bug#4706: 23.1.50; RMAIL mode fails to read mail first time

bug#4706: 23.1.50; RMAIL mode fails to read mail first time

[jpff]

This started happening a few days ago.  When I invoke rmail command it
loads by RMAIL file and then asserts that there it cannot get the po:
password, and there is no new mail.  I then run (rmail-get-new-mail)
explicitely and it works.  I have 
(setq rmail-primary-inbox-list 
      '("po:jpff" "/var/spool/mail/jpff"))
(setenv "MAILHOST" "air.cs.bath.ac.uk")
and the password is set 
          (custom-set-variables
           '(rmail-pop-password-required t)
           '(rmail-pop-password "xxxxxxxx"))
This counts more of an irritation than a bug, but it used to work.  On
my other network which does not use pop there is no problem.


In GNU Emacs 23.1.50.7 (x86_64-unknown-linux-gnu, X toolkit, Xaw3d scroll bars)
 of 2009-10-12 on xenakis
Windowing system distributor `The X.Org Foundation', version 11.0.10502000
Important settings:
  value of $LC_ALL: nil
  value of $LC_COLLATE: nil
  value of $LC_CTYPE: nil
  value of $LC_MESSAGES: nil
  value of $LC_MONETARY: nil
  value of $LC_NUMERIC: nil
  value of $LC_TIME: nil
  value of $LANG: en_GB.UTF-8
  value of $XMODIFIERS: @im=local
  locale-coding-system: utf-8-unix
  default enable-multibyte-characters: t

Major mode: RMAIL

Minor modes in effect:
  auto-image-file-mode: t
  show-paren-mode: t
  display-time-mode: t
  tooltip-mode: t
  mouse-wheel-mode: t
  menu-bar-mode: t
  file-name-shadow-mode: t
  global-font-lock-mode: t
  font-lock-mode: t
  global-auto-composition-mode: t
  auto-composition-mode: t
  auto-encryption-mode: t
  auto-compression-mode: t
  line-number-mode: t
  transient-mark-mode: t

Recent input:
M-r g M-x r e p o <tab> r <tab> <return>

Recent messages:
0 new messages read
Loading mail-extr...done
Getting mail from the remote server ...
Counting new messages...
Counting messages...20
Counting new messages...done (30)
Saving file /bigdisk/jpff/RMAIL...
Wrote /bigdisk/jpff/RMAIL
30 new messages read
Making completion list...

Load-path shadows:
/bigdisk/jpff/GNU_21/local-lisp/t-mouse hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/t-mouse
/bigdisk/jpff/GNU_21/local-lisp/ses hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/ses
/bigdisk/jpff/GNU_21/local-lisp/scroll-lock hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/scroll-lock
/bigdisk/jpff/GNU_21/local-lisp/ruler-mode hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/ruler-mode
/bigdisk/jpff/GNU_21/local-lisp/emerge hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/emerge
/bigdisk/jpff/GNU_21/local-lisp/url-http hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-http
/bigdisk/jpff/GNU_21/local-lisp/url-vars hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-vars
/bigdisk/jpff/GNU_21/local-lisp/url-parse hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-parse
/bigdisk/jpff/GNU_21/local-lisp/url-ns hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-ns
/bigdisk/jpff/GNU_21/local-lisp/url-nfs hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-nfs
/bigdisk/jpff/GNU_21/local-lisp/url-news hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-news
/bigdisk/jpff/GNU_21/local-lisp/url-misc hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-misc
/bigdisk/jpff/GNU_21/local-lisp/url-ldap hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-ldap
/bigdisk/jpff/GNU_21/local-lisp/url-irc hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-irc
/bigdisk/jpff/GNU_21/local-lisp/url-gw hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-gw
/bigdisk/jpff/GNU_21/local-lisp/url-file hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-file
/bigdisk/jpff/GNU_21/local-lisp/url hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url
/bigdisk/jpff/GNU_21/local-lisp/url-cid hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-cid
/bigdisk/jpff/GNU_21/local-lisp/url-cookie hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-cookie
/bigdisk/jpff/GNU_21/local-lisp/url-cache hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-cache
/bigdisk/jpff/GNU_21/local-lisp/url-auth hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/url/url-auth
/bigdisk/jpff/GNU_21/local-lisp/socks hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/net/socks
/bigdisk/jpff/GNU_21/local-lisp/feedmail hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/mail/feedmail
/bigdisk/jpff/GNU_21/local-lisp/spam-stat hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/gnus/spam-stat
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/textmodes/table hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/table
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/semantic
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/progmodes/compile hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/compile
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/progmodes/cpp hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/cpp
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/loaddefs hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/loaddefs
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/texi hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/texi
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/sort hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/sort
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/java hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/java
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/emacs-lisp/chart hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/chart
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/find hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/find
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/format hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/format
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/complete hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/complete
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/emacs-lisp/debug hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/debug
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/ctxt hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/ctxt
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/loaddefs hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/loaddefs
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/wisent hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/wisent/wisent
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/progmodes/grep hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/symref/grep
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/mode hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/decorate/mode
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/srecode/el hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/bovine/el
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/emacs-lisp/debug hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/bovine/debug
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/emacs-lisp/debug hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/analyze/debug
/bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/complete hides /bigdisk/jpff/GNU_21/emacs-22.0.50/lisp/cedet/semantic/analyze/complete

