Like (display . all) but hide expirable messages

Like (display . all) but hide expirable messages

[Kai Großjohann]

People who want to use Gnus like other mail readers want Gnus to
always show all mails, except the deleted ones.  Since expirable
messages are deleted, in a sense, it would be nice to provide a
(display . most) feature which does this.

Opinions?

kai
-- 
~/.signature: No such file or directory

Re: Like (display . all) but hide expirable messages

[Jody Klymak]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> People who want to use Gnus like other mail readers want Gnus to
> always show all mails, except the deleted ones.  Since expirable
> messages are deleted, in a sense, it would be nice to provide a
> (display . most) feature which does this.

I'd like that.  It'd also be nice if something like "C-u C-u Ret" on a
group displayed "most" instead of "all" as well.

Cheers,  Jody

-- 
Jody M. Klymak               mailto:jklymak@oce.orst.edu
College of Oceanic and Atmospheric Sicences
Oregon State University, Corvallis, OR, 97331  

Re: Like (display . all) but hide expirable messages

[Nuutti Kotivuori]

Kai Großjohann wrote:
> People who want to use Gnus like other mail readers want Gnus to
> always show all mails, except the deleted ones.  Since expirable
> messages are deleted, in a sense, it would be nice to provide a
> (display . most) feature which does this.
> 
> Opinions?

Ve-ry useful. My friend was whining just about that. Except that he
decided to just set expiry wait to 0.1 or something.

-- Naked

Re: Like (display . all) but hide expirable messages

[Leonard Stiles]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> People who want to use Gnus like other mail readers want Gnus to
> always show all mails, except the deleted ones.  Since expirable
> messages are deleted, in a sense, it would be nice to provide a
> (display . most) feature which does this.

Or make display more flexible: let it take a list of symbols
representing article marks.  If the car of the list is 'not, the sense
of the match is inverted.

e.g. (display . (not read expirable)) to get the behaviour you suggest.

-- 

Leonard Stiles <ljs@uk2.net>

Re: Like (display . all) but hide expirable messages

[Paul Jarc]

Leonard Stiles <ljs@uk2.net> writes:
> Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
>> People who want to use Gnus like other mail readers want Gnus to
>> always show all mails, except the deleted ones.  Since expirable
>> messages are deleted, in a sense, it would be nice to provide a
>> (display . most) feature which does this.
> 
> Or make display more flexible: let it take a list of symbols
> representing article marks.  If the car of the list is 'not, the sense
> of the match is inverted.
> 
> e.g. (display . (not read expirable)) to get the behaviour you suggest.

To take it even further, e.g.:
(display . (and (not expire) (or tick (not read))))
But this might be more than is worthwhile.


paul

Re: Like (display . all) but hide expirable messages

[Lars Magne Ingebrigtsen]

prj@po.cwru.edu (Paul Jarc) writes:

> To take it even further, e.g.:
> (display . (and (not expire) (or tick (not read))))

Neat.

> But this might be more than is worthwhile.

Has that ever stopped us before?  :-)

So, the question is -- what marks should be predicatable?  (If that's
not a word, I just made it one.)

These are the ones that are stored in the .newsrc.eld file:

(defconst gnus-article-mark-lists
  '((marked . tick) (replied . reply)
    (expirable . expire) (killed . killed)
    (bookmarks . bookmark) (dormant . dormant)
    (scored . score) (saved . save)
    (cached . cache) (downloadable . download)
    (unsendable . unsend) (forwarded . forward)
    (recent . recent)))

I could make a general predicate that takes an article number and a
type, but the function would just be one big cond, or I could define

gnus-article-marked-p, gnus-article-replied-p, gnus-article-expirable-p

and so on.

If a general predicate (which might actually be easier to use, in this
case), it would be called `gnus-article-marked-p'.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Like (display . all) but hide expirable messages

[Kai Großjohann]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> So, the question is -- what marks should be predicatable?  (If that's
> not a word, I just made it one.)

All of them, of course.  Everything else would not be in the spirit of
Gnus.

kai
-- 
~/.signature: No such file or directory

Re: Like (display . all) but hide expirable messages

[Lars Magne Ingebrigtsen]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

>> So, the question is -- what marks should be predicatable?  (If that's
>> not a word, I just made it one.)
>
> All of them, of course.  Everything else would not be in the spirit of
> Gnus.

Oh, sure.  I was just wondering whether we wanted more than the saved
ticks.  Like -- low-scored articles, "old articles" (O-marked), and so
on.  Or not.  (I didn't say what I was wondering about, though.)

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Like (display . all) but hide expirable messages

[Kai Großjohann]

Leonard Stiles <ljs@uk2.net> writes:

> e.g. (display . (not read expirable)) to get the behaviour you suggest.

I'm suggesting (display . (not expirable)).

I like your suggestion.  I also like Paul's.  Hm.

kai
-- 
~/.signature: No such file or directory

Re: Like (display . all) but hide expirable messages

[Lars Magne Ingebrigtsen]

Leonard Stiles <ljs@uk2.net> writes:

> e.g. (display . (not read expirable)) to get the behaviour you suggest.

That would be a syntax problem:

'(display . (not read expirable))
=> (display not read expirable)

So

'(display . (read))
=> (display read)

which means "set the variable `display' to the symbol `read'"; not
"set the `display' group parameter".

