bug#5299: custom-reevaluate-setting / custom-initialize-delay

bug#5299: custom-reevaluate-setting / custom-initialize-delay

[Stefan Monnier]

[ Bringing this back to #5299 where it belongs ]

>> > Drew, does the problem in #5299 go away for you if you set
>> > send-mail-function to mailclient-send-it?
>> 
>> It's message-send-mail-function (the variable) that one needs to set
>> nowadays.
>> 
>> The function message-send-mail-function may also need some adjustment
>> for Windows and Macs to prefer mailclient more strongly than it
>> currently does.

> OK, I tried that too:
> (setq message-send-mail-function 'mailclient-send-it)

> Same thing:

> shell-command-to-string: Searching for program: no such file or directory,
> /bin/bash

Can you get a backtrace of that? [with (setq debug-on-error t) and if
that doesn't work with additionally (setq debug-ignored-errors nil) and/or
maybe even (setq debug-on-signal t)]

Also, can you explain why "bash" shows up here?  Do you have the envvar
SHELL set to "/bin/bash" somehow?


        Stefan

bug#5299: custom-reevaluate-setting / custom-initialize-delay

[Drew Adams]

> [ Bringing this back to #5299 where it belongs ]
> 
> > (setq message-send-mail-function 'mailclient-send-it)
> >
> > shell-command-to-string: Searching for program: no such 
> > file or directory, /bin/bash
> 
> Can you get a backtrace of that?
> [with (setq debug-on-error t) and if
> that doesn't work with additionally (setq 
> debug-ignored-errors nil) and/or
> maybe even (setq debug-on-signal t)]
> 
> Also, can you explain why "bash" shows up here?  Do you have 
> the envvar SHELL set to "/bin/bash" somehow?

Yes, in my Windows settings (since I use Cygwin).

With those 3 variables set, emacs -Q:

Help > Send Bug Report

Type "ffffffffff" as the subject.

Get this backtrace immediately:

Debugger entered--Lisp error: (file-error "Cannot open load file" "idna")
  require(idna)
  (condition-case nil (require (quote idna)) (file-error))
  (and (condition-case nil (require ...) (file-error)) (mm-coding-system-p
(quote utf-8)) (executable-find idna-program) (string= (idna-to-ascii
"räksmörgås") "xn--rksmrgs-5wao1o") t)
  eval((and (condition-case nil (require ...) (file-error)) (mm-coding-system-p
(quote utf-8)) (executable-find idna-program) (string= (idna-to-ascii
"räksmörgås") "xn--rksmrgs-5wao1o") t))
  custom-initialize-reset(message-use-idna (and (condition-case nil (require
...) (file-error)) (mm-coding-system-p (quote utf-8)) (executable-find
idna-program) (string= (idna-to-ascii "räksmörgås") "xn--rksmrgs-5wao1o") t))
  custom-declare-variable(message-use-idna (and (condition-case nil (require
...) (file-error)) (mm-coding-system-p (quote utf-8)) (executable-find
idna-program) (string= (idna-to-ascii "räksmörgås") "xn--rksmrgs-5wao1o") t)
("c:/Emacs-23-1-92/lisp/gnus/message.elc" . 59102) :version "22.1" :group
message-headers :link (custom-manual "(message)IDNA") :type (choice (const :tag
"Ask" ask) (const :tag "Never" nil) (const :tag "Always" t)))
  message-mail("bug-gnu-emacs@gnu.org" "23.1.92; fffffffff" nil nil nil nil nil)
  compose-mail("bug-gnu-emacs@gnu.org" "23.1.92; fffffffff")
  report-emacs-bug("fffffffff" [134217786 40 103 101 116 101 110 118 32 34 83 72
69 76 76 34 41 return (help-echo #<frame emacs@DRADAMS-LAP1 0x2f37800> "Search
forward for a string" nil nil -1) (help-echo #<frame emacs@DRADAMS-LAP1
0x2f37800> "mouse-2: visit this file in other window" #<window 3 on
drews-lisp-20> #<buffer drews-lisp-20> 523) 134217848 115 101 116 45 118 97 114
return 100 101 98 117 103 45 111 110 45 101 114 114 111 114 return 116 return
134217848 115 101 116 ...])
  call-interactively(report-emacs-bug nil nil)

Starting over, emacs -Q, with 2/3 vars set, not `debug-ignored-errors': same
thing.

