GNU bug report logs - #48916
28.0.50; allow windmove to select windows with the 'no-other-window parameter

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 48916 in the body.
You can then email your comments to 48916 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: pillule <pillule <at> riseup.net>
To: bug-gnu-emacs <at> gnu.org
Subject: 28.0.50; allow windmove to select windows with the 'no-other-window
 parameter
Date: Tue, 08 Jun 2021 12:09:21 +0200

[Message part 1 (text/plain, inline)]

Hi,

One may want to customize its `display-buffer-alist' carefully to 
apply the 'no-other-window parameter to eg side windows so they 
are not selected by the `other-window' command and to create 
custom functions of the family (eg creating a 
`find-file-in-other-window' that does not select windows with the 
no-other-window parameter)

Unfortunately by doing so, we actually limit the use of the 
windmove commands since `windmove-find-other-window' always ignore 
windows with this parameter.
It is not a fatality since `window-in-direction' can receive an 
argument to choose if we want to ignore this parameter or not.

This draft implements an user option to let the user choose that 
behavior.

Because it applies to `windmove-find-other-window', it allows the 
commands windmove-right, windmove-left, windmove-down, windmove-up 
to select windows with the no-other-window parameter, but not for 
the `windmove-display-*' functions. I suppose it is the desired 
behavior.

[0001-User-option-to-select-no-other-window-with-windmove.patch (text/x-diff, attachment)]

Message #8 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: pillule <pillule <at> riseup.net>
Cc: 48916 <at> debbugs.gnu.org
Subject: Re: bug#48916: 28.0.50;
 allow windmove to select windows with the 'no-other-window parameter
Date: Tue, 08 Jun 2021 14:44:14 +0300

> From: pillule <pillule <at> riseup.net>
> Date: Tue, 08 Jun 2021 12:09:21 +0200
> 
> One may want to customize its `display-buffer-alist' carefully to 
> apply the 'no-other-window parameter to eg side windows so they 
> are not selected by the `other-window' command and to create 
> custom functions of the family (eg creating a 
> `find-file-in-other-window' that does not select windows with the 
> no-other-window parameter)
> 
> Unfortunately by doing so, we actually limit the use of the 
> windmove commands since `windmove-find-other-window' always ignore 
> windows with this parameter.
> It is not a fatality since `window-in-direction' can receive an 
> argument to choose if we want to ignore this parameter or not.
> 
> This draft implements an user option to let the user choose that 
> behavior.

Thanks.  I don't use windmove, so I will let others comment.  But
please allow me a few minor nits:

> * lisp/windmove.el
>    (windmove-ignore-no-other-window): add this new user option
>    (windmove-find-other-window): uses windmove-ignore-no-other-window
> to choose whether windmove can access to the window with the
> 'no-other-window property.

This isn't formatted accoring to our rules.  In particular, each
sentence after the colon should begin with a capital letter and end
with a period.  See CONTRIBUTE for more details (and I suggest to use
Emacs commands for writing log messages, as they will take care of
some routine parts of the formatting for you).

> +(defcustom windmove-ignore-no-other-window nil

This name is not the best one.  For starters, "ignore-no" is a kind of
double negation, which makes it harder to understand and remember.
Can you come up with a better description of what exactly is ignored
here?

> +  "Whether the windmove commands are allowed to target all type of windows,

The first line of a doc string should be a complete sentence (it is)
and end with a period.

> +If this variable is set to t, `windmove-find-other-window--side' and
   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
"If non-nil, ..."

> +subsequently all interactive windmove commandswill ignore the
                                         ^^^^^^^^^^^^
A typo.

> +no-other-window parameter."

If you reference this parameter, I think you should say that it's a
parameter to be applied by display-buffer-alist's actions.

> +  :type 'boolean
> +  :group 'windmove)

New defcustoms should have a :version tag.  Also, I believe we don't
like redundant :group tags, such as the one here.

Message #11 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: pillule <pillule <at> riseup.net>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: pillule <pillule <at> riseup.net>, 48916 <at> debbugs.gnu.org
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Tue, 08 Jun 2021 15:28:34 +0200

[Message part 1 (text/plain, inline)]

Eli Zaretskii <eliz <at> gnu.org> writes:

> Thanks.  I don't use windmove, so I will let others comment. 
> But
> please allow me a few minor nits:

Your ‘minor nits’ are indeed welcome.