Anyway.  This means that the syntax would have to be something
special-cased, like

(display (param (read)))

where we reserve `param' as our symbol.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Like (display . all) but hide expirable messages

[Kai Großjohann]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Leonard Stiles <ljs@uk2.net> writes:
>
>> e.g. (display . (not read expirable)) to get the behaviour you suggest.
>
> That would be a syntax problem:
>
> '(display . (not read expirable))
> => (display not read expirable)

Ah.  (display . [not read expirable])

kai
-- 
~/.signature: No such file or directory

Re: Like (display . all) but hide expirable messages

[Lars Magne Ingebrigtsen]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> Ah.  (display . [not read expirable])

That's a syntax.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Like (display . all) but hide expirable messages

[Lars Magne Ingebrigtsen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

>> Ah.  (display . [not read expirable])
>
> That's a syntax.

Which has now been implemented and documented.  (And even tested!)  In
CVS in a few jiffies.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Like (display . all) but hide expirable messages

[Dan Christensen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
>
>>> Ah.  (display . [not read expirable])
>>
>> That's a syntax.
>
> Which has now been implemented and documented.  (And even tested!)  In
> CVS in a few jiffies.

Excellent, that's great news.  Things work fine for me with

  (display . [not expire])

Here are some problems/suggestions, though.

1. The docs mention both expire and expirable, but only expire works
   for me.  Both should probably be permitted.

2. The docs should say this feature is implemented by limiting the
   summary buffer, and that this means that `/ w' can be used to
   reveal the hidden articles.

3. Would it be reasonable to have a command and keystroke that applies
   the group's (display . foo) setting to the summary buffer as it
   now stands?  For example, if while reading a group some of
   the articles become marked expirable, I sometimes want to
   hide them to get rid of clutter.

4. Searching the index for the word display doesn't turn up the
   Group Parameters node.

5. While in the group parameters editing screen (`G p'),
   `C-c C-i' doesn't take you to the Group Parameters node.

6. The info page includes three empty predicates:

     `read', `dormant', `', `', `', `expire', `reply', `killed',

7. I can't get (display . [and (not expire) (not read)]) or
   (display . [and (not expire) unread]) to work for me.  With
   the former I get prompted to input a Lisp Expression when I
   enter a group.  With the latter I get the following error when 
   entering a group:

Debugger entered--Lisp error: (void-function unread)
  unread()
  #[nil "ÀÁ!?…	\0Â ‡" [gnus-article-marked-p expire unread] 2]()
  funcall(#[nil "ÀÁ!?…	\0Â ‡" [gnus-article-marked-p expire unread] 2])
  (not (funcall gnus-newsgroup-display))
  (and gnus-newsgroup-display (not (eq gnus-newsgroup-display ...)) (not (funcall gnus-newsgroup-display)))
  (or (and (memq number gnus-newsgroup-dormant) (zerop children)) (and (or ... ...) (gnus-summary-article-ancient-p number) (zerop children)) (and (eq gnus-fetch-old-headers ...) (gnus-summary-article-ancient-p number)) (and (eq gnus-build-sparse-threads ...) (gnus-summary-article-sparse-p number) (zerop children)) (when (and gnus-summary-expunge-below ...) (incf gnus-newsgroup-expunged-tally) (when ... ... ...) t) (and gnus-newsgroup-display (not ...) (not ...)) (if (and gnus-use-nocem ...) (progn ... t)))
  (and (not (memq number gnus-newsgroup-marked)) (or (and ... ...) (and ... ... ...) (and ... ...) (and ... ... ...) (when ... ... ... t) (and gnus-newsgroup-display ... ...) (if ... ...)))
  (if (and (not ...) (or ... ... ... ... ... ... ...)) 0 (push number gnus-newsgroup-limit) 1)
  (let ((children ...) (number ...) score) (if (and ... ...) 0 (push number gnus-newsgroup-limit) 1))
  (progn (let (... ... score) (if ... 0 ... 1)))
  (if thread (progn (let ... ...)))
  (when thread (let (... ... score) (if ... 0 ... 1)))
  gnus-summary-limit-children(([9758 "SHARCNET at UWO" "Allan B MacIsaac <allanb@cmrg.apmaths.uwo.ca>" "Mon, 30 Jul 2001 10:38:44 -0400 (EDT)" "<Pine.LNX.4.10.10107301011230.9162-100000@cmrg.apmaths.uwo.ca>" "<Pine.SOL.4.21.0107301003180.10716-100000@panther.uwo.ca>" 4444 42 "scratchy generic:9758" (...)]))
  mapcar(gnus-summary-limit-children (([9758 "SHARCNET at UWO" "Allan B MacIsaac <allanb@cmrg.apmaths.uwo.ca>" "Mon, 30 Jul 2001 10:38:44 -0400 (EDT)" "<Pine.LNX.4.10.10107301011230.9162-100000@cmrg.apmaths.uwo.ca>" "<Pine.SOL.4.21.0107301003180.10716-100000@panther.uwo.ca>" 4444 42 "scratchy generic:9758" ...])))
  (apply (quote +) (mapcar (quote gnus-summary-limit-children) (cdr thread)))
  (if (cdr thread) (apply (quote +) (mapcar ... ...)) 0)
  (let ((children ...) (number ...) score) (if (and ... ...) 0 (push number gnus-newsgroup-limit) 1))
  (progn (let (... ... score) (if ... 0 ... 1)))
  (if thread (progn (let ... ...)))
  (when thread (let (... ... score) (if ... 0 ... 1)))
  gnus-summary-limit-children(([9757 "Re: SHARCNET Fellowships: Call for proposals" "Cindy Munro <cindy@uwo.ca>" "Mon, 30 Jul 2001 10:07:05 -0400 (EDT)" "<Pine.SOL.4.21.0107301003180.10716-100000@panther.uwo.ca>" "<3B6067CD.4E37C10A@cmrg.apmaths.uwo.ca>" 4777 64 "scratchy generic:9757" (...)] ([9758 "SHARCNET at UWO" "Allan B MacIsaac <allanb@cmrg.apmaths.uwo.ca>" "Mon, 30 Jul 2001 10:38:44 -0400 (EDT)" "<Pine.LNX.4.10.10107301011230.9162-100000@cmrg.apmaths.uwo.ca>" "<Pine.SOL.4.21.0107301003180.10716-100000@panther.uwo.ca>" 4444 42 "scratchy generic:9758" ...])))
  mapcar(gnus-summary-limit-children (([9757 "Re: SHARCNET Fellowships: Call for proposals" "Cindy Munro <cindy@uwo.ca>" "Mon, 30 Jul 2001 10:07:05 -0400 (EDT)" "<Pine.SOL.4.21.0107301003180.10716-100000@panther.uwo.ca>" "<3B6067CD.4E37C10A@cmrg.apmaths.uwo.ca>" 4777 64 "scratchy generic:9757" ...] ([9758 "SHARCNET at UWO" "Allan B MacIsaac <allanb@cmrg.apmaths.uwo.ca>" "Mon, 30 Jul 2001 10:38:44 -0400 (EDT)" "<Pine.LNX.4.10.10107301011230.9162-100000@cmrg.apmaths.uwo.ca>" "<Pine.SOL.4.21.0107301003180.10716-100000@panther.uwo.ca>" 4444 42 "scratchy generic:9758" ...]))))
  (apply (quote +) (mapcar (quote gnus-summary-limit-children) (cdr thread)))
  (if (cdr thread) (apply (quote +) (mapcar ... ...)) 0)
  (let ((children ...) (number ...) score) (if (and ... ...) 0 (push number gnus-newsgroup-limit) 1))
  (progn (let (... ... score) (if ... 0 ... 1)))
  (if thread (progn (let ... ...)))
  (when thread (let (... ... score) (if ... 0 ... 1)))
  gnus-summary-limit-children(([9710 "SHARCNET Fellowships: Call for proposals" "\"Peter H. Poole\" <poole@cmrg.apmaths.uwo.ca>" "Thu, 26 Jul 2001 14:56:13 -0400" "<3B6067CD.4E37C10A@cmrg.apmaths.uwo.ca>" "<382B7631.C041AD59@cmrg.apmaths.uwo.ca> <3B49C29F.D31A0D51@cmrg.apmaths.uwo.ca>" 168256 2309 "scratchy generic:9710" (...)] ([9757 "Re: SHARCNET Fellowships: Call for proposals" "Cindy Munro <cindy@uwo.ca>" "Mon, 30 Jul 2001 10:07:05 -0400 (EDT)" "<Pine.SOL.4.21.0107301003180.10716-100000@panther.uwo.ca>" "<3B6067CD.4E37C10A@cmrg.apmaths.uwo.ca>" 4777 64 "scratchy generic:9757" ...] ([9758 "SHARCNET at UWO" "Allan B MacIsaac <allanb@cmrg.apmaths.uwo.ca>" "Mon, 30 Jul 2001 10:38:44 -0400 (EDT)" "<Pine.LNX.4.10.10107301011230.9162-100000@cmrg.apmaths.uwo.ca>" "<Pine.SOL.4.21.0107301003180.10716-100000@panther.uwo.ca>" 4444 42 "scratchy generic:9758" ...]))))
  (if (and gnus-thread-expunge-below (< ... gnus-thread-expunge-below)) (gnus-expunge-thread (pop nodes)) (setq thread (pop nodes)) (gnus-summary-limit-children thread))
  (while nodes (if (and gnus-thread-expunge-below ...) (gnus-expunge-thread ...) (setq thread ...) (gnus-summary-limit-children thread)))
  (let ((nodes ...) thread) (while nodes (if ... ... ... ...)))
  (if (car (symbol-value node)) nil (let (... thread) (while nodes ...)))
  (unless (car (symbol-value node)) (let (... thread) (while nodes ...)))
  (lambda (node) (unless (car ...) (let ... ...)))(<3B49C29F\.D31A0D51@cmrg\.apmaths\.uwo\.ca>)
  mapatoms((lambda (node) (unless (car ...) (let ... ...))) [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 <3206878724608364@naggum\.net> <200108170928\.DAA14434@xxx\.lanl\.gov> 0 0 <0997629519\.0750506837@inbound\.canada\.com\.criticalpath\.net> 0 0 0 0 <Pine\.GSO\.4\.21\.0108151213470\.389-100000@Zathras\.Stanford\.EDU> 0 0 <5bbslmradh\.fsf@lister\.idonex\.se> 0 0 0 0 0 0 0 0 0 <vafsneojiah\.fsf@INBOX\.auto\.emacs\.pretesters\.tok\.lucy\.cs\.uni-dortmund\.de> 0 0 0 0 <15229\.29817\.637021\.918722@eps\.austin\.ibm\.com> 0 none 0 0 0 0 ...])
  (if (or gnus-inhibit-limiting (and ... ... ... ... ... ... ... ... ... ...)) nil (push gnus-newsgroup-limit gnus-newsgroup-limits) (setq gnus-newsgroup-limit nil) (mapatoms (lambda ... ...) gnus-newsgroup-dependencies) (when (and ... show-if-empty) (setq gnus-newsgroup-limit ...)) gnus-newsgroup-limit)
  gnus-summary-initial-limit(200)
  (let ((gnus-newsgroup-dormant nil)) (gnus-summary-initial-limit show-all))
  (if show-all (let (...) (gnus-summary-initial-limit show-all)) (gnus-summary-initial-limit show-all))
  (if gnus-show-threads (if show-all (let ... ...) (gnus-summary-initial-limit show-all)) (setq gnus-newsgroup-limit (mapcar ... gnus-newsgroup-headers)))
  (cond ((not new-group) (gnus-set-global-variables) (when kill-buffer ...) (gnus-configure-windows ... ...) (gnus-set-mode-line ...) (gnus-summary-position-point) (message "") t) ((null did-select) (when ... ... ...) (let ... ...) nil) ((eq did-select ...) (and ... ... ...) (when kill-buffer ...) (if ... ... ...) (signal ... nil)) (t (gnus-set-global-variables) (setq gnus-newsgroup-active ...) (gnus-run-hooks ...) (gnus-update-format-specifications nil ... ... ...) (gnus-update-summary-mark-positions) (when gnus-use-scoring ...) (when gnus-build-sparse-threads ...) (if gnus-show-threads ... ...) (unless no-display ...) (when gnus-use-trees ... ...) (when ... ...) (gnus-run-hooks ...) (if ... ... ... ... ... ... ... ... t)))
  (let* ((new-group ...) (quit-config ...) (did-select ...)) (cond (... ... ... ... ... ... ... t) (... ... ... nil) (... ... ... ... ...) (t ... ... ... ... ... ... ... ... ... ... ... ... ...)))
  gnus-summary-read-group-1("nnfolder:generic" 200 nil nil nil nil)
  (or (gnus-summary-read-group-1 group show-all no-article kill-buffer no-display select-articles) (setq show-all nil select-articles nil))
  (let ((gnus-auto-select-next nil)) (or (gnus-summary-read-group-1 group show-all no-article kill-buffer no-display select-articles) (setq show-all nil select-articles nil)))
  (setq result (let (...) (or ... ...)))
  (null (setq result (let ... ...)))
  (and group (null (setq result ...)) (eq gnus-auto-select-next (quote quietly)))
  (while (and group (null ...) (eq gnus-auto-select-next ...)) (set-buffer gnus-group-buffer) (when backward (gnus-group-prev-unread-group 2)) (if (not ...) (setq group ...) (setq group nil)))
  (let (result) (while (and group ... ...) (set-buffer gnus-group-buffer) (when backward ...) (if ... ... ...)) result)
  gnus-summary-read-group("nnfolder:generic" 200 nil nil nil nil nil)
  gnus-group-read-group(200 nil nil)
  gnus-topic-read-group(200)
  call-interactively(gnus-topic-read-group)

