Small suggestion for future Oort

Small suggestion for future Oort

[Harry Putnam]

There has always been some confusion surrounding Wide reply (F).  It
is a nice feature as it stands but can lead to bad results, if the
user is not observant.

I get bit by it fairly often... and it can have bad results in this
way.

You grow accustomed to using `F'  for all replies, at least I have.
And justifiably so since many of my groups contain a `to-address' in G p
(group params) 

However when that param is lacking, I send unwanted Cc's fairly often.

In one case it could have had disastrous results.  I mailed a tmp passwd
to someone and didn't notice the Cc in the original message to a mail
list.  Naturally the passwd was posted on the list as well as going
to the intended recipient.  Had I not noticed and had that list been a
big public one, it could have led to a root compromise.

I'm aware of the dangers of sending a passwd in clear text, but it
should be fairly safe, if you keep a tight machine and are not seeing
signs of someone reading your mail.  And it was a temp passwd that
would at best get someone logged in as a restricted user.  Unless that
someone was an experienced hack, it should not result in great harm.

However had I posted it on a busy public group, the chances of
attracting an experienced `black hat' would be far greater.

I set gnus to show all outgoing headers long ago for this very reason
but apparently, just having that Cc visible in the composition buffer
still isn't enough to always get my attention.

I want a feature that queries me about the Cc field on sending.
Something like the warning you get with over 79 chars on a line.  Or
when you left the subject field empty.

It could be toggled on or off.  And would stop the sending action
and ask the user if they want the Cc field as is (with a display in
echo area).  RET would be `no' or blank, or press up-arrow key for the
default (as is) then RET.

People who use `Cc' a lot could turn it off or it could be off by default.

When `on', the query  would only happen when the Cc field was not blank.

Re: Small suggestion for future Oort

[Kai Großjohann]

I wonder: maybe `F' should 

    * in a newsgroup, only post to that group,
    * in a mailing list, only send to that list.

Then `S W' could be used for Cc'ing the author in both cases.

Thoughts?

kai
-- 
~/.signature: No such file or directory

Re: Small suggestion for future Oort

[David S. Goldberg]

> I wonder: maybe `F' should 
>     * in a newsgroup, only post to that group,
This maybe doesn't sound so hard but...

>     * in a mailing list, only send to that list.

How would it know absent a group parameter that said explicitly that
this group is for a mailing list?  Would it look at the names and
assume that one of them is a mailing list?  I hope not.  That would
break a lot of things for me.

-- 
Dave Goldberg
dsg@world.std.com

Re: Small suggestion for future Oort

[Paul Jarc]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
> I wonder: maybe `F' should 
> 
>     * in a newsgroup, only post to that group,
>     * in a mailing list, only send to that list.

Isn't this behavior already available to those who want it with
message-use-followup-to and to-address?

> Then `S W' could be used for Cc'ing the author in both cases.

How do those settings affect S W?

      
paul

Re: Small suggestion for future Oort

[Benjamin Rutt]

Harry Putnam <reader@newsguy.com> writes:

> I want a feature that queries me about the Cc field on sending.
> Something like the warning you get with over 79 chars on a line.  Or
> when you left the subject field empty.
>
> It could be toggled on or off.  And would stop the sending action
> and ask the user if they want the Cc field as is (with a display in
> echo area).  RET would be `no' or blank, or press up-arrow key for the
> default (as is) then RET.

Try setting the following variable (added to Oort in the past few
months) to t, I think that will do most of what you wish:

,----[ C-h v message-wide-reply-confirm-recipients RET ]
| message-wide-reply-confirm-recipients's value is 
| t
| 
| Documentation:
| Whether to confirm a wide reply to multiple email recipients.
| If this variable is nil, don't ask whether to reply to all recipients.
| If this variable is non-nil, pose the question "Reply to all
| recipients?" before a wide reply to multiple recipients.  If the user
| answers yes, reply to all recipients as usual.  If the user answers
| no, only reply back to the author.
| 
| You can customize this variable.
| 
| Defined in `message'.
`----

-- 
Benjamin

Re: Small suggestion for future Oort

[Vincent Bernat]

OoO Pendant le repas du lundi 06 août 2001, vers 19:40, Harry Putnam
<reader@newsguy.com> disait:

> I want a feature that queries me about the Cc field on sending.
> Something like the warning you get with over 79 chars on a line.  Or
> when you left the subject field empty.

Why not add this feature using a pre-send hook ?
-- 
BAGMAN, IS NOT A LEGITIMATE CAREER CHOICE
BAGMAN, IS NOT A LEGITIMATE CAREER CHOICE
BAGMAN, IS NOT A LEGITIMATE CAREER CHOICE
-+- Bart Simpson on chalkboard in episode 2F10

Re: Small suggestion for future Oort

[Kai Großjohann]

dsg@world.std.com (David S. Goldberg) writes:

>>     * in a mailing list, only send to that list.
> 
> How would it know absent a group parameter that said explicitly that
> this group is for a mailing list?

Indeed.  Sorry.  Maybe it should just refrain from sending to the
person in the From header.

kai
-- 
~/.signature: No such file or directory

Re: Small suggestion for future Oort

[Harry Putnam]

Benjamin Rutt <brutt+news@bloomington.in.us> writes:

> Try setting the following variable (added to Oort in the past few
> months) to t, I think that will do most of what you wish:
> 
> ,----[ C-h v message-wide-reply-confirm-recipients RET ]
> | message-wide-reply-confirm-recipients's value is 
> | t

That looks like what I was after.  I don't have that variable on my
Oort3* and everytime I update to a recent Oort4* I loose all font
locking in Messages mode buffers.

I'm running emacs-21 here.  But sort of stuck on Oort-3 until I find
out what is jacking up font-locking.

Just checked on a recent Oort-4 and found the variable you mention,
but for now I prefer having font-locking in compose buffers to having
that variable set to t... It'll come in handy when I get lined out here.

Re: Small suggestion for future Oort

[Nuutti Kotivuori]

Kai Großjohann wrote:
> dsg@world.std.com (David S. Goldberg) writes:
> 
>>>     * in a mailing list, only send to that list.
>> 
>> How would it know absent a group parameter that said explicitly
>> that this group is for a mailing list?
> 
> Indeed.  Sorry.  Maybe it should just refrain from sending to the
> person in the From header.

Well if the message carries a Mail-Post header then we do know the
mailing list. If it doesn't, we could have a group parameter to
facilitate it.

-- Naked