>> * lisp/windmove.el
>>    (windmove-ignore-no-other-window): add this new user option
>>    (windmove-find-other-window): uses 
>>    windmove-ignore-no-other-window
>> to choose whether windmove can access to the window with the
>> 'no-other-window property.
>
> This isn't formatted accoring to our rules.  In particular, each
> sentence after the colon should begin with a capital letter and 
> end
> with a period.  See CONTRIBUTE for more details (and I suggest 
> to use
> Emacs commands for writing log messages, as they will take care 
> of
> some routine parts of the formatting for you).

So that means you are recommending to use VC instead of Magit 
(that I used) for committing changes ?
Maybe Magit have a something that I am not aware for this purpose 
?

>> +(defcustom windmove-ignore-no-other-window nil
>
> This name is not the best one.  For starters, "ignore-no" is a 
> kind of
> double negation, which makes it harder to understand and 
> remember.
> Can you come up with a better description of what exactly is 
> ignored
> here?

Yes, i think 'windmove-move-in-all-windows' will be more explicit.

>> +  "Whether the windmove commands are allowed to target all 
>> type of windows,
>
> The first line of a doc string should be a complete sentence (it 
> is)
> and end with a period.
>
>> +If this variable is set to t, 
>> `windmove-find-other-window--side' and
>    ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> "If non-nil, ..."
>
>> +subsequently all interactive windmove commandswill ignore the
>                                          ^^^^^^^^^^^^
> A typo.
>
>> +no-other-window parameter."
>
> If you reference this parameter, I think you should say that 
> it's a
> parameter to be applied by display-buffer-alist's actions.
>> +  :type 'boolean
>> +  :group 'windmove)
>
> New defcustoms should have a :version tag.
> Also, I believe we don't  like redundant :group tags, such as 
> the one here.

got them. the last one may be a little bit confusing for the 
reader that is not aware of it because all others defcustoms of 
window.el applies a :group tag. Do you want that I remove the 
others unnecessary ones ?

[0002-User-option-to-select-no-other-window-with-windmove.patch (text/x-diff, attachment)]

[Message part 3 (text/plain, inline)]

--

Message #14 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: pillule <pillule <at> riseup.net>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: pillule <pillule <at> riseup.net>, 48916 <at> debbugs.gnu.org
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Tue, 08 Jun 2021 15:57:49 +0200

Eli Zaretskii <eliz <at> gnu.org> writes:

>> +subsequently all interactive windmove commandswill ignore the
>                                          ^^^^^^^^^^^^
> A typo.
>

No, it seems to be a bug on my end. Somewhere on a file 
transfer. Sorry I am not sure I will be able to fix that one 
easily.

--

Message #17 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: pillule <pillule <at> riseup.net>
Cc: 48916 <at> debbugs.gnu.org
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Tue, 08 Jun 2021 17:12:08 +0300

> From: pillule <pillule <at> riseup.net>
> Cc: pillule <pillule <at> riseup.net>, 48916 <at> debbugs.gnu.org
> Date: Tue, 08 Jun 2021 15:28:34 +0200
> 
> > This isn't formatted accoring to our rules.  In particular, each
> > sentence after the colon should begin with a capital letter and
> > end with a period.  See CONTRIBUTE for more details (and I suggest
> > to use Emacs commands for writing log messages, as they will take
> > care of some routine parts of the formatting for you).
> 
> So that means you are recommending to use VC instead of Magit 
> (that I used) for committing changes ?
> Maybe Magit have a something that I am not aware for this purpose 
> ?

AFAIK, Magit does have such a command, but don't ask me what it is...

> > Also, I believe we don't  like redundant :group tags, such as 
> > the one here.
> 
> got them. the last one may be a little bit confusing for the 
> reader that is not aware of it because all others defcustoms of 
> window.el applies a :group tag. Do you want that I remove the 
> others unnecessary ones ?

Fine with me if you remove them.

(The patch you sent was the old one, right?)

Thanks.

