Re: (possibly silly) problem sending drafts

[George McNinch]

Some of my earlier post was badly edited, sorry:

I wrote:

    George> If I compose a message and save (C-x C-s) before sending (to
    George> the drafts folder), and subsequently send the mail, it sends
    George> fine, but I get the following error 

Here is a hopefully better version of that sentence:

If I am composing a message, save it to the drafts folder (C-x C-s) and
then immediately send the mail (C-c C-c) I get the indicated error:

,----
| Sending...
| Mark set [2 times]
| Sending via mail...
| Sending email 
| Sending email done
| Sending...done
| Opening connection to personal via tls...
| open-network-stream: personal/993 Name or service not known
`----

I get the same error if I (say) delete the *unsent mail* buffer, go to
the Drafts group, and send the saved message from Drafts via (D s).

I'm a bit mystified why gnus is trying to establish a "connection
... via tls" at all. I get this behavior when I compose e.g. from the
*Groups* buffer (so there seems to me to be no reason why gnus is
connecting to anywhere after having sent the email...?).

For what it is worth, I have

   (setq message-send-mail-function 'smtpmail-send-it)

and 

   (setq gnus-message-archive-group nil)

in my config.   


As I mentioned in the earlier post, I have lines like


    (eval (setq smtpmail-smtp-user "gmcninch"))

and

    (eval (setq smtpmail-smtp-user "gmcninchprof"))

in gnus-posting-styles which control "who I'm posting as".

And mysteriously (to me at least) I get a different error when posting
"as gmcninchprof" then when posting "as gmcninch" -- namely the error

| Opening connection to personal via tls...
| open-network-stream: personal/993 Name or service not known

becomes

| Opening connection to gmcninch-tufts via tls...
| open-network-stream: gmcninch-tufts/993 Name or service not known

Indeed the condition testing in gnus-posting-styles is essentially
whether I'm posting from the "personal" group or the "gmcninch-tufts"
group, so there is some sense to this but I don't see the mechanism...



Well, hopefully my report is at least somewhat more coherent....

Thanks for any suggestions!

best,
gm



-- 
  ,--<>----
  | George McNinch <gmcninch@gmail.com>
  | http://gmcninch.math.tufts.edu
  `--<>----