Re: Envelope sender lost from sending after gnus-draft-edit-message

Re: Envelope sender lost from sending after gnus-draft-edit-message

[Jesper Harder]

"Andrew A. Raines" <drew@poured.net> writes:

> I notice, Jesper, that you added a variable and function,
> message-sendmail-envelope-from, which, when set to 'header, will
> create the envelope using the From: address.  For some reason,
> though, it doesn't work me:
>
> ,----[ C-h v message-sendmail-envelope-from RET ]
>
> Are you still working on this implementation?

It's supposed to be working :-) What do you get when you eval:

     (let (message-sendmail-envelope-from)
       (message-sendmail-envelope-from))

and

     (let ((message-sendmail-envelope-from 'header))
       (with-temp-buffer
         (insert "From: J Random Hacker <jrh@example.com>")
         (message-sendmail-envelope-from)))
  

Re: Envelope sender lost from sending after gnus-draft-edit-message

[Andrew A. Raines]

Jesper Harder <harder@myrealbox.com> writes:

> "Andrew A. Raines" <drew@poured.net> writes:
>
>> I notice, Jesper, that you added a variable and function,
>> message-sendmail-envelope-from, which, when set to 'header, will
>> create the envelope using the From: address.  For some reason,
>> though, it doesn't work me:

[...]

> It's supposed to be working :-) What do you get when you eval:
>
>      (let (message-sendmail-envelope-from)
>        (message-sendmail-envelope-from))

This returns what I have set for user-mail-address in .gnus.el.

>      (let ((message-sendmail-envelope-from 'header))
>        (with-temp-buffer
>          (insert "From: J Random Hacker <jrh@example.com>")
>          (message-sendmail-envelope-from)))

This returns `jrh@example.com'.

I just threw those into *scratch* and C-x C-e'd each of them.  Is
that what you meant?  I'm not sure if I was supposed to see more
than the one-liners on the message bar.

Re: Envelope sender lost from sending after gnus-draft-edit-message

[Jesper Harder]

"Andrew A. Raines" <drew@poured.net> writes:

> Jesper Harder <harder@myrealbox.com> writes:
>
>> It's supposed to be working :-) What do you get when you eval:
>>
>>      (let (message-sendmail-envelope-from)
>>        (message-sendmail-envelope-from))
>
> This returns what I have set for user-mail-address in .gnus.el.
>
>>      (let ((message-sendmail-envelope-from 'header))
>>        (with-temp-buffer
>>          (insert "From: J Random Hacker <jrh@example.com>")
>>          (message-sendmail-envelope-from)))
>
> This returns `jrh@example.com'.

OK, so this suggests that `message-sendmail-envelope-from' is working
correctly.

I wonder what is wrong in your case.  What's your value of
`message-send-mail-function'?  The option only applies if it is
`message-send-mail-with-sendmail'.

> I just threw those into *scratch* and C-x C-e'd each of them.  Is
> that what you meant?

Yes.

Re: Envelope sender lost from sending after gnus-draft-edit-message

[Andrew A. Raines]

Jesper Harder <harder@myrealbox.com> writes:

> OK, so this suggests that `message-sendmail-envelope-from' is
> working correctly.
>
> I wonder what is wrong in your case.  What's your value of
> `message-send-mail-function'?  The option only applies if it is
> `message-send-mail-with-sendmail'.

Darn.  I wish I had it set to something else, but I don't.

,----[ C-h v message-send-mail-function RET ]
| `message-send-mail-function' is a variable declared in Lisp.
|   -- loaded from "message"
| 
| Value: message-send-mail-with-sendmail
[...]
`----

Bummer, I don't know enough elisp to diagnose what's wrong.

How did earlier Gnuses handle this issue?  It seems like this only
started recently.  Surely I'm not the first person expecting the
envelope to match the header contents.

-- 
http://poured.net/drew/dot-gnus.txt

Re: Envelope sender lost from sending after gnus-draft-edit-message

[Jesper Harder]

[-- Attachment #1: Type: text/plain, Size: 553 bytes --]

"Andrew A. Raines" <drew@poured.net> writes:

> How did earlier Gnuses handle this issue?  

Basically the same way as now: It passes the envelope sender to sendmail
with '-f' on the command line.

If `message-sendmail-envelope-from' is nil (the default), nothing should
have changed -- it'll use `message-make-address' in the envelope.

> It seems like this only started recently.  

I don't see how `message-sendmail-envelope-from' could make anything
different with the default setting.  But you could try this patch and
see if it changes anything:


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: message.el.diff --]
[-- Type: text/x-patch, Size: 473 bytes --]

--- gnus/lisp/message.el	Sat Mar  1 15:28:48 2003
+++ message.el	Sun Mar  2 17:28:12 2003
@@ -3559,7 +3559,7 @@
 			;; But some systems are more broken with -f, so
 			;; we'll let users override this.
 			(if (null message-sendmail-f-is-evil)
-			    (list "-f" (message-sendmail-envelope-from)))
+			    (list "-f" (message-make-address)))
 			;; These mean "report errors by mail"
 			;; and "deliver in background".
 			(if (null message-interactive) '("-oem" "-odb"))

[-- Attachment #3: Type: text/plain, Size: 346 bytes --]


Hmm, maybe your sendmail just doesn't like the '-f' option -- the
comment in message.el suggests that this is sometimes the case:

;; Always specify who from,
;; since some systems have broken sendmails.
;; But some systems are more broken with -f, so
;; we'll let users override this.

