Re: Value for gnus-auto-expirable-newsgroups

Re: Value for gnus-auto-expirable-newsgroups

[Thomas Neumann]

> 	Now that we know that nnmail-expiry-wait-function should not
>  contain the method ("other" rather than "nnml:other"), does it make
>  sense to have gnus-auto-expirable-newsgroups set to "^nnml:"? 

I'm doing exactly that. Works like a charm.

-t

Re: Value for gnus-auto-expirable-newsgroups

[Manoj Srivastava]

X-Geek-3: GE/CS d+(--) s:++>: a C++++$ ULUHO++++$ P+++$ L+++ E+++ 
   W+++$ N+++ K-? !w--- O-? !M-- !V-- PS+ PE- Y+ PGP++ t@ 5++ !X R++ 
   b+++ DI+++ D- G e+++ h+ r++ y+
X-Face:  #q.#]5@vq!Jz+E0t_/;Y^gTjR\T^"B'fbeuVGiyKrvbfKJl!^e|e:iu(kJ6c|QYB57LP*|t&YlP~HF/=h:GA6o6W@I#deQL-%#.6]!z:6Cj0kd#4]>*D,|0djf'CVlXkI,>aV4\}?d_KEqsN{Nnt778"OsbQ["56/!nisvyB/uA5Q.{)gm6?q.j71ww.>b9b]-sG8zNt%KkIa>xWg&1VcjZk[hBQ>]j~`WqXl,y1a!(>6`UM{~'X[Y_,Bv+}=L\SS*mA8=s;!=O`ja|@PEzb&i0}Qp,`Z\:6:OmRi*
X-Organization: University of Massachusetts, Amherst, MA 01003
X-Time: Tue Dec  5 19:30:59 1995
X-Mail-System: Vm 5.95 (beta) for GNU Emacs 19.14 XEmacs Lucid (beta6)

Hi,
>>"Thomas" == Thomas Neumann <tom@smart.ruhr.de> writes:

 Manoj> Now that we know that nnmail-expiry-wait-function should not
 Manoj> contain the method ("other" rather than "nnml:other"), does it
 Manoj> make sense to have gnus-auto-expirable-newsgroups set to
 Manoj> "^nnml:"?

 Thomas> I'm doing exactly that. Works like a charm.

	Unfortunately, it only partially works for me.  I mean, all
 the messages I read _are_ marked with E, but any messages with low
 scores from my scoring process are marked with Y, and I have a
 feeling that they are not expired.

	manoj

-- In the future, etiquette will become more and more important.  That
 doesn't mean knowing which fork to pick up -- I mean basic
 consideration for the rights of other animals (human beings included)
 and the willingness, whenever practical, to tolerate the other guy's
 idiosyncrasies. Frank Zappa, _The Real Frank Zappa Book_

-- 
Manoj Srivastava               Systems Research Programmer, Project Pilgrim,
Phone: (413) 545-3918                A143B Lederle Graduate Research Center,
Fax:   (413) 545-1249         University of Massachusetts, Amherst, MA 01003
<srivasta@pilgrim.umass.edu> <URL:http://www.pilgrim.umass.edu/%7Esrivasta/>

Re: Value for gnus-auto-expirable-newsgroups

[Mark Borges]

>>>>> "M" == Manoj Srivastava <srivasta@pilgrim.umass.edu> writes:

    M> Hi, Now that we know that nnmail-expiry-wait-function should
    M> not contain the method ("other" rather than "nnml:other"), does

Do we? I thought it didn't matter:

>>>>> "Lars" == Lars Magne Ingebrigtsen <larsi@ifi.uio.no> writes:

    Lars> mdb@cdc.noaa.gov (Mark Borges) writes:
    >> It doesn't seem to matter -- it appears to work fine with the
    >> method left in. Should it? Lars?

    Lars> No, it shouldn't.

I took that as "it shouldn't matter".

    M> it make sense to have gnus-auto-expirable-newsgroups set to
    M> "^nnml:"?

That works for me, though I have "nnml:" in my wait function as well.

    M> 	Or do I have to maintain the list by hand? Somewhare in thed
    M> documentation I read that the forst element in the newsrc.eld
    M> is the group name, and that certainly contains nnml: prefix.

The code has this (at least for Gnus-5.0.12, I haven't tried sgnus) in
gnus.el:

      ;; Check whether auto-expire is to be done in this group.
      (setq gnus-newsgroup-auto-expire
	    (or (and (stringp gnus-auto-expirable-newsgroups)
		     (string-match gnus-auto-expirable-newsgroups group))
		(memq 'auto-expire (nth 5 info))))

so it should be OK without the method, and (if group includes the
method, which I think it does) with the method as well.

  -mb-

Re: Value for gnus-auto-expirable-newsgroups

[Steven L. Baur]

>>>>> "mb" == Mark Borges <mdb@cdc.noaa.gov> writes:

>>>>> "M" == Manoj Srivastava <srivasta@pilgrim.umass.edu> writes:
    M> Hi, Now that we know that nnmail-expiry-wait-function should
    M> not contain the method ("other" rather than "nnml:other"), does

    mb> Do we? I thought it didn't matter:

>>>>> "Lars" == Lars Magne Ingebrigtsen <larsi@ifi.uio.no> writes:

    Lars> mdb@cdc.noaa.gov (Mark Borges) writes:
    >>> It doesn't seem to matter -- it appears to work fine with the
    >>> method left in. Should it? Lars?

