Problem with 0.53/0.54

Problem with 0.53/0.54

[Lars Balker Rasmussen]

In message.el there is a variable message-generate-new-buffers which has
the default value 'unique.

In message-buffer-name the following code is causing me grief:

  (cond
   ;; Check whether `message-generate-new-buffers' is a function,
   ;; and if so, call it.
   ((message-functionp message-generate-new-buffers)
    (funcall message-generate-new-buffers type to group))

Since I have a function unique defined in my running Emacs:

	unique is a compiled Lisp function in `font'.
	(unique LIST)
	not documented

I get:

	Signaling: (wrong-number-of-arguments #[(list) "[...]" [nil cur retval list] 2] 3)
	  unique("followup" "Troels Arvin <tarvin@mdb.ku.dk>" "dk.edb.unix")
	  message-buffer-name("followup" "Troels Arvin <tarvin@mdb.ku.dk>" "dk.edb.unix")
	  message-followup(nil)
	  gnus-post-news(nil "dk.edb.unix" [10923 "Redirections" "Troels Arvin <tarvin@mdb.ku.dk>" "Wed, 25 Nov 1998 17:06:44 +0100" "<365C2B14.2F538723@mdb.ku.dk>" "<365ae3ed.13037251@news.get2net.dk> <yd8u2zpuhyb.fsf@zastava.kom.auc.dk> <365b225e.8054027@news.get2net.dk> <x2lnl037aq.fsf@blueberry.kubism.ku.dk>" 990 12 "Xref: news.daimi.au.dk dk.edb.unix:10923" nil] "*Article*" (10923) nil nil)
	  gnus-summary-followup((10923) nil)
	  gnus-summary-followup-with-original(nil)
	* call-interactively(gnus-summary-followup-with-original)

when trying to do an 'F'.

I suppose this is as much a problem with font as it is with pgnus, but
isn't there some way around this problem?

BTW, This problem only appears after our local broken w3-installation
attempts and fails to parse html-articles.  Is there some way to tell
gnus never to use w3, even if it is available?
-- 
Lars Balker Rasmussen, Software Engineer, Mjolner Informatics ApS
lbr@mjolner.dk

Re: Problem with 0.53/0.54

[Lars Magne Ingebrigtsen]

Lars Balker Rasmussen <lbr@mjolner.dk> writes:

> Since I have a function unique defined in my running Emacs:
> 
> 	unique is a compiled Lisp function in `font'.
> 	(unique LIST)
> 	not documented

This is pure evil.  Get thee behind me!  Etc!

Anyway, it does bring up a general question -- many of the Gnus
variables can be either nil, to some symbols or a function.  The
symbols have been cleverly chosen not to map to any Emacs functions
(`some', `unique', `more'), but those things change over time, so
should I stop using symbols in these types of situations?

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

Re: Problem with 0.53/0.54

[Colin Rafferty]

Lars Magne Ingebrigtsen writes:

> Anyway, it does bring up a general question -- many of the Gnus
> variables can be either nil, to some symbols or a function.  The
> symbols have been cleverly chosen not to map to any Emacs functions
> (`some', `unique', `more'), but those things change over time, so
> should I stop using symbols in these types of situations?

Symbols are not a problem if you test for them before testing
functionp.  I don't see why this isn't the simplest solution.

Alternatively, you could implement the `unique' capability as a
function, and have the users choose that instead of the symbol
`unique'.

For example:

(defcustom message-generate-new-buffers 'message-generate-new-buffers-unique
  "*Non-nil ...."
  :group 'message-buffers
  :type '(choice (const :tag "off" nil)
		 (const :tag "unique" message-generate-new-buffers-unique)
		 (const :tag "unsuniqueent" message-generate-new-buffers-unsent)
		 (function fun)))

;; Colin

Re: Problem with 0.53/0.54

[Lars Magne Ingebrigtsen]

Colin Rafferty <craffert@ms.com> writes:

> Symbols are not a problem if you test for them before testing
> functionp.  I don't see why this isn't the simplest solution.

Er, yes.  :-)  Fix in Pterodactyl Gnus v0.55.

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

Re: Problem with 0.53/0.54

[Felix Lee]

Lars Magne Ingebrigtsen <larsi@gnus.org>:
> Anyway, it does bring up a general question -- many of the Gnus
> variables can be either nil, to some symbols or a function.  The
> symbols have been cleverly chosen not to map to any Emacs functions
> (`some', `unique', `more'), but those things change over time, so
> should I stop using symbols in these types of situations?

a. use funky symbols like ":unique" that are always never
   almost usually not going to be function names.

b. use strings for special values.

c. always check if it's a special symbol before checking if
   it's a function.  if someone really wants the setting to
   be the function 'unique', you can
     (fset 'my-unique 'unique)

d. have function values be: '(function unique)

e. use vectors, just because.

f. have non-function values be: '(unique)

g. use text properties to mark function values in red.
--

Re: Problem with 0.53/0.54

[Lars Balker Rasmussen]

Lars Balker Rasmussen <lbr@mjolner.dk> writes:
> BTW, This problem only appears after our local broken w3-installation
> attempts and fails to parse html-articles.  Is there some way to tell
> gnus never to use w3, even if it is available?

I'd still like to know this :-)

(HTML-articles are spawns of the devil, and only deserve to be ignored.)
-- 
Lars Balker Rasmussen, Software Engineer, Mjolner Informatics ApS
lbr@mjolner.dk

Re: Problem with 0.53/0.54

[Robert Bihlmeyer]

Hi,

>>>>> On 26 Nov 1998 09:56:23 +0100
>>>>> Lars Balker Rasmussen <lbr@mjolner.dk> said:

 Lars> Is there some way to tell gnus never to use w3, even if it is
 Lars> available?

 Lars> I'd still like to know this :-)

 Lars> (HTML-articles are spawns of the devil, and only deserve to be
 Lars> ignored.)

Hmm, does putting

	text/html; nuke %{Sender-IP}

in your .mailcap not work?

Oh, gnus does not provide Sender-IP...

	Robbe

-- 
Robert Bihlmeyer	reads: Deutsch, English, MIME, Latin-1, NO SPAM!
<robbe@orcus.priv.at>	<http://stud2.tuwien.ac.at/~e9426626/sig.html>