Does '-f' work if you use sendmail from the command line?

Proper Gnus rebuilding technique? (was: Envelope sender lost from sending after gnus-draft-edit-message)

[Andrew A. Raines]

Jesper Harder <harder@myrealbox.com> writes:

> I don't see how `message-sendmail-envelope-from' could make
> anything different with the default setting.  But you could try
> this patch and see if it changes anything:

I didn't need to try the patch.

Your assertion that it should work got me curious.  I deleted all
the *.elc files from gnus/lisp and rebuilt everything.  It seems to
be working now.  Even the draft/delayed envelope senders are
working properly (the original thread topic).  Cool!

I guess I should `make distclean' instead of just `make clean'ing
when rebuilding Gnus.  Also, do you guys restart Emacs when you
rebuild Gnus?

Thanks for your help, Kai and Jesper.  I'm probably not
knowledgeable enough about Emacs to run CVS Gnus, but I'm too
addicted to the features in Oort to run anything else!

-- 
(http://fyinic.com)

Re: Proper Gnus rebuilding technique?

[Jesper Harder]

"Andrew A. Raines" <drew@poured.net> writes:

> I guess I should `make distclean' instead of just `make clean'ing when
> rebuilding Gnus.

I don't do either.  I just do 'make' in the gnus directory.  (If you
know what you're doing, you can sometimes get away with 'make some').

> Also, do you guys restart Emacs when you rebuild Gnus?

Yes, that's necessary.

Re: Proper Gnus rebuilding technique?

[Frank Haun]

Jesper Harder <harder@myrealbox.com> writes:

> "Andrew A. Raines" <drew@poured.net> writes:
>
>> Also, do you guys restart Emacs when you rebuild Gnus?
>
> Yes, that's necessary.

What about `gnus-unload'?

I've made a test:

`M-x gnus-unload RET'

After typing `M-x gnus RET' I get some  errors:

gnus-shutdown: Symbol's function definition is void: gnus-topic-close
gnus-shutdown: Symbol's function definition is void: gnus-backlog-shutdown
gnus-shutdown: Symbol's function definition is void: gnus-async-close

OK, I load the library where the functions are defined:

`M-x load-library RET gnus-async RET'
`M-x load-library RET gnus-bcklg RET'
`M-x load-library RET gnus-topic RET'

Seems to work.

Frank

Re: Proper Gnus rebuilding technique?

[Jesper Harder]

Frank Haun <pille3003@fhaun.de> writes:

> Jesper Harder <harder@myrealbox.com> writes:
>
>> "Andrew A. Raines" <drew@poured.net> writes:
>>
>>> Also, do you guys restart Emacs when you rebuild Gnus?
>>
>> Yes, that's necessary.
>
> What about `gnus-unload'?

It doesn't work reliably.  

It only unloads stuff that begins with `gnus-', `nn', `mm-' or `rfc' --
there's a lot of other Gnus related stuff which doesn't have those
prefixes.

Re: Proper Gnus rebuilding technique?

[Michael Shields]

In article <m3healz5y6.fsf@defun.localdomain>,
Jesper Harder <harder@myrealbox.com> wrote:
> Frank Haun <pille3003@fhaun.de> writes:
[...]
>> What about `gnus-unload'?
>
> It doesn't work reliably.  

Is it useful, then?  Shouldn't it be either fixed or removed?
-- 
Shields.

Re: Proper Gnus rebuilding technique?

[Frank Haun]

Michael Shields <shields@msrl.com> writes:

> In article <m3healz5y6.fsf@defun.localdomain>,
> Jesper Harder <harder@myrealbox.com> wrote:
>> Frank Haun <pille3003@fhaun.de> writes:
> [...]
>>> What about `gnus-unload'?
>>
>> It doesn't work reliably.  
>
> Is it useful, then?  Shouldn't it be either fixed or removed?

,----[ C-h f gnus-unload RET ]
| [...]
| Currently, features whose names have prefixes `gnus-', `nn', `mm-' or
| `rfc' are unloaded.
| [...]
`----

I vote for remove or moving to contrib.

After having a look at
  http://anc.ed.ac.uk/~stephen/emacs/ell.html

these files are affected
  rfc.el rfc-util.el rfc-view.el rfc2104.el rfcr.el rfcview.el

For instance rfcview.el doesn't work after `gnus-unload'.

Yes, nnir.el and some more too, but it make no sense to use them after
`gnus-unload'. :-)

Frank

Re: Proper Gnus rebuilding technique?

[Lars Magne Ingebrigtsen]

Frank Haun <pille3003@fhaun.de> writes:

> ,----[ C-h f gnus-unload RET ]
> | [...]
> | Currently, features whose names have prefixes `gnus-', `nn', `mm-' or
> | `rfc' are unloaded.
> | [...]
> `----
>
> I vote for remove or moving to contrib.

I've now removed it.

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

Re: Proper Gnus rebuilding technique?

[Frank Haun]

Frank Haun <pille3003@fhaun.de> writes:

[gnus-unload]

> Seems to work.

Addition, I got:
  mailcap-mime-info: Symbol's function definition is void:
  mail-header-parse-content-type

so I have to load `mail-parse'.

,----[ C-h f gnus-unload RET ]
| [...]
| Use cautiously -- unloading may cause trouble.
`----

Yes.

Frank