Message #20 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: Eli Zaretskii <eliz <at> gnu.org>
To: pillule <pillule <at> riseup.net>
Cc: 48916 <at> debbugs.gnu.org
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Tue, 08 Jun 2021 17:13:07 +0300

> From: pillule <pillule <at> riseup.net>
> Cc: pillule <pillule <at> riseup.net>, 48916 <at> debbugs.gnu.org
> Date: Tue, 08 Jun 2021 15:57:49 +0200
> 
> >> +subsequently all interactive windmove commandswill ignore the
> >                                          ^^^^^^^^^^^^
> > A typo.
> >
> 
> No, it seems to be a bug on my end. Somewhere on a file 
> transfer. Sorry I am not sure I will be able to fix that one 
> easily.

If that's the case, I suggest to send the patch as an attachment.

Message #23 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: pillule <pillule <at> riseup.net>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: pillule <pillule <at> riseup.net>, 48916 <at> debbugs.gnu.org
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Tue, 08 Jun 2021 16:26:51 +0200

[Message part 1 (text/plain, inline)]

Eli Zaretskii <eliz <at> gnu.org> writes:

>> got them. the last one may be a little bit confusing for the
>> reader that is not aware of it because all others defcustoms of
>> window.el applies a :group tag. Do you want that I remove the
>> others unnecessary ones ?
>
> Fine with me if you remove them.
>
> (The patch you sent was the old one, right?)
>> [...]
>
> If that's the case, I suggest to send the patch as an 
> attachment.

My bad the bug was me ;)

[0004-User-option-to-select-no-other-window-with-windmove.patch (text/x-diff, attachment)]

[Message part 3 (text/plain, inline)]

--

Message #26 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: Robert Pluim <rpluim <at> gmail.com>
To: pillule <pillule <at> riseup.net>
Cc: Eli Zaretskii <eliz <at> gnu.org>, 48916 <at> debbugs.gnu.org
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Tue, 08 Jun 2021 16:48:45 +0200

>>>>> On Tue, 08 Jun 2021 17:12:08 +0300, Eli Zaretskii <eliz <at> gnu.org> said:

    >> From: pillule <pillule <at> riseup.net>
    >> Cc: pillule <pillule <at> riseup.net>, 48916 <at> debbugs.gnu.org
    >> Date: Tue, 08 Jun 2021 15:28:34 +0200
    >> 
    >> > This isn't formatted accoring to our rules.  In particular, each
    >> > sentence after the colon should begin with a capital letter and
    >> > end with a period.  See CONTRIBUTE for more details (and I suggest
    >> > to use Emacs commands for writing log messages, as they will take
    >> > care of some routine parts of the formatting for you).
    >> 
    >> So that means you are recommending to use VC instead of Magit 
    >> (that I used) for committing changes ?

How you commit is up to you. This is about how to create the commit
message.

    >> Maybe Magit have a something that I am not aware for this purpose 
    >> ?

    Eli> AFAIK, Magit does have such a command, but don't ask me what it is...

'magit-generate-changelog', which you can run from your Magit commit
buffer. It's pretty much the same as C-x 4 A
(diff-add-change-log-entries-other-window), which you'd run from a
*vc-diff* buffer. The main difference is that C-x 4 A and C-x 4 a will
put entries into a ChangeLog file or buffer.

Robert
-- 

Message #29 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: Juri Linkov <juri <at> linkov.net>
To: pillule <pillule <at> riseup.net>
Cc: 48916 <at> debbugs.gnu.org
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Tue, 08 Jun 2021 19:46:23 +0300

>>> +(defcustom windmove-ignore-no-other-window nil
>>
>> This name is not the best one.  For starters, "ignore-no" is a kind of
>> double negation, which makes it harder to understand and remember.
>> Can you come up with a better description of what exactly is ignored
>> here?

> Yes, i think 'windmove-move-in-all-windows' will be more explicit.

Repeating words "...move-move..." doesn't look nice.  Maybe a better name
would be 'windmove-allow-all-windows'.

> Because it applies to `windmove-find-other-window', it allows the commands
> windmove-right, windmove-left, windmove-down, windmove-up to select windows
> with the no-other-window parameter, but not for the `windmove-display-*'
> functions. I suppose it is the desired behavior.

You could try using new option in all functions that call window-in-direction:
in windmove-display-in-direction, in windmove-delete-in-direction,
in windmove-swap-states-in-direction.

