From: Bruce Lilly (blilly@erols.com)
Date: Wed Mar 17 2004 - 06:28:03 CST
Charles Lindsey wrote:
> True, but why should it be necessary to know which are the immediate
> predecessors?
Various possibilities:
To implement a don't-expire-recently-referenced policy for
expiration.
To be able to move to immediate predecessors (to view quoted
text in context, etc.).
To be able to identify articles which are followups to
multiple messages.
To be able to quantify the number of articles which are
immediate predecessors of a given article.
[Like References, these are not "necessary" in a protocol sense,
but they may be "useful"]
> Actually, it is rather easy in NNTP, by making use of the overview
> information, which is what newsreaders usually do. There is still the
> problem that some earlier messages from the thread may have expired.
Are References fields universally available via overview (at
least one person recently active on this mailing list has said
that that is not the case)?
In any event, it is simpler for the followup agent to indicate
one time (via In-Reply-To) which articles are the immediate
predecessors (information obviously known to that followup
agent) than it is for thousands of individual reading agents to
request a References field, request References fields for all
articles whose message-ids are listed therein, receive some
fraction of those, attempt to build a dependency graph based on
that partial data, and then attempt to guess the number of
immediate predecessors and to guess which of the predecessors
are the immediate ones -- and to do that each time the same
article is (re-)read.
If we're going to address followups to multiple messages, we're
going to need to examine and modify the draft text related to
Newsgroups, Followup-To, Distribution, etc.; the content of
those fields may differ among the multiple articles being
followed up.
Example:
article A may have
From: a@b.com
Newsgroups: foo, bar
Followup-To: bar
Subject: frammis
Distribution: world
Mail-Copies-To: poster
article B may have
From: c@d.edu
Newsgroups: foo, baz
Followup-To: baz
Subject: Whatsit (Was: frammis)
Distribution: ru
Mail-Copies-To:xyz@isp.net
article C may have
From: e@f.gov
Newsgroups: foo
Followup-To: poster
Subject: thingamajig
Distribution: uk
Mail-Copies-To: nobody
Leaving aside References and In-Reply-To for the moment, consider
a followup to some combination of two or more of A, B, and/or C.
Which newsgroups should be placed in the Newsgroups field of the
followup by a followup agent? The Usefor draft doesn't address
that issue (for a followup to multiple messages). If the
followup agent presents some Subject field content for the
followup, what should it be? The Usefor draft doesn't address
that issue. What should the followup agent do for Distribution
for the followup? The Usefor draft doesn't address that issue.
Where should mailed copies of the followup be sent? The Usefor
draft doesn't address that issue.
Side issue:
Is the combination of
Followup-To: poster
and
Mail-Copies-To: nobody
in article C legal and semantically sensible? The Usefor draft
doesn't address those issues (even for the case of article
C considered in isolation). Perhaps the two fields should be
combined into a single field:
Whatever: baz, bar ; followup to newsgroups baz and bar
Whatever: baz, poster ; followup to newsgroup baz, mailed copy to poster
Whatever: poster ; no posted followups, responses mailed to poster
Whatever: baz, foo@bar.com ; followup to baz, mailed copy to foo@bar.com
Whatever: foo@bar.com ; no posted followups, responses mailed to foo@bar.com
Whatever: baz, poster, foo@bar.com ; followups to newsgroup baz, mailed copies to poster and foo@bar.com
Whatever: poster, foo@bar.com ; no posted followups, responses mailed to poster and foo@bar.com
Whatever: ; no followups or mailed copies
field absent ; followups to newsgroups in Newsgroups field(s), no mailed copies
#################################################################
#################################################################
#################################################################
#####
#####
#####
#################################################################
#################################################################
#################################################################