-- 
Dan Christensen
jdc+news@uwo.ca

Re: Like (display . all) but hide expirable messages

[Lars Magne Ingebrigtsen]

Dan Christensen <jdc+news@uwo.ca> writes:

> 1. The docs mention both expire and expirable, but only expire works
>    for me.  Both should probably be permitted.

No, only the mark name is permitted.  (`expire' is the name, and what
it does is mark things as expirable.  :-)  I've now fixed the
documentation. 

> 2. The docs should say this feature is implemented by limiting the
>    summary buffer, and that this means that `/ w' can be used to
>    reveal the hidden articles.

I've now done so.

> 3. Would it be reasonable to have a command and keystroke that applies
>    the group's (display . foo) setting to the summary buffer as it
>    now stands?  For example, if while reading a group some of
>    the articles become marked expirable, I sometimes want to
>    hide them to get rid of clutter.

Sure.  I've now added this.  `/ p'.

> 4. Searching the index for the word display doesn't turn up the
>    Group Parameters node.

Fixed.

> 5. While in the group parameters editing screen (`G p'),
>    `C-c C-i' doesn't take you to the Group Parameters node.

Ditto.

> 6. The info page includes three empty predicates:
>
>      `read', `dormant', `', `', `', `expire', `reply', `killed',

Even more ditto.

> 7. I can't get (display . [and (not expire) (not read)]) or
>    (display . [and (not expire) unread]) to work for me.  With
>    the former I get prompted to input a Lisp Expression when I
>    enter a group.

So much ditto that I don't know how to express it.

>    With the latter I get the following error when 
>    entering a group:
>
> Debugger entered--Lisp error: (void-function unread)
>   unread()

I've now made that invalid and removed it from the documentation.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Like (display . all) but hide expirable messages

[Dan Christensen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Dan Christensen <jdc+news@uwo.ca> writes:
>
>> 3. Would it be reasonable to have a command and keystroke that applies
>>    the group's (display . foo) setting to the summary buffer as it
>>    now stands?  For example, if while reading a group some of
>>    the articles become marked expirable, I sometimes want to
>>    hide them to get rid of clutter.
>
> Sure.  I've now added this.  `/ p'.

Doing `/ w' and `/ p' leaves the dormant articles in the display.

>> 5. While in the group parameters editing screen (`G p'),
>>    `C-c C-i' doesn't take you to the Group Parameters node.