    Lars> No, it shouldn't.

    mb> I took that as "it shouldn't matter".

I don't want to split grammatical hairs, let's look at the source.

(The expiration selection code in sgnus v0.18 is unchanged from Gnus
5.0.12.)

As I read the code, all article expiration is done through the
function gnus-request-expire-articles, which selects the appropriate
backend expiration function, *and* calls gnus-group-real-name on the
newsgroup parameter.  Therefore, if your expiry-wait function is
getting the newsgroup with the method left in, that's a bug.


[ These are cut & pasted from 5.0.12]

[ This function gets group with the method still attached, as returned
by gnus-group-group-name. ]

(defun gnus-request-expire-articles (articles group &optional force)
  (let ((method (gnus-find-method-for-group group)))
    (funcall (gnus-get-function method 'request-expire-articles) 
             articles (gnus-group-real-name group) (nth 1 method)
             force)))


(defun gnus-group-real-name (group)
  "Find the real name of a foreign newsgroup."
  (if (string-match ":[^:]+$" group)
      (substring group (1+ (match-beginning 0)))
    group))



    M> it make sense to have gnus-auto-expirable-newsgroups set to
    M> "^nnml:"?

    mb> That works for me, though I have "nnml:" in my wait function
    mb> as well.

Are you sure your wait function really works, or is everything
expiring after the (same) default amount of time?


I match explicitly on group name sans method in my
expiry-wait-function, and I've verified that expiration does occur
based on the value returned by it.

Regards,
-- 
steve@miranova.com baur

Re: Value for gnus-auto-expirable-newsgroups

[Lars Magne Ingebrigtsen]

Manoj Srivastava <srivasta@pilgrim.umass.edu> writes:

> 	Now that we know that nnmail-expiry-wait-function should not
>  contain the method ("other" rather than "nnml:other"), does it make
>  sense to have gnus-auto-expirable-newsgroups set to "^nnml:"? 

Yes.  `nnmail-expiry-wait-function' is a function used by the
backends.  The backends have no concept of "prefixed group names".
`gnus-auto-expirable-newsgroups', on the other hand, is used by Gnus,
which uses prefixed group names exclusively.

-- 
Home is where the cat is.

Re: Value for gnus-auto-expirable-newsgroups

[Lars Magne Ingebrigtsen]

mdb@cdc.noaa.gov (Mark Borges) writes:

>     >> It doesn't seem to matter -- it appears to work fine with the
>     >> method left in. Should it? Lars?
> 
>     Lars> No, it shouldn't.
> 
> I took that as "it shouldn't matter".

I meant it as "it shouldn't work", though.  :-)

-- 
Home is where the cat is.

Value for gnus-auto-expirable-newsgroups

[Unknown]

X-Geek-3: GE/CS d+(--) s:++>: a C++++$ ULUHO++++$ P+++$ L+++ E+++ 
   W+++$ N+++ K-? !w--- O-? !M-- !V-- PS+ PE- Y+ PGP++ t@ 5++ !X R++ 
   b+++ DI+++ D- G e+++ h+ r++ y+
X-Face:  #q.#]5@vq!Jz+E0t_/;Y^gTjR\T^"B'fbeuVGiyKrvbfKJl!^e|e:iu(kJ6c|QYB57LP*|t&YlP~HF/=h:GA6o6W@I#deQL-%#.6]!z:6Cj0kd#4]>*D,|0djf'CVlXkI,>aV4\}?d_KEqsN{Nnt778"OsbQ["56/!nisvyB/uA5Q.{)gm6?q.j71ww.>b9b]-sG8zNt%KkIa>xWg&1VcjZk[hBQ>]j~`WqXl,y1a!(>6`UM{~'X[Y_,Bv+}=L\SS*mA8=s;!=O`ja|@PEzb&i0}Qp,`Z\:6:OmRi*
X-Organization: University of Massachusetts, Amherst, MA 01003
X-Time: Tue Dec  5 17:48:07 1995
X-Mail-System: Vm 5.95 (beta) for GNU Emacs 19.14 XEmacs Lucid (beta6)
From: Manoj Srivastava <srivasta@pilgrim.umass.edu>
Date: 05 Dec 1995 17:48:07 -0500
Message-ID: <gvx3fazxh3s.fsf@belthil.pilgrim.umass.edu>
Organization: Project Pilgrim, University of Massachusetts at Amherst
Lines: 21
X-Mailer: September Gnus v0.18/XEmacs 19.14

Hi,

	Now that we know that nnmail-expiry-wait-function should not
 contain the method ("other" rather than "nnml:other"), does it make
 sense to have gnus-auto-expirable-newsgroups set to "^nnml:"? 

	Or do I have to maintain the list by hand? Somewhare in thed
 documentation I read that the forst element in the newsrc.eld is the
 group name, and that certainly contains nnml: prefix. 

	confused,

		manoj

-- "From there to here, from here to there, funny things are
 everywhere." Dr. Seuss

Manoj Srivastava               Systems Research Programmer, Project Pilgrim,
Phone: (413) 545-3918                A143B Lederle Graduate Research Center,
Fax:   (413) 545-1249         University of Massachusetts, Amherst, MA 01003
<srivasta@pilgrim.umass.edu> <URL:http://www.pilgrim.umass.edu/%7Esrivasta/>