Re: Advanced scoring rule omitted from trace buffer when traditional rules also apply

[Reiner Steib <reinersteib+gmane@imap.cc>]

On Tue, Oct 11 2005, Jonathan Hankins wrote:

> Reiner,
>
> Thanks for the quick reply.  The information you requested is included
> below.

Please don't strip the list address when replying (see
Mail-Followup-To header).  As bugs@gnus.org is quite overloaded with
spam, I move the discussion to the ding list (Gnus development list)
instead.

I always get lost when trying to understand `gnus-score.el', so I hope
someone else can look into this... ;-)

Bye, Reiner

Full-quote of Jonathan's message follows...

> Reiner Steib <reinersteib+from-uce@imap.cc> writes:
>
>> On Tue, Oct 11 2005, Jonathan Hankins wrote:
>>
>>> However, when I have a combination of "traditional" and "advanced"
>>> scoring rules that all apply to a message, the "V S" command shows the
>>> correct score, but when I use the "V t" command, the trace shows only
>>> the "traditional" rules.
>>>
>>> Let me know if I can provide any further information.
>>
>> - The relevant score rules
>
> (((&
>  ("subject" "No Gnus" s)
>  ("from" "Magne" s))
>  500)
>  ("from"
>   ("Lars" 1 nil s))
>  ("subject"
>   ("Oort" 1 nil s)))
>
>>
>> - Message-Id of a public article (usenet or from some mailing list
>>   carried by gmane.org) which triggers the incorrect `V t' output.
>>   Alternatively post such a test case article to gmane.test on
>>   news.gmane.org.
>
> (nntp+news.gmane.org:gmane.emacs.gnus.announce)
>
> From: Lars Magne Ingebrigtsen <larsi@gnus.org>
> Newsgroups: gmane.emacs.gnus.announce
> Subject: No Gnus v0.3 is released
> Date: Sun, 01 May 2005 19:02:43 +0200
> Message-ID: <m3acneyj5t.fsf@quimbies.gnus.org>
>
> V-S shows: 501
> V-t shows:
>     ("Lars" 1 nil s)
>     [nntp+news.gmane.org:gmane.emacs.gnus.announce.SCORE] ->
>     ~/News/nntp+news.gmane.org:gmane.emacs.gnus.announce.SCORE
>
> The rule:
>     ((&
>       ("subject" "No Gnus" s)
>       ("from" "Magne" s))
>      500)
>
> adds the 500 points, but is not displayed in the trace.
>
> --
>
> To contrast, if I delete the rule:
>     ("from"
>      ("Lars" 1 nil s))
>
> leaving me with this score file:
>     (((&
>        ("subject" "No Gnus" s)
>        ("from" "Magne" s))
>       500)
>      ("subject"
>       ("Oort" 1 nil s)))
>
> after V-R, V-S shows: 500
> V-t shows: 
>     ((& ("subject" "No Gnus" s) ("from" "Magne" s)) 500) 
>       [A file] -> A file