Starting from the group buffer, `G p C-c C-i' gives me:

Wrong type argument: commandp, (lambda nil "Enter the info system at node (gnus)Group Parameters" (Info-goto-node "(gnus)Group Parameters") (setq gnus-info-buffer (current-buffer)) (gnus-configure-windows (quote info)))

>> 7. I can't get (display . [and (not expire) (not read)]) or
>>    (display . [and (not expire) unread]) to work for me. 
>>    With the latter I get the following error when 
>>    entering a group:
>>
>> Debugger entered--Lisp error: (void-function unread)
>>   unread()
>
> I've now made that invalid and removed it from the documentation.

Except for one example in that node...  :-)

Thanks for the very useful functionality!  

-- 
Dan Christensen
jdc+news@uwo.ca

Re: Like (display . all) but hide expirable messages

[ShengHuo ZHU]

Dan Christensen <jdc+news@uwo.ca> writes:

> Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
>
>> Dan Christensen <jdc+news@uwo.ca> writes:

[...]

>>> 5. While in the group parameters editing screen (`G p'),
>>>    `C-c C-i' doesn't take you to the Group Parameters node.
>
> Starting from the group buffer, `G p C-c C-i' gives me:
>
> Wrong type argument: commandp, (lambda nil "Enter the info system at node (gnus)Group Parameters" (Info-goto-node "(gnus)Group Parameters") (setq gnus-info-buffer (current-buffer)) (gnus-configure-windows (quote info)))

Fixed.

ShengHuo

Re: Like (display . all) but hide expirable messages

[Lars Magne Ingebrigtsen]

Dan Christensen <jdc+news@uwo.ca> writes:

> Doing `/ w' and `/ p' leaves the dormant articles in the display.

What `display' parameter do you use?

>> I've now made that invalid and removed it from the documentation.
>
> Except for one example in that node...  :-)

Oops.  :-)  Fixed.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Like (display . all) but hide expirable messages

[Dan Christensen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Dan Christensen <jdc+news@uwo.ca> writes:
>
>> Doing `/ w' and `/ p' leaves the dormant articles in the display.
>
> What `display' parameter do you use?

(display . [not expire])

But now I'm confused about how this is all supposed to work.
If I enter a group with SPC, I don't see expirable articles
or dormant articles.  If I hit `/ w' the dormants are added
to the summary, but I still don't see the expirables.  And
there are no more limits to pop.

If I enter a group with `99 SPC', say, then I see the dormants
already, and not the expirables.  And then '/ w' will bring
the expirables in.

Maybe this is how things are supposed to work?

-- 
Dan Christensen
jdc+news@uwo.ca

Re: Like (display . all) but hide expirable messages

[Björn Torkelsson]

Dan Christensen <jdc+news@uwo.ca> writes:

> Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
>
>> Dan Christensen <jdc+news@uwo.ca> writes:
>>
>>> Doing `/ w' and `/ p' leaves the dormant articles in the display.
>>
>> What `display' parameter do you use?
>
> (display . [not expire])
>
> But now I'm confused about how this is all supposed to work.
> If I enter a group with SPC, I don't see expirable articles
> or dormant articles.  If I hit `/ w' the dormants are added
> to the summary, but I still don't see the expirables.  And
> there are no more limits to pop.
>
> If I enter a group with `99 SPC', say, then I see the dormants
> already, and not the expirables.  And then '/ w' will bring
> the expirables in.
>
> Maybe this is how things are supposed to work?

I sure don't hope so. ;-)

With the following group parameters:

((uidvalidity . "991199422")
 (expiry-wait . 7)
 (visible . t)
 (display . [not expire]))

and with new messages waiting in the group I only see the new messages,
not the old ones. 

/torkel

Re: Like (display . all) but hide expirable messages

[Dan Christensen]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Dan Christensen <jdc+news@uwo.ca> writes:
>
>> 7. I can't get (display . [and (not expire) (not read)]) or
>>    (display . [and (not expire) unread]) to work for me.  With
>>    the former I get prompted to input a Lisp Expression when I
>>    enter a group.
>
> So much ditto that I don't know how to express it.
>
>>    With the latter I get the following error when 
>>    entering a group:
>>
>> Debugger entered--Lisp error: (void-function unread)
>>   unread()
>
> I've now made that invalid and removed it from the documentation.

This still doesn't work for me, with a cvs from an hour ago.  (not
read) checks that the article doesn't appear in gnus-newsgroup-reads,
but this variable seems to keep track of the articles read in the
*current* session, so I think it is always nil when the summary is
being generated.  What I want is to check that the article is in
gnus-newsgroup-unreads.  So I think read should be made invalid, and
unread should be made valid, since it is what one is actually
interested in.

Dan

-- 
Dan Christensen
jdc+news@uwo.ca

Re: Like (display . all) but hide expirable messages

[ShengHuo ZHU]

Dan Christensen <jdc+news@uwo.ca> writes:

> Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
>
>> Dan Christensen <jdc+news@uwo.ca> writes:
>>
>>> 7. I can't get (display . [and (not expire) (not read)]) or
>>>    (display . [and (not expire) unread]) to work for me.  With
>>>    the former I get prompted to input a Lisp Expression when I
>>>    enter a group.
>>
>> So much ditto that I don't know how to express it.
>>
>>>    With the latter I get the following error when 
>>>    entering a group:
>>>
>>> Debugger entered--Lisp error: (void-function unread)
>>>   unread()
>>
>> I've now made that invalid and removed it from the documentation.
>
> This still doesn't work for me, with a cvs from an hour ago.  (not
> read) checks that the article doesn't appear in gnus-newsgroup-reads,
> but this variable seems to keep track of the articles read in the
> *current* session, so I think it is always nil when the summary is
> being generated.  What I want is to check that the article is in
> gnus-newsgroup-unreads.  So I think read should be made invalid, and
> unread should be made valid, since it is what one is actually
> interested in.