Features:
(shadow mailabbrev mailalias emacsbug sendmail mime-compose
mail-alias-menu mailcrypt mail-extr comint ring rmail mail-utils eldoc
delsel image-file cus-start cus-load logger crypt crypt++
crypt+pgp-pub paren uniquify advice help-fns help-mode view
advice-preload holidays hol-loaddefs regexp-opt cal-menu easymenu
calendar cal-loaddefs time tooltip ediff-hook vc-hooks lisp-float-type
mwheel x-win x-dnd tool-bar dnd fontset image fringe lisp-mode
register page menu-bar rfn-eshadow timer select scroll-bar mldrag
mouse jit-lock font-lock syntax facemenu font-core frame cham georgian
utf-8-lang misc-lang vietnamese tibetan thai tai-viet lao korean
japanese hebrew greek romanian slovak czech european ethiopic indian
cyrillic chinese case-table epa-hook jka-cmpr-hook help simple abbrev
loaddefs button minibuffer faces cus-face text-properties overlay md5
base64 format env code-pages mule custom widget
hashtable-print-readable backquote make-network-process x-toolkit x
multi-tty emacs)

==John ffitch

bug#4706: 23.1.50; RMAIL mode fails to read mail first time

[Glenn Morris]

jpff wrote:

> This started happening a few days ago.  When I invoke rmail command it
> loads by RMAIL file and then asserts that there it cannot get the po:
> password, and there is no new mail.
[...]
> and the password is set 
>           (custom-set-variables
>            '(rmail-pop-password-required t)
>            '(rmail-pop-password "xxxxxxxx"))

This is probably my mistake. Does it work with 's/pop/remote/' in
those variables names?

bug#4706: customize and aliaes [was Re: bug#4706: 23.1.50; RMAIL mode fails to read mail first time]

[Glenn Morris]

retitle 4706 customize and aliases
stop

I could have sworn this was supposed to work, eg:

http://lists.gnu.org/archive/html/emacs-devel/2005-02/msg00543.html
http://lists.gnu.org/archive/html/emacs-devel/2008-04/msg00716.html

Now I can't see that it ever has done. Consider:

foo.el:
(define-obsolete-variable-alias 'oldname 'newname "23.1")
(defcustom newname nil "" :type 'boolean :group 'customize)

~/.emacs:
(custom-set-variables
  '(oldname t))

(It works if you use setq rather than custom in your .emacs.)

Loading .emacs gives 'oldname a 'saved-value property. If foo.el has
not been loaded at this time, there is no way custom can know it
should really be applying this property to 'newname. Later on when
foo.el is loaded, custom-initialize-default checks if 'newname has a
'saved-value property, finds it does not, and so uses the default
value of nil rather than the desired value of t.

It will work if you load foo.el before .emacs, so one solution is to
autoload all defvaraliases that point to defcustoms. However, this is
not a solution for packages not distributed with emacs.

Another solution would be to make defvaralias copy any properties.
At present:

(put 'oldname 'foobar t)
(defvaralias 'oldname 'newname)
(get 'newname 'foobar)    ; -> nil

But changing this doesn't seem right in general, and might break some
existing behaviour.

If custom-initialize-default could somehow check if 'symbol has any
aliases, and if any of these have a 'saved-value property, that would
seem like the right thing to do. But I can't see how to do this. Is
there a "reverse indirect-variable" that lists all aliases of a given
variable?

rmail used to have a slightly complicated workaround for this. I
mistakenly thought it was not needed. Rather than put it back, I'd
prefer to just autoload the alias statements. There are other places
besides rmail that define aliases pointing to defcustoms, so some kind
of general solution is needed.


Any suggestions? 

Processed: customize and aliaes [was Re: bug#4706: 23.1.50; RMAIL mode fails to read mail first time]

[Emacs bug Tracking System]

Processing commands for control@emacsbugs.donarmstrong.com:

> retitle 4706 customize and aliases
Bug #4706 [emacs] 23.1.50; RMAIL mode fails to read mail first time
Changed bug title to 'customize and aliases' from '23.1.50; RMAIL mode fails to read mail first time'
> stop
Stopping processing here.

Please contact me if you need assistance.

Don Armstrong
(administrator, Emacs bugs database)

bug#4706: customize and aliaes

[Glenn Morris]

Glenn Morris wrote:

> Any suggestions? 

Perhaps if new_alias has a 'saved-value property, and base_variable
does not, Fdefvaralias could copy the property to base_variable?
(Additionally, perhaps only if base_variable is not bound?)

That seems like it might be a safe, general solution.

bug#4706: customize and aliaes

[Stefan Monnier]

>> Any suggestions? 
> Perhaps if new_alias has a 'saved-value property, and base_variable
> does not, Fdefvaralias could copy the property to base_variable?
> (Additionally, perhaps only if base_variable is not bound?)
> That seems like it might be a safe, general solution.

