bug? mm-display-external does delete-other-windows

bug? mm-display-external does delete-other-windows

[Ingo Lohmar]

Hi Emacs,

Due to some other changes, I am recently affected by the "aggressive"
behavior of gnus' mm-display-external function (in mm-decode.el): When
displaying a part in its own window (not inline), it calls
(delete-other-windows), destroying my window configuration.

Is this behavior wanted when used from inside gnus?  Are mm-decode.el's
functions meant to be used by other packages as well?  In my case the
notmuch Emacs client [1] uses mm-display-part which calls
mm-display-external under certain conditions.

If it's ok to use the function (I hope it is), I would consider
(delete-other-windows) a bug.  Otherwise, are there any plans to extract
the MIME functionality into a separate package?

[1] http://notmuchmail.org

Re: bug? mm-display-external does delete-other-windows

[Mike Kupfer]

Ingo Lohmar wrote:

> Are mm-decode.el's
> functions meant to be used by other packages as well?  In my case the
> notmuch Emacs client [1] uses mm-display-part which calls
> mm-display-external under certain conditions.

Hi Ingo, yes, the mm functions are documented as general-purpose.
There's at least one other package that uses them (MH-E).

I don't see any problems with MH-E when I use an external viewer.  I
guess this is because MH-E is invoking mm-display-external directly in
response to a user keystroke, rather than going through mm-display-part.
And MH-E passes in a string, not a function, as the method.

I don't know how much the above helps you, but I wanted to reassure you
that the mm layer is supposed to be general-purpose.

regards,
mike

Re: bug? mm-display-external does delete-other-windows

[Lars Ingebrigtsen]

Ingo Lohmar <i.lohmar@gmail.com> writes:

> If it's ok to use the function (I hope it is), I would consider
> (delete-other-windows) a bug.

Yes, I think it's a bug.  Please report it with `report-emacs-bug'.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Re: bug? mm-display-external does delete-other-windows

[Ingo Lohmar]

On Mon, Feb 29 2016 14:23 (+1100), Lars Ingebrigtsen wrote:

> Ingo Lohmar <i.lohmar@gmail.com> writes:
>
>> If it's ok to use the function (I hope it is), I would consider
>> (delete-other-windows) a bug.
>
> Yes, I think it's a bug.  Please report it with `report-emacs-bug'.

[I erroneously sent this as a private mail at first, sorry.]

Reported as #22861.  I cannot think of a scenario where
(delete-other-windows) would be ok --- I will just remove it as the fix
(if nobody objects here).

I assume it's still correct to do this on emacs-25, or isn't it?

Re: bug? mm-display-external does delete-other-windows

[Lars Ingebrigtsen]

Ingo Lohmar <i.lohmar@gmail.com> writes:

> Reported as #22861.  I cannot think of a scenario where
> (delete-other-windows) would be ok --- I will just remove it as the fix
> (if nobody objects here).

It looks like the intention is to do the window manipulation if the
function is called from Gnus.  I don't know whether that's a good thing
to do from Gnus or not, though, but just removing it is probably the
wrong thing to do.

If this is something that should be done, then Gnus should pass
something in to make all that...  happen...

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Re: bug? mm-display-external does delete-other-windows

[Ingo Lohmar]

On Wed, Mar 02 2016 16:41 (+0000), Lars Ingebrigtsen wrote:

> It looks like the intention is to do the window manipulation if the
> function is called from Gnus.  I don't know whether that's a good thing
> to do from Gnus or not, though, but just removing it is probably the
> wrong thing to do.
>
> If this is something that should be done, then Gnus should pass
> something in to make all that...  happen...

Sooooo...  that would mean only somebody who uses Gnus, or better yet,
has intimate knowledge of its inner workings (wink, wink) can take care
of this bug?

I'll keep my private hack then, for the time being.

Thanks,
Ingo