emacs-lock enhancements

emacs-lock enhancements

[Juanma Barranquero]

I'm interested in incorporating into Emacs some of the functionality
(not code) of Noah Friedman's protbuf.el, i.e., the ability to protect
a buffer against killing.

Currently, emacs-lock.el does something similar, though it has some problems:

 - It's not a proper minor-mode, so the locked status doesn't show in
the modeline.
 - It conflates two meanings of "locked": it protects against killing,
and it avoids Emacs exiting while a buffer is locked.
 - It deals with telnet and shell buffers as special cases, instead of
handling all buffers with live processes.

I'm planning on enhancing emacs-lock's functionality to make possible
separating the dont-kill and dont-exit-emacs flags (so a buffer can be
non-kill, but does not stop Emacs from exiting, for example).

What is the best interface for that: having two different minor modes
(let's say "emacs-lock-dont-kill-mode' and
`emacs-lock-dont-exit-mode'), or just one and somehow setting a flag
to indicate the desired level of protection, sort of like
overwrite-mode distinguishes between "normal" and "binary" modes?

    Juanma

Re: emacs-lock enhancements

[Stefan Monnier]

> I'm interested in incorporating into Emacs some of the functionality
> (not code) of Noah Friedman's protbuf.el, i.e., the ability to protect
> a buffer against killing.

We already have a fair bit of Noah's code, so at least from a copyright
point of view, copying code shouldn't be a problem (we'd still need Noah
to agreed, of course, but the prognosis is pretty good).

Of course, reimplementing might actually be simpler from a technical
point of view.

> Currently, emacs-lock.el does something similar, though it has some problems:
>  - It's not a proper minor-mode, so the locked status doesn't show in
>    the modeline.

It would be good to fix it so it uses define-minor-mode.

>  - It deals with telnet and shell buffers as special cases, instead of
>    handling all buffers with live processes.

Sounds like it would be good to improve that as well.

> What is the best interface for that: having two different minor modes
> (let's say "emacs-lock-dont-kill-mode' and
> `emacs-lock-dont-exit-mode'), or just one and somehow setting a flag
> to indicate the desired level of protection, sort of like
> overwrite-mode distinguishes between "normal" and "binary" modes?

I single major mode sounds sufficient.  The distinction between the two
cases could even be done with a global variable, rather than being
per-buffer.


        Stefan

Re: emacs-lock enhancements

[Juanma Barranquero]

On Tue, Sep 22, 2009 at 17:02, Stefan Monnier <monnier@iro.umontreal.ca> wrote:

> We already have a fair bit of Noah's code, so at least from a copyright
> point of view, copying code shouldn't be a problem (we'd still need Noah
> to agreed, of course, but the prognosis is pretty good).
>
> Of course, reimplementing might actually be simpler from a technical
> point of view.

I'll take a look. I'm not sure whether it's easier to incorporate
emacs-lock into protbuf, or vice versa.

> I single major mode sounds sufficient.  The distinction between the two
> cases could even be done with a global variable, rather than being
> per-buffer.

OK (assuming you mean "single minor mode" :-)

    Juanma

Re: emacs-lock enhancements

[Stefan Monnier]

>> I single major mode sounds sufficient.  The distinction between the two
>> cases could even be done with a global variable, rather than being
>> per-buffer.
> OK (assuming you mean "single minor mode" :-)

Wihch gparted of me answering ist clear?


        Stefan