I added unread.

ShengHuo

Re: Like (display . all) but hide expirable messages

[Dan Christensen]

ShengHuo ZHU <zsh@cs.rochester.edu> writes:

> Dan Christensen <jdc+news@uwo.ca> writes:
>
>> Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
>>
>>> Dan Christensen <jdc+news@uwo.ca> writes:
>>>
>>>> 7. I can't get (display . [and (not expire) (not read)]) or
>>>>    (display . [and (not expire) unread]) to work for me.  With
>>>>    the former I get prompted to input a Lisp Expression when I
>>>>    enter a group.
>>>
>>> So much ditto that I don't know how to express it.
>>>
>>>>    With the latter I get the following error when 
>>>>    entering a group:
>>>>
>>>> Debugger entered--Lisp error: (void-function unread)
>>>>   unread()
>>>
>>> I've now made that invalid and removed it from the documentation.
>>
>> This still doesn't work for me, with a cvs from an hour ago.  (not
>> read) checks that the article doesn't appear in gnus-newsgroup-reads,
>> but this variable seems to keep track of the articles read in the
>> *current* session, so I think it is always nil when the summary is
>> being generated.  What I want is to check that the article is in
>> gnus-newsgroup-unreads.  So I think read should be made invalid, and
>> unread should be made valid, since it is what one is actually
>> interested in.
>
> I added unread.

It still doesn't work for me, with the latest cvs.

Debugger entered--Lisp error: (void-function unread)
  unread()
  gnus-summary-limit-children(([16512 "Re: expiring by copying instead of by removing" "Kai Grossjohann <grossjohann@ls6.cs.uni-dortmund.de>" "13 Sep 1997 17:19:10 +0200" "<vafn2lh44y9.fsf@ls6.cs.uni-dortmund.de>" "<52rab6iu2f.fsf@sean.ebone.net> <w8su3g1mavw.fsf@levding.ifi.uio.no> <kigiuwhm9ce.fsf@jagor.srce.hr> <vafrab5gmaf.fsf@ls6.cs.uni-dortmund.de> <kig7mcxm4rm.fsf@jagor.srce.hr> <m2d8mm11cz.fsf@proletcult.slip.ifi.uio.no> <52sovgsv52.fsf@sean.ebone.net> <87oh64nukw.fsf@isdn-uhl4.rz.uni-karlsruhe.de> <m24t7vvd1x.fsf@proletcult.slip.ifi.uio.no> <87d8mea3n7.fsf@isdn-uhl4.rz.uni-karlsruhe.de> <m2g1rauw59.fsf@proletcult.slip.ifi.uio.no> <87wwkl341m.fsf@isdn-uhl4.rz.uni-karlsruhe.de>" 2032 21 "scratchy ding:16512" (...)]))
  #[(node) "\bJ@?…7\0\bJAÄ\x19\x1a\n…6\0\vƒ)\0Å\n@!\vWƒ)\0Æ\n‰A\x12@!ˆ‚\r\0\n‰A\x12@\x11Ç	!ˆ‚\r\0*‡" [node thread nodes gnus-thread-expunge-below nil gnus-thread-total-score gnus-expunge-thread gnus-summary-limit-children] 3](<87wwkl341m\.fsf@isdn-uhl4\.rz\.uni-karlsruhe\.de>)
  mapatoms(#[(node) "\bJ@?…7\0\bJAÄ\x19\x1a\n…6\0\vƒ)\0Å\n@!\vWƒ)\0Æ\n‰A\x12@!ˆ‚\r\0\n‰A\x12@\x11Ç	!ˆ‚\r\0*‡" [node thread nodes gnus-thread-expunge-below nil gnus-thread-total-score gnus-expunge-thread gnus-summary-limit-children] 3] [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ...])
  gnus-summary-initial-limit(nil)
  gnus-summary-read-group-1("nnfolder:ding" nil nil nil nil nil)
  gnus-summary-read-group("nnfolder:ding" nil nil nil nil nil nil)
  gnus-group-read-group(nil nil nil)
  gnus-topic-read-group(nil)
  call-interactively(gnus-topic-read-group)

-- 
Dan Christensen
jdc+news@uwo.ca

Re: Like (display . all) but hide expirable messages

[ShengHuo ZHU]

Dan Christensen <jdc+news@uwo.ca> writes:

> ShengHuo ZHU <zsh@cs.rochester.edu> writes:
>
>> Dan Christensen <jdc+news@uwo.ca> writes:
>>
>>> Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
>>>
>>>> Dan Christensen <jdc+news@uwo.ca> writes:
>>>>
>>>>> 7. I can't get (display . [and (not expire) (not read)]) or
>>>>>    (display . [and (not expire) unread]) to work for me.  With
>>>>>    the former I get prompted to input a Lisp Expression when I
>>>>>    enter a group.
>>>>
>>>> So much ditto that I don't know how to express it.
>>>>
>>>>>    With the latter I get the following error when 
>>>>>    entering a group:
>>>>>
>>>>> Debugger entered--Lisp error: (void-function unread)
>>>>>   unread()
>>>>
>>>> I've now made that invalid and removed it from the documentation.
>>>
>>> This still doesn't work for me, with a cvs from an hour ago.  (not
>>> read) checks that the article doesn't appear in gnus-newsgroup-reads,
>>> but this variable seems to keep track of the articles read in the
>>> *current* session, so I think it is always nil when the summary is
>>> being generated.  What I want is to check that the article is in
>>> gnus-newsgroup-unreads.  So I think read should be made invalid, and
>>> unread should be made valid, since it is what one is actually
>>> interested in.
>>
>> I added unread.
>
> It still doesn't work for me, with the latest cvs.

