Define groups of recipients

Define groups of recipients

[Michael Heerdegen]

Hi,

say I often want to send a message to a constant larger group of
recipients, what is the simplest way to do this with Emacs and Gnus?

I use BBDB and vaguely remember that it could be used to define mail
aliases for groups of mail addresses but it seems BBDB now comes without
documentation.


TIA,

Michael.

Re: Define groups of recipients

[Enrico Schumann]

>>>>> "MH" == Michael Heerdegen <michael_heerdegen@web.de> writes:

    MH> Hi,
    MH> say I often want to send a message to a constant larger group of
    MH> recipients, what is the simplest way to do this with Emacs and Gnus?

    MH> I use BBDB and vaguely remember that it could be used to define mail
    MH> aliases for groups of mail addresses but it seems BBDB now comes without
    MH> documentation.


    MH> TIA,

    MH> Michael.

With BBDB 3.2, I have

  'mail-alias: <grp>'

fields in the contacts that belong to <grp> in my .bbdb.
In my .gnus.el, I have

  (add-hook 'message-setup-hook 'bbdb-mail-aliases)

When composing a message, the mail-alias in "To:", say,
expands (with SPC) into the list of emails.


kind regards
     Enrico
     
-- 
Enrico Schumann
Lucerne, Switzerland
http://enricoschumann.net

Re: Define groups of recipients

[Michael Heerdegen]

Enrico Schumann <es@enricoschumann.net> writes:

> With BBDB 3.2, I have
>
>   'mail-alias: <grp>'
>
> fields in the contacts that belong to <grp> in my .bbdb.
> In my .gnus.el, I have
>
>   (add-hook 'message-setup-hook 'bbdb-mail-aliases)
>
> When composing a message, the mail-alias in "To:", say,
> expands (with SPC) into the list of emails.

Ok, thanks.  Do I get it right that this approach involves to add an
individual contact for any person that should be part of the group?  Is
there any way to avoid this?

Michael.

Re: Define groups of recipients

[Eric Abrahamsen]

Michael Heerdegen <michael_heerdegen@web.de> writes:

> Enrico Schumann <es@enricoschumann.net> writes:
>
>> With BBDB 3.2, I have
>>
>>   'mail-alias: <grp>'
>>
>> fields in the contacts that belong to <grp> in my .bbdb.
>> In my .gnus.el, I have
>>
>>   (add-hook 'message-setup-hook 'bbdb-mail-aliases)
>>
>> When composing a message, the mail-alias in "To:", say,
>> expands (with SPC) into the list of emails.
>
> Ok, thanks.  Do I get it right that this approach involves to add an
> individual contact for any person that should be part of the group?  Is
> there any way to avoid this?

How would you rather have it work? I think you can also do it manually
by adding lines to a ~/.mailrc file, see the "Mail Aliases" section of
the Message manual. It looks like it's meant for expanding a single
contact address, but I'll bet you can put whatever string you like in
there.

Eric

Re: Define groups of recipients

[Michael Heerdegen]

Eric Abrahamsen <eric@ericabrahamsen.net> writes:

> How would you rather have it work? I think you can also do it manually
> by adding lines to a ~/.mailrc file, see the "Mail Aliases" section of
> the Message manual. It looks like it's meant for expanding a single
> contact address, but I'll bet you can put whatever string you like in
> there.

Ok, that would be the last resort.  I had hoped for a nice solution in
Emacs.

BTW, anyone knowing why BBDB comes with a manual that looks like

1 First Chapter
***************

This is the first chapter.

   Here is a numbered list.

  1. This is the first item.

  2. This is the second item.

?  I can't find any other documentation installed.


Thanks,

Michael.

Re: Define groups of recipients

[Eric Abrahamsen]

Michael Heerdegen <michael_heerdegen@web.de> writes:

> Eric Abrahamsen <eric@ericabrahamsen.net> writes:
>
>> How would you rather have it work? I think you can also do it manually
>> by adding lines to a ~/.mailrc file, see the "Mail Aliases" section of
>> the Message manual. It looks like it's meant for expanding a single
>> contact address, but I'll bet you can put whatever string you like in
>> there.
>
> Ok, that would be the last resort.  I had hoped for a nice solution in
> Emacs.