Message #34 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: pillule <pillule <at> riseup.net>
To: Robert Pluim <rpluim <at> gmail.com>
Cc: pillule <pillule <at> riseup.net>, Eli Zaretskii <eliz <at> gnu.org>,
 48916 <at> debbugs.gnu.org, Juri Linkov <juri <at> linkov.net>
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Wed, 09 Jun 2021 19:37:15 +0200

[Message part 1 (text/plain, inline)]

Robert Pluim <rpluim <at> gmail.com> writes:

>     >> Maybe Magit have a something that I am not aware for this 
>     >> purpose
>     >> ?
>
>     Eli> AFAIK, Magit does have such a command, but don't ask me 
>     what it is...
>
> 'magit-generate-changelog', which you can run from your Magit 
> commit
> buffer. It's pretty much the same as C-x 4 A
> (diff-add-change-log-entries-other-window), which you'd run from 
> a
> *vc-diff* buffer. The main difference is that C-x 4 A and C-x 4 
> a will
> put entries into a ChangeLog file or buffer.

Thanks you for the guidance, I think I would have had difficulties 
to find it
myself because it seems dependant of the context (at least the 
magit one)
and I couldn't find them directly with eg 'helpfull-function'.
However it don't seems to always works and another annoyance but 
it is
another topic.

Very much appreciated.

Juri Linkov <juri <at> linkov.net> writes:

>>>> +(defcustom windmove-ignore-no-other-window nil
>>>
>>> This name is not the best one.  For starters, "ignore-no" is a 
>>> kind of
>>> double negation, which makes it harder to understand and 
>>> remember.
>>> Can you come up with a better description of what exactly is 
>>> ignored
>>> here?
>
>> Yes, i think 'windmove-move-in-all-windows' will be more 
>> explicit.
>
> Repeating words "...move-move..." doesn't look nice.  Maybe a 
> better name
> would be 'windmove-allow-all-windows'.

That is better, indeed, thank you.

>> Because it applies to `windmove-find-other-window', it allows 
>> the commands
>> windmove-right, windmove-left, windmove-down, windmove-up to 
>> select windows
>> with the no-other-window parameter, but not for the 
>> `windmove-display-*'
>> functions. I suppose it is the desired behavior.
>
> You could try using new option in all functions that call 
> window-in-direction:
> in windmove-display-in-direction, in 
> windmove-delete-in-direction,
> in windmove-swap-states-in-direction.

Done.
I am in terra incognita for theses commands, The opportunity to 
add new
keybindings in my configuration is always nice :p

It was really simple and my tests seems to works fine with
'windmove-allow-all-windows' for all functions.

See the patch attached.

[0005-User-option-to-select-no-other-window-with-windmove.patch (text/x-diff, attachment)]

[Message part 3 (text/plain, inline)]

As a side note, while reading these commands,

I think 'windmove-display-same-window', 
'windmove-display-new-frame',
'windmove-display-new-tab' are not really at they own place in 
windmove.

We have already "C-x 4 1" (same-window-prefix) and "C-x 5 5"
(other-frame-prefix) and (other-tab-prefix) --that is not bound to 
any
key on my system but would eventually have its place in the "C-x 
t" map.

Should we --for these windmove-commands :
1. keep them as they are
2. make aliases and clean-up 'windmove-display-in-direction'
3. delete them and clean-up 'windmove-display-in-direction'
?

Since it becomes to get off the topic with the removal of the
:group tag, I can start another patch "clean-up" if requested.

Message #37 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: Juri Linkov <juri <at> linkov.net>
To: pillule <pillule <at> riseup.net>
Cc: Robert Pluim <rpluim <at> gmail.com>, 48916 <at> debbugs.gnu.org,
 Eli Zaretskii <eliz <at> gnu.org>
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Wed, 09 Jun 2021 23:13:04 +0300

>> You could try using new option in all functions that call
>> window-in-direction:
>> in windmove-display-in-direction, in windmove-delete-in-direction,
>> in windmove-swap-states-in-direction.
>
> Done.
> I am in terra incognita for theses commands, The opportunity to add new
> keybindings in my configuration is always nice :p
>
> It was really simple and my tests seems to works fine with
> 'windmove-allow-all-windows' for all functions.
>
> See the patch attached.

Thanks, now it's much better.

