    > 
    > I have no idea why anyone would want to run an intranet Web any
    > differently than the Internet Web.  But in any case, this is a
    > "should", not a MUST, and so I see no reason to change the suggestion
    > from the draft-05 version.
    
    No.  As Ben mentioned several times, it is ludicrous to suggest that
    HTTP *implementations* should use NTP -- instead, the hosts running
    HTTP should also run NTP.  If you can rephrase that in a better way, fine,
    but the existing text is just plain bogus.

OK, so your intention was not to exclude intranet hosts, but merely
to shift the locus of the NTP suggestion from "HTTP implementation"
to "host that uses HTTP".  (I re-read Ben's message, and this was NOT
his main objection; he seems to have incorrectly assumed that use of
NTP requires Internet connectivity.)

So I think you and I would be in agreement that this wording should
be used:

    In this discussion, we use the term "now" to mean "the current
    value of the clock at the host performing the calculation."
    Internet or intranet hosts that use HTTP, particularly those
    hosting origin servers and caches, should use NTP [28] or some
    similar protocol to synchronize their clocks to a globally accurate
    time standard.

Or simply "Hosts that use HTTP, ..."

-Jeff
Received on Wednesday, 3 July 1996 16:12:33 EDT

This archive was generated by hypermail pre-2.1.9 : Wednesday, 24 September 2003 06:32:04 EDT