Puzzle: Why do some follow-ups thread and some not

Puzzle: Why do some follow-ups thread and some not

[Steinar Bang]

Emacs 19.31, Gnus 5.2.14 (but this behaviour I've observed since 5.1)

What makes some articles in mail folders thread, and some not?

Articles made with a follow-up with Gnus seem to thread, articles
created with a reply in RMAIL seem to thread.  Articles made with a
reply with Mew seem to thread.

Articles made with a reply with mh-e doesn't thread.  Ditto for exmh.
I've created a replcomps file that gives me In-reply-to headers and
References headers.

At the start, I *think* I got these to thread, but then they suddenly
stopped threading.

But all of this may be caused by erronous headers.  So here comes the
mysterious part:
 I have two articles.  One made by RMAIL (the one from Trond
 Kristiansen), one made by mh-e (the one from me).

I've tried hand editing the one from me, to see if I could get it to
thread, and I've been unsuccessful in doing so.

I've done the testing by editing, then starting a new emacs, then
killing all Gnus buffers and exiting emacs, answering "yes" on the
"unsaved buffers exits, do you really...etc.".  Then editing, and
starting a new emacs, answering "n" on the offer to use auto save
file (even when I've deleted "#.newsrc-dribble" it displays this
prompt). 

So here are the two headers.  As near as I can see, all headers that
should be relevant for threading are identical.  But one thread and
the other doesn't.

This isn't like, important.  But it really *bugs* me not knowing *why*
this happens...:-(

Is the information about the threading cached in some obscure place?
I've tried stripped everything relating to this mailing list out of
.newsrc.eld (without finding anything, when I refused the offer to
save). 



Header 1

Replied: Fri, 14 Jun 1996 11:33:20 +0200
Replied: "Trond Kristiansen <tk@metis.no> dk,jh,lro,ht"
Return-Path: tk@metis.no 
Received: by hub.metis.no (8.6.12/8.6.12) with ESMTP id MAA01848; Thu, 13 Jun 1996 12:30:31 +0200
Received: by client.metis.no (8.6.11/8.6.12) id MAA01099; Thu, 13 Jun 1996 12:30:29 +0200
Date: Thu, 13 Jun 1996 12:30:29 +0200
Message-Id: <199606131030.MAA01099@client.metis.no>
From: Trond Kristiansen <tk@metis.no>
To: dk@metis.no
CC: jh@metis.no, lro@metis.no, ht@metis.no, sb@metis.no, lro@metis.no
In-reply-to: <31BFDACB.2698@metis.no> (message from Dag Karlsen on Thu, 13 Jun
	1996 11:09:31 +0200)
References: <31BFDACB.2698@metis.no>
Subject: Re: New 1.x API


Header 2

Received: by hub.metis.no (8.6.12/8.6.12) with ESMTP id LAA09768; Fri, 14 Jun 1996 11:16:27 +0200
Received: by client.metis.no (8.6.11/8.6.12) with ESMTP id LAA20466; Fri, 14 Jun 1996 11:16:26 +0200
Return-Path: sb@metis.no
Date: Fri, 14 Jun 1996 11:16:25 +0200
Message-Id: <199606140916.LAA20466@client.metis.no>
From: Steinar Bang <sb@metis.no>
To: Dag Karlsen <dk@metis.no>
CC: jh@metis.no, lro@metis.no, ht@metis.no, sb@metis.no, lro@metis.no
In-reply-to: <31BFDACB.2698@metis.no> (message from Dag Karlsen on Thu, 13 Jun
	1996 11:09:31 +0200)
References: <31BFDACB.2698@metis.no>
Subject: Re: New 1.x API 

Re: Puzzle: Why do some follow-ups thread and some not

[Jens Lautenbacher]

From: Jens Lautenbacher <jens@lemcbed.lem.uni-karlsruhe.de>
Date: 14 Jun 1996 19:34:56 +0200
In-Reply-To: Steinar Bang's message of 14 Jun 1996 18:28:42 +0200
Message-ID: <x2lohqb873.fsf@lemcbed.lem.uni-karlsruhe.de>
Lines: 8
X-Mailer: Gnus v5.2.17/XEmacs 19.14


In your case the answer is quite easy: The second subject line has a
space at the end, the first one not. 

I see this happening many times, maybe we could ignore whitespace at
the end of subject lines? (Lars?)

Re: Puzzle: Why do some follow-ups thread and some not

[Sten Drescher]

>>>>> Jens Lautenbacher writes:

JL> In your case the answer is quite easy: The second subject line has
JL> a space at the end, the first one not.

JL> I see this happening many times, maybe we could ignore whitespace
JL> at the end of subject lines? (Lars?)

	And at the beginning - I'm on a mailing list that insists on
using Header:<tab> instead of Header:<space>.

-- 
+----------------------  Tivoli Customer Support  ----------------------+
|   Sten Drescher                     Tivoli Systems, Inc               |
|   email: sten.drescher@tivoli.com   9442 Capital of Texas Hwy North   |
|   phone: (512) 794-9070             Arboretum Plaza One, Suite 500    |
|   fax  : (512) 345-2784             Austin, Texas 78759               |
+-----------------------------------------------------------------------+

Re: Puzzle: Why do some follow-ups thread and some not

[Lars Magne Ingebrigtsen]

Jens Lautenbacher <jens@lemcbed.lem.uni-karlsruhe.de> writes:

> I see this happening many times, maybe we could ignore whitespace at
> the end of subject lines? (Lars?)

Well, we could allow a new value to `gnus-thread-ignore-subject' --
`spaces', or something.  (We could even default to that.)  And then
subjects that differ in white space only could be considered the
"same" subject for threading purposes.  I've now added this to the Red
Gnus todo list.

-- 
  "Yes.  The journey through the human heart 
     would have to wait until some other time."