Starting over, emacs -Q, with only `debug-on-error' set:

I get to buffer *mail to bug-gnu-emacs@gnu.org*. I type some text "xxxxxxxxxxx",
then `C-c C-c', reply "yes". I get this backtrace:

Debugger entered--Lisp error: (file-error "Searching for program" "no such file
or directory" "/bin/bash")
  call-process("/bin/bash" nil t nil "-c" "uncompface")
  shell-command-to-string("uncompface")
  (string-match "^0x" (shell-command-to-string "uncompface"))
  (and (string-match "^0x" (shell-command-to-string "uncompface"))
(executable-find "icontopbm"))
  (if (featurep (quote xemacs)) (featurep (quote xface)) (and (string-match
"^0x" ...) (executable-find "icontopbm")))
  (and (not noninteractive) (gnus-image-type-available-p (quote xbm)) (if
(featurep ...) (featurep ...) (and ... ...)) (quote head))
  eval((and (not noninteractive) (gnus-image-type-available-p (quote xbm)) (if
(featurep ...) (featurep ...) (and ... ...)) (quote head)))
  custom-initialize-reset(gnus-treat-display-x-face (and (not noninteractive)
(gnus-image-type-available-p (quote xbm)) (if (featurep ...) (featurep ...) (and
... ...)) (quote head)))
  custom-declare-variable(gnus-treat-display-x-face (and (not noninteractive)
(gnus-image-type-available-p (quote xbm)) (if (featurep ...) (featurep ...) (and
... ...)) (quote head)) ("c:/Emacs-23-1-92/lisp/gnus/gnus-art.elc" . 52832)
:group gnus-article-treat :version "21.1" :link (custom-manual
"(gnus)Customizing Articles") :link (custom-manual "(gnus)X-Face") :type (choice
(const :tag "Off" nil) (const :tag "Header" head)) :set #[(symbol value)
"\302\b\303\b!\204\x0e


HTH.

bug#5299: custom-reevaluate-setting / custom-initialize-delay

[Stefan Monnier]

>>>>> "Drew" == Drew Adams <drew.adams@oracle.com> writes:

>> [ Bringing this back to #5299 where it belongs ]
>> 
>> > (setq message-send-mail-function 'mailclient-send-it)
>> >
>> > shell-command-to-string: Searching for program: no such 
>> > file or directory, /bin/bash
>> 
>> Can you get a backtrace of that?
>> [with (setq debug-on-error t) and if
>> that doesn't work with additionally (setq 
>> debug-ignored-errors nil) and/or
>> maybe even (setq debug-on-signal t)]
>> 
>> Also, can you explain why "bash" shows up here?  Do you have 
>> the envvar SHELL set to "/bin/bash" somehow?

> Yes, in my Windows settings (since I use Cygwin).

> With those 3 variables set, emacs -Q:

> Help > Send Bug Report

> Type "ffffffffff" as the subject.
[...]
> Starting over, emacs -Q, with only `debug-on-error' set:

> I get to buffer *mail to bug-gnu-emacs@gnu.org*. I type some text "xxxxxxxxxxx",
> then `C-c C-c', reply "yes". I get this backtrace:

> Debugger entered--Lisp error: (file-error "Searching for program" "no such file
> or directory" "/bin/bash")
>   call-process("/bin/bash" nil t nil "-c" "uncompface")
>   shell-command-to-string("uncompface")
>   (string-match "^0x" (shell-command-to-string "uncompface"))

Thanks, so I think the patch below is needed, can you try it to see if
it helps?

There's still another bug in sendmail.el where we use
custom-initialize-delay on a defcustom that's autoloaded (the problem
there being that autoloaded defcustoms aren't copied verbatim, so only
some of the defcustom features are supported for autoloaded defcustoms,
and currently custom-initialize-delay is not one of them), but this
patch may be sufficient for you and seems useful in anny case.


        Stefan


=== modified file 'lisp/gnus/gnus-art.el'
--- lisp/gnus/gnus-art.el	2010-01-13 08:35:10 +0000
+++ lisp/gnus/gnus-art.el	2010-02-24 16:25:07 +0000
@@ -1422,8 +1422,12 @@
        (gnus-image-type-available-p 'xbm)
        (if (featurep 'xemacs)
 	   (featurep 'xface)
+	 (condition-case nil
 	 (and (string-match "^0x" (shell-command-to-string "uncompface"))
-	      (executable-find "icontopbm")))
+                  (executable-find "icontopbm"))
+           ;; shell-command-to-string may signal an error, e.g. if
+           ;; shell-file-name is not found.
+           (error nil)))
        'head)
   "Display X-Face headers.
 Valid values are nil and `head'.

bug#5299: custom-reevaluate-setting / custom-initialize-delay

[Drew Adams]

> Thanks, so I think the patch below is needed, can you try it to see if
> it helps?