But what would a "nice solution" look like? I suppose you could always
write your own completion-at-point function.

> BTW, anyone knowing why BBDB comes with a manual that looks like

BBDB's lack of manual is half of why I ended up writing EBDB (only
partly joking).

Re: Define groups of recipients

[Michael Heerdegen]

Eric Abrahamsen <eric@ericabrahamsen.net> writes:

> > Ok, that would be the last resort.  I had hoped for a nice solution in
> > Emacs.
>
> But what would a "nice solution" look like? I suppose you could always
> write your own completion-at-point function.

That I could somewhere create a record for a group consisting of a name,
a mail alias, and a number of Email addresses.  Writing my own
completion-at-point function doesn't count as "nice".

> > BTW, anyone knowing why BBDB comes with a manual that looks like
>
> BBDB's lack of manual is half of why I ended up writing EBDB (only
> partly joking).

Ok, I may be switching because of the manual (also only partly joking).
No, I wanted to have a look at it for a longer time anyway ;-)


Thanks,

Michael.

Re: Define groups of recipients

[Emanuel Berg]

Michael Heerdegen wrote:

> say I often want to send a message to
> a constant larger group of recipients, what
> is the simplest way to do this with Emacs
> and Gnus?

Easy! With abbrevs, use the ~/.mailrc file:

    # Google Translate for Emacs
    alias trans-1 "Oleksandr Manzyuk <not-the-mail@gmail.com>"
    alias trans-2 "Andrey Tykhonov <not-the-mail@gmail.com>"
    alias trans trans-1 trans-2

now just type "trans" and it gets expanded to
both guys :)

I hope Gaspadins Tykhonov and Manzyuk don't
mind me using them as examples :)

-- 
underground experts united
http://user.it.uu.se/~embe8573

Re: Define groups of recipients

[Eric Abrahamsen]

Michael Heerdegen <michael_heerdegen@web.de> writes:

> Eric Abrahamsen <eric@ericabrahamsen.net> writes:
>
>> > Ok, that would be the last resort.  I had hoped for a nice solution in
>> > Emacs.
>>
>> But what would a "nice solution" look like? I suppose you could always
>> write your own completion-at-point function.
>
> That I could somewhere create a record for a group consisting of a name,
> a mail alias, and a number of Email addresses.  Writing my own
> completion-at-point function doesn't count as "nice".

No, I suppose not. You can also build mail abbrevs without too much
trouble using the function `define-mail-abbrev', that might be your
easiest entry-point. I'm still not quite clear on what your ideal
workflow would be here.

>> > BTW, anyone knowing why BBDB comes with a manual that looks like
>>
>> BBDB's lack of manual is half of why I ended up writing EBDB (only
>> partly joking).
>
> Ok, I may be switching because of the manual (also only partly joking).
> No, I wanted to have a look at it for a longer time anyway ;-)

I've been sort of tracking the vCard RFC when considering what record
classes to add, and vCard does have the concept of a "group" record,
which is just a collection of other records. This wouldn't save you any
work, however, since you'd still need to create the person records, and
add them to the group. I suppose I could allow adding mail fields
instead of "people" to the group, but it still seems roundabout.

Eric

Re: Define groups of recipients

[Bob Newell]

Michael Heerdegen <michael_heerdegen@web.de> writes:
> Ok, thanks.  Do I get it right that this approach involves to add an
> individual contact for any person that should be part of the group?
> Is
> there any way to avoid this?

As others have pointed out, there are workarounds that I think
in the end are not as good as using BBDB as designed. At
first, adding a mail-alias for each user in a group may seem
like a bit of a pain, and it is, but ultimately it is much
easier to maintain when members enter and leave a group. It is
very simple to call up a BBDB record and add or remove an
alias.

It would be possible to do some coding to accept a list of
email addresses and add an alias for each one, but unless you
have dozens of members on the list that would likely take
longer (although it would be reusable).

On the other hand, I recently had a list of 50 to create and
it took me around five minutes. I didn't think that was
excessive.

