Crypto key bindings considered harmful

Crypto key bindings considered harmful

[Florian Weimer]

Key sequences such as 'C-c C-m c' are a *very* bad idea.  Guess
why. 8-/

The crypto support has more serious usability problems (e.g. if the
'#secure' tag is not at the beginning of the message, it is ignored
without warning *sigh*).  Is anybody looking at such issues?  I plan
to write a general check list to evaluate crypto-enabled mail clients,
but I can't promise anything right now (too much other tasks).

Re: Crypto key bindings considered harmful

[Kai Großjohann]

Florian Weimer <fw@deneb.enyo.de> writes:

> Key sequences such as 'C-c C-m c' are a *very* bad idea.  Guess
> why. 8-/

It's too close to C-x C-c?

At least I did guess.  Don't blame me when the guess was wrong ;-p
-- 
This line is not blank.

Re: Crypto key bindings considered harmful

[Andreas Fuchs]

Today, Florian Weimer <fw@deneb.enyo.de> wrote:
> Key sequences such as 'C-c C-m c' are a *very* bad idea.  Guess
> why. 8-/

Ugh, right.

> The crypto support has more serious usability problems (e.g. if the
> '#secure' tag is not at the beginning of the message, it is ignored
> without warning *sigh*).  

Add "if there are more than one #secure tags present, only the first
one is used" - much fun can be had with an "auto-sign?" question when
sending the mail, but the mail should be encrypted.

> Is anybody looking at such issues?  I plan to write a general check
> list to evaluate crypto-enabled mail clients, but I can't promise
> anything right now (too much other tasks).

Sounds like a great idea. 

-- 
Andreas Fuchs, <asf@acm.org>, asf@jabber.at, antifuchs
irc.freenode.net's #emacs - online emacs advice from IRC addicts

Re: Crypto key bindings considered harmful

[Chris Green]

Andreas Fuchs <asf@void.at> writes:

> Today, Florian Weimer <fw@deneb.enyo.de> wrote:
>> Key sequences such as 'C-c C-m c' are a *very* bad idea.  Guess
>> why. 8-/
>
> Ugh, right.

Yup.  When doing crypto in gnus, it's one of the few times I make
myself use the Attachements Menu since it's not going have any
possibility of me just hitting C-c C-c
-- 
Chris Green <cmg@sourcefire.com>
To err is human, to moo bovine.

Re: Crypto key bindings considered harmful

[Romain FRANCOISE]

Chris Green <cmg@sourcefire.com> writes:

> Yup.  When doing crypto in gnus, it's one of the few times I make
> myself use the Attachements Menu since it's not going have any
> possibility of me just hitting C-c C-c

I just hit C-c RET, then wait to see "C-c RET-" in the minibuffer to
ensure that I'm doing the right thing.  I don't remember ever hitting
C-c C-c by accident.

-- 
Romain FRANCOISE <romain@orebokech.com> | It was fourteen degrees below
it's a miracle -- http://orebokech.com/ | on a screeching march 23.

Re: Crypto key bindings considered harmful

[Andreas Fuchs]

Today, Romain FRANCOISE <romain@orebokech.com> wrote:
> I just hit C-c RET, then wait to see "C-c RET-" in the minibuffer to
> ensure that I'm doing the right thing.  I don't remember ever hitting
> C-c C-c by accident.

This can happen with an over-eager keyboard. I've started using a
y-or-n-p query in `message-send-hook', which (error )s if "n" is
pressed, thus preventing gnus from sending a message..

Anyway, this feels somehow ... imperfect (:

-- 
Andreas Fuchs, <asf@acm.org>, asf@jabber.at, antifuchs
irc.freenode.net's #emacs - online emacs advice from IRC addicts

Re: Crypto key bindings considered harmful

[Simon Josefsson]

Florian Weimer <fw@deneb.enyo.de> writes:

> Key sequences such as 'C-c C-m c' are a *very* bad idea.  Guess
> why. 8-/

Right.  How about C-c q *?  Or C-RET *?  The latter is shorter.

> The crypto support has more serious usability problems (e.g. if the
> '#secure' tag is not at the beginning of the message, it is ignored
> without warning *sigh*).  Is anybody looking at such issues?

Oops.

> I plan to write a general check list to evaluate crypto-enabled mail
> clients, but I can't promise anything right now (too much other
> tasks).

This would be quite useful.

Re: Crypto key bindings considered harmful

[lawrence mitchell]

[crypto signing in message]

Simon Josefsson wrote:

> Florian Weimer <fw@deneb.enyo.de> writes:

>> Key sequences such as 'C-c C-m c' are a *very* bad idea.  Guess
>> why. 8-/

> Right.  How about C-c q *?  Or C-RET *?  The latter is shorter.

C-RET is a non-starter on terminals, so I'd go for the former.
And, for some reason, I think q is mnemonic for crypt :), go
figure.

[...]

-- 
lawrence mitchell <s0198183+ding@sms.ed.ac.uk>

Re: Crypto key bindings considered harmful

[Florian Weimer]

Simon Josefsson <jas@extundo.com> writes:

> How about C-c q *?

Is this an FSF-compatible key binding?

IIRC we moved away from 'C-c letter' sequences for policy reasons.

Re: Crypto key bindings considered harmful

[Simon Josefsson]

Florian Weimer <fw@deneb.enyo.de> writes:

> Simon Josefsson <jas@extundo.com> writes:
>
>> How about C-c q *?
>
> Is this an FSF-compatible key binding?
>
> IIRC we moved away from 'C-c letter' sequences for policy reasons.

Ah, right.  Was there a reason we don't use M-<whatever> in message
mode?  M-o and M-p are undefined.

Re: Crypto key bindings considered harmful

[Lars Magne Ingebrigtsen]

Simon Josefsson <jas@extundo.com> writes:

> Ah, right.  Was there a reason we don't use M-<whatever> in message
> mode?  M-o and M-p are undefined.

I think there was some obscure mode that used those, or something... 

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