I think that extending defvaralias this way would be wrong, but I think it
would fit just fine in define-obsolete-variable-alias.


        Stefan

bug#4706: customize and aliaes

[Glenn Morris]

Stefan Monnier wrote:

>> Perhaps if new_alias has a 'saved-value property, and base_variable
>> does not, Fdefvaralias could copy the property to base_variable?
[...]
> I think that extending defvaralias this way would be wrong, but I think it
> would fit just fine in define-obsolete-variable-alias.

What then to do about non-obsolete aliases to defcustoms?

Such things are used eg in find-file, emacs-lisp/warnings.el, etc, and
probably outside of Emacs.

bug#4706: customize and aliaes

[Stefan Monnier]

>>> Perhaps if new_alias has a 'saved-value property, and base_variable
>>> does not, Fdefvaralias could copy the property to base_variable?
> [...]
>> I think that extending defvaralias this way would be wrong, but I think it
>> would fit just fine in define-obsolete-variable-alias.
> What then to do about non-obsolete aliases to defcustoms?

Depends on the intention between the alias, I guess.
I mean: why does that variable need more than one non-obsolete name?


        Stefan

bug#4706: customize and aliaes

[Glenn Morris]

Stefan Monnier wrote:

>> What then to do about non-obsolete aliases to defcustoms?
>
> Depends on the intention between the alias, I guess.
> I mean: why does that variable need more than one non-obsolete name?

I don't know.

I can create `define-custom-variable-alias' that DTRT, and replace
defvaralias with it where appropriate, if you like.

bug#4706: customize and aliaes

[Stefan Monnier]

>>> What then to do about non-obsolete aliases to defcustoms?
>> Depends on the intention between the alias, I guess.
>> I mean: why does that variable need more than one non-obsolete name?
> I don't know.
> I can create `define-custom-variable-alias' that DTRT, and replace
> defvaralias with it where appropriate, if you like.

I'd rather wait for bug-reports before doing such a thing.
Then we'll know better what needs to be done,


        Stefan

bug#4706: customize and aliaes

[Glenn Morris]

submitter 4706 rgm@gnu.org
severity 4706 minor
stop

Fixed for aliases defined via define-obsolete-variable-alias.

AFAICS, the only possible remaining issue is:

i) You have customized oldvar.

ii) (defcustom oldvar ...) is replaced by
(defvaralias 'oldvar 'newvar)
(defcustom newvar ...)

(in a non-dumped file)

where oldvar is either not obsolete, or is made obsolete "by hand"
rather than through define-obsolete-variable-alias.

Processed: Re: bug#4706: customize and aliaes

[Emacs bug Tracking System]

Processing commands for control@emacsbugs.donarmstrong.com:

> submitter 4706 rgm@gnu.org
Bug #4706 [emacs] customize and aliases
Changed bug submitter to 'rgm@gnu.org' from 'jpff <jpff@cs.bath.ac.uk>'
> severity 4706 minor
Bug #4706 [emacs] customize and aliases
Severity set to 'minor' from 'normal'

> stop
Stopping processing here.

Please contact me if you need assistance.

Don Armstrong
(administrator, Emacs bugs database)

bug#4706: 23.1.50; RMAIL mode fails to read mail first time

[jpff]

>>>>> "Glenn" == Glenn Morris <rgm@gnu.org> writes:

 Glenn> jpff wrote:
 >> This started happening a few days ago.  When I invoke rmail command it
 >> loads by RMAIL file and then asserts that there it cannot get the po:
 >> password, and there is no new mail.
 Glenn> [...]
 >> and the password is set 
 >> (custom-set-variables
 >> '(rmail-pop-password-required t)
 >> '(rmail-pop-password "xxxxxxxx"))

 Glenn> This is probably my mistake. Does it work with 's/pop/remote/' in
 Glenn> those variables names?

Not that I can determine.

I added

(custom-set-variables
  '(rmail-remote-password-required t)
  '(rmail-remote-password "xxxxxxxx"))

but it behaved just the same.  I am wondering if it could be to
exactly when the rmail-mode-hook is obeyed?
==John ffitch

bug#4706: 23.1.50; RMAIL mode fails to read mail first time

[Glenn Morris]

jpff wrote:

> I added
>
> (custom-set-variables
>   '(rmail-remote-password-required t)
>   '(rmail-remote-password "xxxxxxxx"))
>
> but it behaved just the same.  I am wondering if it could be to
> exactly when the rmail-mode-hook is obeyed?

Literally "added"? You may have only one custom-set-variables section.

I'm afraid it works for me with a .emacs consisting of just:

(setq rmail-primary-inbox-list '("po:USERNAME"))
(setenv "MAILHOST" "MAIL.SERVER")
(custom-set-variables
 '(rmail-remote-password-required t)
 '(rmail-remote-password "PASSWORD"))

BTW, this is all rather obsolete. The modern method would be:

(setq rmail-primary-inbox-list
   '("pop://USERNAME:PASSWORD@MAIL.SERVER"))

But either method works fine for me.


Why do you think rmail-mode-hook has anything to do with this?