-- 
Bob Newell
Honolulu, Hawai`i
* Via Gnus/BBDB/Org/Emacs/Linux *

Re: Define groups of recipients

[Michael Heerdegen]

Bob Newell <bobnewell@bobnewell.net> writes:

> > Ok, thanks.  Do I get it right that this approach involves to add an
> > individual contact for any person that should be part of the group?
> > Is there any way to avoid this?
>
> As others have pointed out, there are workarounds that I think
> in the end are not as good as using BBDB as designed. At
> first, adding a mail-alias for each user in a group may seem
> like a bit of a pain, and it is, but ultimately it is much
> easier to maintain when members enter and leave a group. It is
> very simple to call up a BBDB record and add or remove an
> alias.
>
> It would be possible to do some coding to accept a list of
> email addresses and add an alias for each one, but unless you
> have dozens of members on the list that would likely take
> longer (although it would be reusable).
>
> On the other hand, I recently had a list of 50 to create and
> it took me around five minutes. I didn't think that was
> excessive.

Ok, thanks for your opinion.  Maybe I'll just do what you suggest.  BTW,
one problem is that I only have a list of addresses, I don't even know
the names of these people.  It feels a bit strange to add a record for
them.


Michael.

Re: Define groups of recipients

[Dan Christensen]

On Mar 17, 2019, Michael Heerdegen <michael_heerdegen@web.de> wrote:

> one problem is that I only have a list of addresses, I don't even know
> the names of these people.  It feels a bit strange to add a record for
> them.

Can you say what you don't like about the ~/.mailrc solution?  You just
add a single line to that file like this:

alias groupname   user1@domain1.com, user2@domain2.com, etc@etc.com

and then you compose a message to groupname and hit SPC or , after
groupname and it expands.

After editing .mailrc, you have to restart Gnus, or use something like:

(defun jdc-rebuild-mail-abbrevs ()
  (interactive)
  (rebuild-mail-abbrevs (concat (getenv "HOME") "/.mailrc"))
  nil)

(global-set-key [f11 ?R] 'jdc-rebuild-mail-abbrevs)

Dan

Re: Define groups of recipients

[Emanuel Berg]

Dan Christensen wrote:

> Can you say what you don't like about the
> ~/.mailrc solution? You just add a single
> line to that file like this:
>
> alias groupname user1@domain1.com,
> user2@domain2.com, etc@etc.com

Indeed, but it is even better to do it this
way:

    # parents
    alias dad "Dad <some@mail.com>"
    alias mom "Mom <some-other@mail.com>"
    alias parents dad mom

:)

-- 
underground experts united
http://user.it.uu.se/~embe8573

Re: Define groups of recipients

[Michael Heerdegen]

Dan Christensen <jdc@uwo.ca> writes:

> Can you say what you don't like about the ~/.mailrc solution?

The only thing I dislike is that it isn't completely in Emacs.

> You just add a single line to that file like this:
>
> alias groupname   user1@domain1.com, user2@domain2.com, etc@etc.com
>
> and then you compose a message to groupname and hit SPC or , after
> groupname and it expands.

Ok, thanks, that works.  I think I'll just go with that.

Michael.

Re: Define groups of recipients

[Michael Heerdegen]

Dan Christensen <jdc@uwo.ca> writes:

> and then you compose a message to groupname and hit SPC or , after
> groupname and it expands.

FWIW I guess this could also work with normal completion but
`message-expand-name' doesn't try when you use bbdb, even if bbdb
doesn't have a completion:

#+begin_src emacs-lisp
(defun message-expand-name ()
  (cond ((and (memq 'eudc message-expand-name-databases)
		    (boundp 'eudc-protocol)
		    eudc-protocol)
	 (eudc-expand-inline))
	((and (memq 'bbdb message-expand-name-databases)
	      (fboundp 'bbdb-complete-name))
         (let ((starttick (buffer-modified-tick)))
           (or (bbdb-complete-name)
               ;; Apparently, bbdb-complete-name can return nil even when
               ;; completion took place.  So let's double check the buffer was
               ;; not modified.
               (/= starttick (buffer-modified-tick)))))
	(t
	 (expand-abbrev))))
#+end_src

I wonder if this is intended.  I guess it is, at least it has been like
that for a long time.


Michael.

