Re: ASCII encryption

Re: ASCII encryption

[Jambunathan K]

Richard H Lee <ricardohenrylee@gmail.com> writes:

> On 16/04/12 12:51, Bernardo wrote:
>> there's a related epa-armor variable; (haven't tried it myself)
>>
>> ,----
>> | epa-armor is a variable defined in `epa.el'.
>> | Its value is nil
>> |
>> | Documentation:
>> | *If non-nil, epa commands create ASCII armored output.
>> |
>> | You should bind this variable with `let', but do not set it globally.
>> `----
>
> Yes, thanks, I tried this out yesterday. It works.
>
> However the EasyPG assistant integration with armor ascii files is not
> as nice with gpg binary files.

If `epa-armor' is not supported it should be removed or be made in to a
defconst.  Or it could be crptically renamed so that people don't
discover it accidentally.  The info manual can be augmented to mention
why the author thinks using it globally is not a good idea.

epa-armor is like the biblical forbidden fruit with hints of both
pleasure and pain.  The ultimate end it achieves is to leave mere
mortals in a state of perpetual confusion on private matters.
-- 

Re: ASCII encryption

[Richard H Lee]

On 16/04/12 15:18, Jambunathan K wrote:
> If `epa-armor' is not supported it should be removed or be made in to a
> defconst.  Or it could be crptically renamed so that people don't
> discover it accidentally.  The info manual can be augmented to mention
> why the author thinks using it globally is not a good idea.
>
> epa-armor is like the biblical forbidden fruit with hints of both
> pleasure and pain.  The ultimate end it achieves is to leave mere
> mortals in a state of perpetual confusion on private matters.

So I that (setq epa-armor t) in my emacs file is not a good idea?

Re: ASCII encryption

[Jambunathan K]

Richard H Lee <ricardohenrylee@gmail.com> writes:

> On 16/04/12 15:18, Jambunathan K wrote:
>> If `epa-armor' is not supported it should be removed or be made in to a
>> defconst.  Or it could be crptically renamed so that people don't
>> discover it accidentally.  The info manual can be augmented to mention
>> why the author thinks using it globally is not a good idea.
>>
>> epa-armor is like the biblical forbidden fruit with hints of both
>> pleasure and pain.  The ultimate end it achieves is to leave mere
>> mortals in a state of perpetual confusion on private matters.
>
> So I that (setq epa-armor t) in my emacs file is not a good idea?

That's what the last line of the docstring says.  It is confusing and
also scary.

,----[ C-h v epa-armor RET ]
| epa-armor is a variable defined in `epa.el'.
| Its value is nil
| 
| Documentation:
| *If non-nil, epa commands create ASCII armored output.
| 
| You should bind this variable with `let', but do not set it globally.
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
`----

Re: ASCII encryption

[Daiki Ueno]

Jambunathan K <kjambunathan@gmail.com> writes:

>>> epa-armor is like the biblical forbidden fruit with hints of both
>>> pleasure and pain.  The ultimate end it achieves is to leave mere
>>> mortals in a state of perpetual confusion on private matters.
>>
>> So I that (setq epa-armor t) in my emacs file is not a good idea?
>
> That's what the last line of the docstring says.  It is confusing and
> also scary.

> | You should bind this variable with `let', but do not set it globally.

I think it is normal as coding-system-for-write, etc.  What's the problem?

epa-encrypt-file could surely be improved for ASCII armor though.  One
idea is to query user with "Save File: " prompt and if the answer ends
with ".asc", encrypt in ASCII armor instead of binary.

Regards,
-- 
Daiki Ueno