What are the evaluations of gnus-summary-display-cache and
gnus-category-predicate-cache ?

ShengHuo

Re: Like (display . all) but hide expirable messages

[Dan Christensen]

ShengHuo ZHU <zsh@cs.rochester.edu> writes:

> Dan Christensen <jdc+news@uwo.ca> writes:
>
>> ShengHuo ZHU <zsh@cs.rochester.edu> writes:
>>
>>> Dan Christensen <jdc+news@uwo.ca> writes:
>>>
>>>> Lars Magne Ingebrigtsen <larsi@gnus.org> writes:
>>>>
>>>>> Dan Christensen <jdc+news@uwo.ca> writes:
>>>>>
>>>>>> 7. I can't get (display . [and (not expire) (not read)]) or
>>>>>>    (display . [and (not expire) unread]) to work for me.  With
>>>>>>    the former I get prompted to input a Lisp Expression when I
>>>>>>    enter a group.
>>>>>
>>>>> So much ditto that I don't know how to express it.
>>>>>
>>>>>>    With the latter I get the following error when 
>>>>>>    entering a group:
>>>>>>
>>>>>> Debugger entered--Lisp error: (void-function unread)
>>>>>>   unread()
>>>>>
>>>>> I've now made that invalid and removed it from the documentation.
>>>>
>>>> This still doesn't work for me, with a cvs from an hour ago.  (not
>>>> read) checks that the article doesn't appear in gnus-newsgroup-reads,
>>>> but this variable seems to keep track of the articles read in the
>>>> *current* session, so I think it is always nil when the summary is
>>>> being generated.  What I want is to check that the article is in
>>>> gnus-newsgroup-unreads.  So I think read should be made invalid, and
>>>> unread should be made valid, since it is what one is actually
>>>> interested in.
>>>
>>> I added unread.
>>
>> It still doesn't work for me, with the latest cvs.
>
> What are the evaluations of gnus-summary-display-cache and
> gnus-category-predicate-cache ?

Here they are.  The first doesn't mention unread.

gnus-summary-display-cache's value is shown below.

Documentation:
not documented as a variable.

Defined in `gnus-sum'.

Value:
((seen .
       #[nil "\300\301!\207"
	     [gnus-article-marked-p seen]
	     2])
 (recent .
	 #[nil "\300\301!\207"
	       [gnus-article-marked-p recent]
	       2])
 (forward .
	  #[nil "\300\301!\207"
		[gnus-article-marked-p forward]
		2])
 (unsend .
	 #[nil "\300\301!\207"
	       [gnus-article-marked-p unsend]
	       2])
 (download .
	   #[nil "\300\301!\207"
		 [gnus-article-marked-p download]
		 2])
 (cache .
	#[nil "\300\301!\207"
	      [gnus-article-marked-p cache]
	      2])
 (save .
       #[nil "\300\301!\207"
	     [gnus-article-marked-p save]
	     2])
 (score .
	#[nil "\300\301!\207"
	      [gnus-article-marked-p score]
	      2])
 (dormant .
	  #[nil "\300\301!\207"
		[gnus-article-marked-p dormant]
		2])
 (bookmark .
	   #[nil "\300\301!\207"
		 [gnus-article-marked-p bookmark]
		 2])
 (killed .
	 #[nil "\300\301!\207"
	       [gnus-article-marked-p killed]
	       2])
 (expire .
	 #[nil "\300\301!\207"
	       [gnus-article-marked-p expire]
	       2])
 (reply .
	#[nil "\300\301!\207"
	      [gnus-article-marked-p reply]
	      2])
 (tick .
       #[nil "\300\301!\207"
	     [gnus-article-marked-p tick]
	     2])
 (unseen .
	 #[nil "\300\301!\207"
	       [gnus-article-marked-p unseen]
	       2])
 (read .
       #[nil "\300\301!\207"
	     [gnus-article-marked-p read]
	     2]))

gnus-category-predicate-cache's value is shown below.

Documentation:
not documented as a variable.

Defined in `gnus-agent'.

Value:
(((not expire)
  .
  #[nil "\300\301!?\207"
	[gnus-article-marked-p expire]
	2])
 ((and
   (not expire)
   (not read))
  .
  #[nil "\300\301!?\205\v
	[gnus-article-marked-p expire read]
	2])
 ((not read)
  .
  #[nil "\300\301!?\207"
	[gnus-article-marked-p read]
	2])
 (unread . unread)
 ((and
   (not expire)
   unread)
  .
  #[nil "\300\301!?\205	
	[gnus-article-marked-p expire unread]
	2])
 ((and
   (not expire)
   (unread))
  .
  #[nil "\300\301!?\205	
	[gnus-article-marked-p expire nil]
	2]))

-- 
Dan Christensen
jdc+news@uwo.ca

Re: Like (display . all) but hide expirable messages

[ShengHuo ZHU]

Dan Christensen <jdc+news@uwo.ca> writes:


[...]