Re: Define groups of recipients

[Eric Abrahamsen]

Michael Heerdegen <michael_heerdegen@web.de> writes:

> Dan Christensen <jdc@uwo.ca> writes:
>
>> and then you compose a message to groupname and hit SPC or , after
>> groupname and it expands.
>
> FWIW I guess this could also work with normal completion but
> `message-expand-name' doesn't try when you use bbdb, even if bbdb
> doesn't have a completion:
>
> #+begin_src emacs-lisp
> (defun message-expand-name ()
>   (cond ((and (memq 'eudc message-expand-name-databases)
> 		    (boundp 'eudc-protocol)
> 		    eudc-protocol)
> 	 (eudc-expand-inline))
> 	((and (memq 'bbdb message-expand-name-databases)
> 	      (fboundp 'bbdb-complete-name))
>          (let ((starttick (buffer-modified-tick)))
>            (or (bbdb-complete-name)
>                ;; Apparently, bbdb-complete-name can return nil even when
>                ;; completion took place.  So let's double check the buffer was
>                ;; not modified.
>                (/= starttick (buffer-modified-tick)))))
> 	(t
> 	 (expand-abbrev))))
> #+end_src
>
> I wonder if this is intended.  I guess it is, at least it has been like
> that for a long time.

It's also very much not the way that Emacs completion is meant to work,
though. I don't hold out much hope that it could be changed (as you
note, it's been like this a long time), but for EBDB at least I've had
to jump through a lot of hoops to get completion to work.

Re: Define groups of recipients

[Emanuel Berg]

Eric Abrahamsen wrote:

> It's also very much not the way that Emacs
> completion is meant to work, though. I don't
> hold out much hope that it could be changed
> (as you note, it's been like this a long
> time), but for EBDB at least I've had to jump
> through a lot of hoops to get completion
> to work.

It works great with ~/.mailrc!

What's to change?

Unnecessary work with zero or close to zero
gain, the risk of introducing bugs,
incompatibility issues, people's own little
hacks gets screwed, and so on.

-- 
underground experts united
http://user.it.uu.se/~embe8573

Re: Define groups of recipients

[Eric Abrahamsen]

Emanuel Berg <moasenwood@zoho.eu> writes:

> Eric Abrahamsen wrote:
>
>> It's also very much not the way that Emacs
>> completion is meant to work, though. I don't
>> hold out much hope that it could be changed
>> (as you note, it's been like this a long
>> time), but for EBDB at least I've had to jump
>> through a lot of hoops to get completion
>> to work.
>
> It works great with ~/.mailrc!
>
> What's to change?
>
> Unnecessary work with zero or close to zero
> gain, the risk of introducing bugs,
> incompatibility issues, people's own little
> hacks gets screwed, and so on.

Oh, I don't mean Gnus' use of ~/.mailrc at all -- that's a solid
mechanism that shouldn't be messed with. It's been there for 20 years,
and could be there for 20 more. One of the things I appreciate about
Gnus is that it will happily use these old conventions.

What's weird is message-mode's interaction with completion-at-point
functions. Completion at point is supposed to work by adding a bunch of
functions to `completion-at-point-functions', and then each one gets a
chance to complete or not -- simple enough.

Instead message-mode binds TAB to its own function, which then calls
`completion-at-point', which calls message-mode's own completion at
point function, which looks at where it is in the composition buffer,
and either completes a group name or asks either BBDB or EUDC if it
wants to complete a contact name. There's a customization option to
set the contact name database, but the only valid values are 'bbdb or
'eudc.

It makes it fairly difficult to use anything else.

Eric

Re: Define groups of recipients

[Emanuel Berg]

Eric Abrahamsen wrote:

>> It works great with ~/.mailrc! What's to
>> change? Unnecessary work with zero or close
>> to zero gain, the risk of introducing bugs,
>> incompatibility issues, people's own little
>> hacks gets screwed, and so on.
>
> Oh, I don't mean Gnus' use of ~/.mailrc at
> all -- that's a solid mechanism that
> shouldn't be messed with. It's been there for
> 20 years, and could be there for 20 more.
> One of the things I appreciate about Gnus is
> that it will happily use these
> old conventions.

OK, sorry I misunderstood so you had to spend
time explaining it to me.

-- 
underground experts united
http://user.it.uu.se/~embe8573