Toolbar mode on/off

Toolbar mode on/off

[David Kastrup]

Hi,

at the current point of time, enabling the toolbar and disabling it
again is not a noop, but rather causes shrinking.  That's ugly.  On
the other side, Emacs should not increase its frame size beyond the
given initial frame size.

So I think that Emacs should record its startup frame size in _pixels_
(or just record whether it started its geometry calculation with the
toolbar on or off).  Turning the toolbar on and off should then cause
the text size to have the largest size fitting inside of the original
startup frame size.

Of course, it would be nicer (and simpler altogether) if Emacs did not
change its frame pixel size at all when turning toolbar (or menubar or
whatever) on and off.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

RE: Toolbar mode on/off

[Drew Adams]

> at the current point of time, enabling the toolbar and disabling it
> again is not a noop, but rather causes shrinking.

Shrinking of what? The frame? The font?

In Emacs 22.1, I don't see shrinking of anything. The number of lines
displayed is reduced or increased - that's the only change I see.

> That's ugly.  On
> the other side, Emacs should not increase its frame size beyond the
> given initial frame size.
>
> So I think that Emacs should record its startup frame size in _pixels_
> (or just record whether it started its geometry calculation with the
> toolbar on or off).  Turning the toolbar on and off should then cause
> the text size to have the largest size fitting inside of the original
> startup frame size.
>
> Of course, it would be nicer (and simpler altogether) if Emacs did not
> change its frame pixel size at all when turning toolbar (or menubar or
> whatever) on and off.

Just to be sure I understand. Are you saying that turning the tool-bar on or
off should have the (side-)effect of changing the font size?

If so, my answer would be no, it should not. Better to change the frame size
(if necessary) than the font size, IMO. But I don't see either one change in
Emacs 22.1 (and that's good).

Re: Toolbar mode on/off

[David Kastrup]

"Drew Adams" <drew.adams@oracle.com> writes:

>> at the current point of time, enabling the toolbar and disabling it
>> again is not a noop, but rather causes shrinking.
>
> Shrinking of what? The frame? The font?

The frame.

> In Emacs 22.1, I don't see shrinking of anything. The number of
> lines displayed is reduced or increased - that's the only change I
> see.

If you say
M-x tool-bar-mode RET

4 times in a row, you end up with the same size frame as before?
Guess you are lucky then.

>> That's ugly.  On the other side, Emacs should not increase its
>> frame size beyond the given initial frame size.
>>
>> So I think that Emacs should record its startup frame size in _pixels_
>> (or just record whether it started its geometry calculation with the
>> toolbar on or off).  Turning the toolbar on and off should then cause
>> the text size to have the largest size fitting inside of the original
>> startup frame size.
>>
>> Of course, it would be nicer (and simpler altogether) if Emacs did not
>> change its frame pixel size at all when turning toolbar (or menubar or
>> whatever) on and off.
>
> Just to be sure I understand. Are you saying that turning the
> tool-bar on or off should have the (side-)effect of changing the
> font size?

Nonsense.

> If so, my answer would be no, it should not. Better to change the
> frame size (if necessary) than the font size, IMO. But I don't see
> either one change in Emacs 22.1 (and that's good).

You don't see the frame size change at all?  Then you must have a
particularly lucky relation between toolbar height and font size, or
your toolkit behaves different from mine (GTK+).

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: Toolbar mode on/off

[Jan Djärv]

David Kastrup skrev:
> Hi,
> 
> at the current point of time, enabling the toolbar and disabling it
> again is not a noop, but rather causes shrinking.  That's ugly.  On
> the other side, Emacs should not increase its frame size beyond the
> given initial frame size.
> 
> So I think that Emacs should record its startup frame size in _pixels_
> (or just record whether it started its geometry calculation with the
> toolbar on or off).  Turning the toolbar on and off should then cause
> the text size to have the largest size fitting inside of the original
> startup frame size.
> 
> Of course, it would be nicer (and simpler altogether) if Emacs did not
> change its frame pixel size at all when turning toolbar (or menubar or
> whatever) on and off.
> 

It is on my todo-list, but I'm still waiting for the unicode2 merger to happen 
first before committing large changes.