Nope.

I loaded a patched version of gnus-art.el, then tried to send a bug report.

I got only this (in *Messages*):

Sending...
Sending via mail...
message-send-mail-with-sendmail: Sending...failed with exit value 1

bug#5299: Default mailer (was: custom-reevaluate-setting / custom-initialize-delay)

[Stefan Monnier]

>> Thanks, so I think the patch below is needed, can you try it to see if
>> it helps?
> Nope.
> I loaded a patched version of gnus-art.el, then tried to send a bug report.
> I got only this (in *Messages*):

> Sending...
> Sending via mail...
> message-send-mail-with-sendmail: Sending...failed with exit value 1

I guess that good: it means that my patch worked and then you bumped
into the next problem which is that message-send-mail-function is not
set to use mailclient.

I think we need to revisit the setting for the default mailer: I think
neither the mailclient-send-it nor any other setting is a good default.
I think we should instead use as default a special function that prompts
the user whether he wants to use his external MUA to send email
(i.e. use mailclient-send-it), or to use /usr/sbin/sendmail, or to use
smtpmail in which case it'll ask the user to configure his SMTP server.

And this should be used on *all* systems rather than only under w32
and/or macosx, since sadly /usr/sbin/sendmail is too often unusable even
under GNU/Linux (my /usr/sbin/sendmail is usable, but I must admit that
it took me a significant amount of work on my laptop, so I can't expect
most people to go through that trouble).


        Stefan

bug#5299: Default mailer (was: custom-reevaluate-setting / custom-initialize-delay)

[Lennart Borgman]

On Wed, Feb 24, 2010 at 7:51 PM, Stefan Monnier
<monnier@iro.umontreal.ca> wrote:
>
> I think we need to revisit the setting for the default mailer: I think
> neither the mailclient-send-it nor any other setting is a good default.
> I think we should instead use as default a special function that prompts
> the user whether he wants to use his external MUA to send email
> (i.e. use mailclient-send-it), or to use /usr/sbin/sendmail, or to use
> smtpmail in which case it'll ask the user to configure his SMTP server.
>
> And this should be used on *all* systems rather than only under w32
> and/or macosx, since sadly /usr/sbin/sendmail is too often unusable even
> under GNU/Linux (my /usr/sbin/sendmail is usable, but I must admit that
> it took me a significant amount of work on my laptop, so I can't expect
> most people to go through that trouble).


That sounds good and in line with what we previously did.

bug#5299: Default mailer (was: custom-reevaluate-setting / custom-initialize-delay)

[Drew Adams]

> I think we need to revisit the setting for the default mailer: I think
> neither the mailclient-send-it nor any other setting is a 
> good default.
> I think we should instead use as default a special function 
> that prompts
> the user whether he wants to use his external MUA to send email
> (i.e. use mailclient-send-it), or to use /usr/sbin/sendmail, or to use
> smtpmail in which case it'll ask the user to configure his 
> SMTP server.
> 
> And this should be used on *all* systems rather than only under w32
> and/or macosx, since sadly /usr/sbin/sendmail is too often 
> unusable even
> under GNU/Linux (my /usr/sbin/sendmail is usable, but I must 
> admit that
> it took me a significant amount of work on my laptop, so I 
> can't expect
> most people to go through that trouble).

I would hope that if a user is interrupted by such questions this happens only
the first time, not each time s?he sends a mail.

IOW, questions are OK as a means to help a user configure things, but once
defaults are configured according to the user's preference, then there shouldn't
be any more prompting.

If there is a need to let users override their default settings, then that could
perhap be done either via a different command or using C-u.

bug#5299: Default mailer

[Stefan Monnier]

> I would hope that if a user is interrupted by such questions this
> happens only the first time, not each time s?he sends a mail.

Goes without saying,


        Stefan

bug#5299:

[Glenn Morris]

[ See discussion at
  http://lists.gnu.org/archive/html/emacs-devel/2010-02/msg00550.html ]

Drew Adams wrote (on Fri, 26 Feb 2010 at 09:54 -0800):

> Loading a patched gnus-art.el plus setting message-send-mail-function to
> message-send-mail-with-mailclient works.

OK, then with the changes from this report for gnus-art and from #5643
for the default message-send-mail-function, this should be working
again in the latest 23.1.93 pretest.