>> What are the evaluations of gnus-summary-display-cache and
>> gnus-category-predicate-cache ?
>
> Here they are.  The first doesn't mention unread.
>
> gnus-summary-display-cache's value is shown below.

Emm. I've added a patch into the CVS. Check out it and try again. If
it doesn't work, try

   (setq gnus-summary-display-cache  nil
         gnus-category-predicate-cache nil)

ShengHuo

Re: Like (display . all) but hide expirable messages

[Dan Christensen]

ShengHuo ZHU <zsh@cs.rochester.edu> writes:

> Emm. I've added a patch into the CVS. Check out it and try again. If
> it doesn't work, try
>
>    (setq gnus-summary-display-cache  nil
>          gnus-category-predicate-cache nil)

Thanks,

 (display .
	  [and
	   (not expire)
	   unread])

now works perfectly for me!

Dan

-- 
Dan Christensen
jdc+news@uwo.ca

Re: Like (display . all) but hide expirable messages

[Leonard Stiles]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> > e.g. (display . (not read expirable)) to get the behaviour you suggest.
> 
> That would be a syntax problem:
> 
> '(display . (read))
> => (display read)
> 
> which means "set the variable `display' to the symbol `read'"; not
> "set the `display' group parameter".

Ah, yes.  However:

(Info-goto-node "(gnus)Group Parameters") =>

    `posting-style'
     [...]

     For instance, if you want a funky name and signature in this
     group only, instead of hacking `gnus-posting-styles', you could
     put something like this in the group parameters:

          (posting-style
            (name "Funky Name")
            (signature "Funky Signature"))

Which does not set the variable `posting-style'.  (I suppose you could
argue that this is because ((name "Funky Name") (signature "Funky
Signature")) is not an eval-able form).

Also, to be pedantic, in the above example the variable `display'
would be set to the value of the variable `read', not the symbol
`read'.

-- 

Leonard Stiles <ljs@uk2.net>

Re: Like (display . all) but hide expirable messages

[Lars Magne Ingebrigtsen]

Leonard Stiles <ljs@uk2.net> writes:

> Ah, yes.  However:
>
> (Info-goto-node "(gnus)Group Parameters") =>
>
>     `posting-style'
>      [...]
>
>      For instance, if you want a funky name and signature in this
>      group only, instead of hacking `gnus-posting-styles', you could
>      put something like this in the group parameters:
>
>           (posting-style
>             (name "Funky Name")
>             (signature "Funky Signature"))
>
> Which does not set the variable `posting-style'.  (I suppose you could
> argue that this is because ((name "Funky Name") (signature "Funky
> Signature")) is not an eval-able form).

That's definitely a major bug

(posting-style (name "Funky name"))

will call the function `name'.  Which will lead to Emacs
beeping at you.  Or it would have if it weren't for the
`ignore-errors' that's wrapped around the code.

We need a new syntax for posting-style.

> Also, to be pedantic, in the above example the variable `display'
> would be set to the value of the variable `read', not the symbol
> `read'.

Exactly.  :-)

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Like (display . all) but hide expirable messages

[Wes Hardaker]

>>>>> On Thu, 09 Aug 2001 18:49:29 +0200, Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) said:

Kai> People who want to use Gnus like other mail readers want Gnus to
Kai> always show all mails, except the deleted ones.  Since expirable
Kai> messages are deleted, in a sense, it would be nice to provide a
Kai> (display . most) feature which does this.

Kai> Opinions?

I made a patch to do this years ago, but Lars rejected it.  I'd be
happy to resubmit it if the desire was re-found.  I've had it applied
to my CVS tree ever since i created it years ago.  The property value
I picked is "not-expired", but whatever...

What I wanted to do (that Lars said he would accept) would be to make
it a string or list such that you could do (display . "-E"), which
would functionally say display everything but expired articles.
Obviously the full range of marks would have to be supported.

-- 
"Ninjas aren't dangerous.  They're more afraid of you than you are of them."

Re: Like (display . all) but hide expirable messages

[Dan Christensen]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> People who want to use Gnus like other mail readers want Gnus to
> always show all mails, except the deleted ones.  Since expirable
> messages are deleted, in a sense, it would be nice to provide a
> (display . most) feature which does this.

Yes, I've mentioned several times how much I'd like this.  
I currently achieve this with the code below, but it causes
the summary to be generated twice, which is slow in large
groups.

Dan

-- 
Dan Christensen
jdc+news@uwo.ca

; This runs only when the summary is first prepared, and not after
; other limiting commands.  Thus `/ w' pop's this limit, and others,
; as one would like it to.
(add-hook 'gnus-summary-prepared-hook 
	  (lambda ()
	    (if (or (string-match "^nnfolder:" gnus-newsgroup-name)
		    (string-match "^nnml:" gnus-newsgroup-name))
		(jdc-gnus-summary-limit-exclude-unwanted))))

(defun jdc-gnus-summary-limit-exclude-unwanted ()
  (interactive)
  (gnus-summary-limit-to-marks "EG" t))

(add-hook 'gnus-summary-mode-hook 
'(lambda nil
   (define-key gnus-summary-mode-map "/e" 'jdc-gnus-summary-limit-exclude-unwanted)))