	Jan D.

RE: Toolbar mode on/off

[Drew Adams]

> > In Emacs 22.1, I don't see shrinking of anything. The number of
> > lines displayed is reduced or increased - that's the only change I
> > see.
>
> If you say
> M-x tool-bar-mode RET
>
> 4 times in a row, you end up with the same size frame as before?
> Guess you are lucky then.

Nope. Absolutely no change in frame size, including with emacs -Q. But as I
say, I'm using Emacs 22.1, the released version. Perhaps you are using a
more recent version, in which a bug has been introduced?

> > Just to be sure I understand. Are you saying that turning the
> > tool-bar on or off should have the (side-)effect of changing the
> > font size?
>
> Nonsense.

What's nonsense? That was a question. You seem to be saying that you do see
such a frame size change, but that you agree that it is not desirable. Is
that right?

> > If so, my answer would be no, it should not. Better to change the
> > frame size (if necessary) than the font size, IMO. But I don't see
> > either one change in Emacs 22.1 (and that's good).
>
> You don't see the frame size change at all?  Then you must have a
> particularly lucky relation between toolbar height and font size, or
> your toolkit behaves different from mine (GTK+).

No, I do not see any frame size change, regardless of what font or font size
I use. I'm using 22.1 on MS Windows.

Re: Toolbar mode on/off

[Robert J. Chassell]

with today's GNU Emacs CVS snapshot, Tue, 2007 Sep 25   9:54 UTC
GNU Emacs 23.0.50.22 (i686-pc-linux-gnu, GTK+ Version 2.10.13)
started with

/usr/local/src/emacs/src/emacs -q --no-site-file \
     --eval '(setq-default mode-line-buffer-identification
                           (quote (#("%14b" 0 4 (face (:weight normal))))))' \
     -fn "-Misc-Fixed-Medium-R-Normal--20-200-75-75-C-100-ISO8859-1"   \
     --visit=/usr/local/bin/emacs-test-little-q                        \
     --eval "(setq frame-title-format '(\"Emacs test with -q:  %b\"))" \
     --eval '(set-frame-name "Emacs test with -q")'


"Drew Adams" <drew.adams@oracle.com> writes:

    >> at the current point of time, enabling the toolbar and disabling it
    >> again is not a noop, but rather causes shrinking.

That is true.  On an instance with a tool bar, M-x tool-bar-mode RET
makes the frame smaller and when tool-bar-mode is reenabled, larger.
The command does not change the font size.

-- 
    Robert J. Chassell                          GnuPG Key ID: 004B4AC8
    bob@rattlesnake.com                         bob@gnu.org
    http://www.rattlesnake.com                  http://www.teak.cc

RE: Toolbar mode on/off

[Drew Adams]

> "Drew Adams" <drew.adams@oracle.com> writes:
>
>     >> at the current point of time, enabling the toolbar and disabling it
>     >> again is not a noop, but rather causes shrinking.
>
> That is true.  On an instance with a tool bar, M-x tool-bar-mode RET
> makes the frame smaller and when tool-bar-mode is reenabled, larger.
> The command does not change the font size.

That looks very authoritative: `"Drew Adams" <drew.adams@oracle.com>
writes:', but it's just plain wrong.

Please, there is no sense citing a person and quoting something that s?he
didn't say. It's inaccurate, confusing, and unethical.

I did not say what you have quoted above. My only contribution to this
thread was to state, in fact, the opposite, that I do _not_ see, in Emacs
22.1 on Windows, any shrinking.

Please don't bother citing someone at all, or at least get the person right.

Re: Toolbar mode on/off

[Robert J. Chassell]

    That looks very authoritative: `"Drew Adams" <drew.adams@oracle.com>
    writes:', but it's just plain wrong.

My apologies.  I thought you had written what I cited.  My
understanding is traditional.  I don't know about angle brackets.

My only information about citations that use them comes from a very
short look at emacs/lisp/mail/supercite.el where
`sc-citation-delimiter' is a variable set in your .emacs that could be
several angle brackets in a row or anything else.

