Help with Adaptive Scoring Problems (2nd try)

Help with Adaptive Scoring Problems (2nd try)

[IPmonger]

Folks,

    I thought I sent this late last night, but I haven't seen it go
  through yet, so here goes...

    I used to use the following adaptive-score-alist:

(defvar gnus-default-adaptive-score-alist
  '((gnus-unread-mark (thread 10))
    (gnus-ticked-mark (from 4))
    (gnus-dormant-mark (from 5))
    (gnus-del-mark (from -4) (subject -1) (thread -1))
    (gnus-read-mark (from 4) (subject 2) (thread 25))
    (gnus-expirable-mark)
    (gnus-killed-mark (from -10) (subject -3))
    (gnus-kill-file-mark)
    (gnus-ancient-mark)
    (gnus-low-score-mark)
    (gnus-catchup-mark (from -1) (subject -1))))


    But, when I set it up this way now, the (thread <score-delta>)
  entries cause an Elisp error when I attempt to exit a group.
  gnus-score-adaptive() converts the above alist into a list which
  contains functions in one of two forms:  either (lambda ....)
  lists or the original (tag arg) list.

    This conversion begins on line 2208 of gnus-score.el.  The test to
  convert to a (lambda ...) list is to append the scoring tag name
  to "message-header-" and then to look up the symbol in obarray via a
  call to intern.  the symbol 'message-header-thread is not predefined,
  so (intern "message-header-thread") returns a symbol that is unbound.
  This, in turn, means that the list (thread <score-delta>) is *not*
  converted to a (lambda ...) list. 
  
    Later, at line 2227 of gnus-score.el, we try to use the scoring
  functions.  We do this by  using (funcall (caar elem) headers).  This
  means that we end up trying to call the function #'thread which
  doesn't exist!

    Where is this supposed to be set up?  I did check the info files and
  you are supposed to be able to score adaptively by thread, so I think
  that this is a bug.  

-IPmonger
-- 
------------------
IPmonger
ipmonger@delamancha.org
CCIE #8338

Help with Adaptive Scoring Problems (2nd try)

[IPmonger]

Folks,

    I thought I sent this late last night, but I haven't seen it go
  through yet, so here goes...

    I used to use the following adaptive-score-alist:

(defvar gnus-default-adaptive-score-alist
  '((gnus-unread-mark (thread 10))
    (gnus-ticked-mark (from 4))
    (gnus-dormant-mark (from 5))
    (gnus-del-mark (from -4) (subject -1) (thread -1))
    (gnus-read-mark (from 4) (subject 2) (thread 25))
    (gnus-expirable-mark)
    (gnus-killed-mark (from -10) (subject -3))
    (gnus-kill-file-mark)
    (gnus-ancient-mark)
    (gnus-low-score-mark)
    (gnus-catchup-mark (from -1) (subject -1))))


    But, when I set it up this way now, the (thread <score-delta>)
  entries cause an Elisp error when I attempt to exit a group.
  gnus-score-adaptive() converts the above alist into a list which
  contains functions in one of two forms:  either (lambda ....)
  lists or the original (tag arg) list.

    This conversion begins on line 2208 of gnus-score.el.  The test to
  convert to a (lambda ...) list is to append the scoring tag name
  to "message-header-" and then to look up the symbol in obarray via a
  call to intern.  the symbol 'message-header-thread is not predefined,
  so (intern "message-header-thread") returns a symbol that is unbound.
  This, in turn, means that the list (thread <score-delta>) is *not*
  converted to a (lambda ...) list. 
  
    Later, at line 2227 of gnus-score.el, we try to use the scoring
  functions.  We do this by  using (funcall (caar elem) headers).  This
  means that we end up trying to call the function #'thread which
  doesn't exist!

    Where is this supposed to be set up?  I did check the info files and
  you are supposed to be able to score adaptively by thread, so I think
  that this is a bug.  

-IPmonger
-- 
------------------
IPmonger
ipmonger@delamancha.org
CCIE #8338