> As a side note, while reading these commands,
>
> I think 'windmove-display-same-window', 'windmove-display-new-frame',
> 'windmove-display-new-tab' are not really at they own place in windmove.
>
> We have already "C-x 4 1" (same-window-prefix) and "C-x 5 5"
> (other-frame-prefix) and (other-tab-prefix) --that is not bound to any
> key on my system but would eventually have its place in the "C-x t" map.

It's very strange that your configuration has unbound other-tab-prefix.
By default, it's bound to 'C-x t t'.

> Should we --for these windmove-commands :
> 1. keep them as they are
> 2. make aliases and clean-up 'windmove-display-in-direction'
> 3. delete them and clean-up 'windmove-display-in-direction'
> ?

Please keep them as they are.  It's very convenient
to bind these commands to use the same prefix keys
that are configured by windmove customization
(that was greatly improved recently!)

Message #40 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: pillule <pillule <at> riseup.net>
To: Juri Linkov <juri <at> linkov.net>
Cc: pillule <pillule <at> riseup.net>, Robert Pluim <rpluim <at> gmail.com>,
 48916 <at> debbugs.gnu.org, Eli Zaretskii <eliz <at> gnu.org>
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Thu, 10 Jun 2021 10:58:11 +0200

Juri Linkov <juri <at> linkov.net> writes:

>> Should we --for these windmove-commands :
>> 1. keep them as they are
>> 2. make aliases and clean-up 'windmove-display-in-direction'
>> 3. delete them and clean-up 'windmove-display-in-direction'
>> ?
>
> Please keep them as they are.  It's very convenient
> to bind these commands to use the same prefix keys
> that are configured by windmove customization
> (that was greatly improved recently!)

Ok fine.
--

Message #43 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: pillule <pillule <at> riseup.net>
To: pillule <pillule <at> riseup.net>
Cc: Robert Pluim <rpluim <at> gmail.com>, 48916 <at> debbugs.gnu.org,
 Eli Zaretskii <eliz <at> gnu.org>, Juri Linkov <juri <at> linkov.net>
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Tue, 15 Jun 2021 21:50:03 +0200

[Message part 1 (text/plain, inline)]

References: <878s3kzm66.fsf <at> riseup.net> <83czswfunl.fsf <at> gnu.org> <87zgw0xxya.fsf <at> riseup.net> <83y2bke98n.fsf <at> gnu.org> <878s3k76pe.fsf <at> gmail.com> <87mtryj28g.fsf <at> riseup.net> <87lf7ibxxf.fsf <at> mail.linkov.net> <87a6nyks9y.fsf <at> riseup.net>
User-agent: mu4e 1.5.11; emacs 28.0.50
In-reply-to: <87a6nyks9y.fsf <at> riseup.net>

pillule <pillule <at> riseup.net> writes:

> Juri Linkov <juri <at> linkov.net> writes:
>
>>> Should we --for these windmove-commands :
>>> 1. keep them as they are
>>> 2. make aliases and clean-up 'windmove-display-in-direction'
>>> 3. delete them and clean-up 'windmove-display-in-direction'
>>> ?
>>
>> Please keep them as they are.  It's very convenient
>> to bind these commands to use the same prefix keys
>> that are configured by windmove customization
>> (that was greatly improved recently!)
>
> Ok fine.

Hi, here the patch for this one  with v28.1 and an updated docstring.

[0003-User-option-to-select-no-other-window-with-windmove-.patch (text/x-diff, attachment)]

[Message part 3 (text/plain, inline)]

--

Message #46 received at 48916 <at> debbugs.gnu.org (full text, mbox):

From: Juri Linkov <juri <at> linkov.net>
To: pillule <pillule <at> riseup.net>
Cc: Robert Pluim <rpluim <at> gmail.com>, 48916 <at> debbugs.gnu.org,
 Eli Zaretskii <eliz <at> gnu.org>
Subject: Re: bug#48916: 28.0.50; allow windmove to select windows with the
 'no-other-window parameter
Date: Wed, 16 Jun 2021 02:27:29 +0300

tags 48916 fixed
close 48916 28.0.50
thanks

>> Ok fine.
>
> Hi, here the patch for this one  with v28.1 and an updated docstring.

Thanks for the patch, now it's pushed.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.