I will try to avoid mis-attributing. 

Thank you.

(I wrote, 

    with today's GNU Emacs CVS snapshot, Tue, 2007 Sep 25   9:54 UTC
    GNU Emacs 23.0.50.22 (i686-pc-linux-gnu, GTK+ Version 2.10.13) 
    started with [-q --no-site-file] ...
    On an instance with a tool bar, M-x tool-bar-mode RET
    makes the frame smaller ...
    The command does not change the font size.

(Drew Adams said the opposite,

    I do _not_ see, in Emacs 22.1 on Windows, any shrinking.
)

-- 
    Robert J. Chassell                          GnuPG Key ID: 004B4AC8
    bob@rattlesnake.com                         bob@gnu.org
    http://www.rattlesnake.com                  http://www.teak.cc

RE: Toolbar mode on/off

[Drew Adams]

>     That looks very authoritative: `"Drew Adams" <drew.adams@oracle.com>
>     writes:', but it's just plain wrong.
>
> My apologies.  I thought you had written what I cited.  My
> understanding is traditional.  I don't know about angle brackets.
>
> My only information about citations that use them comes from a very
> short look at emacs/lisp/mail/supercite.el where
> `sc-citation-delimiter' is a variable set in your .emacs that could be
> several angle brackets in a row or anything else.
>
> I will try to avoid mis-attributing.
>
> Thank you.

Thanks. It's an easy mistake to make; we all make it occasionally. I've
given up citing people here, because I think it just makes following threads
more confusing.

Re: Toolbar mode on/off

[Richard Stallman]

    at the current point of time, enabling the toolbar and disabling it
    again is not a noop, but rather causes shrinking.

That seems like bug, if it happens.

      That's ugly.  On
    the other side, Emacs should not increase its frame size beyond the
    given initial frame size.

I don't agree.  If the initial frame size was specified in lines,
and the toolbar was off, turning it on should increase the frame size.

    Of course, it would be nicer (and simpler altogether) if Emacs did not
    change its frame pixel size at all when turning toolbar (or menubar or
    whatever) on and off.

That would be incorrect if the size of the text area has been
specified in lines.

Re: Toolbar mode on/off

[David Kastrup]

Richard Stallman <rms@gnu.org> writes:

>     at the current point of time, enabling the toolbar and disabling it
>     again is not a noop, but rather causes shrinking.
>
> That seems like bug, if it happens.
>
>       That's ugly.  On
>     the other side, Emacs should not increase its frame size beyond the
>     given initial frame size.
>
> I don't agree.  If the initial frame size was specified in lines,
> and the toolbar was off, turning it on should increase the frame
> size.

A disabled toolbar comes about only by user configuration of the
startup, and this configuration is likely to also specify a window
size filling the screen vertically.  So later enabling a toolbar
without adjusting the number of lines will cause the bottom of the
frame to go off-screen: not really useful behavior whether or not one
considers it "correct".

>     Of course, it would be nicer (and simpler altogether) if Emacs did not
>     change its frame pixel size at all when turning toolbar (or menubar or
>     whatever) on and off.
>
> That would be incorrect if the size of the text area has been
> specified in lines.

The startup configuration will usually specify both lines and the
presence and absence of the toolbar, and implicitly the frame size.
The question is whether it is more important to the user to keep the
frame size or the line count when toggling toolbar mode.

It is my conviction that in the overwhelming number of cases, the
frame size will be more important to the user, since it will be almost
always be chosen to fit the available screen area.

-- 
David Kastrup

RE: Toolbar mode on/off

[Drew Adams]

>     at the current point of time, enabling the toolbar and disabling it
>     again is not a noop, but rather causes shrinking.
>
> That seems like bug, if it happens.

I agree.

>     That's ugly.  On the other side, Emacs should not increase
>     its frame size beyond the given initial frame size.
>
> I don't agree.  If the initial frame size was specified in lines,
> and the toolbar was off, turning it on should increase the frame size.

I don't think it should increase its size, and I don't think it should have
anything to do with the _initial_ frame.

The behavior I see in Emacs 22.1 (on Windows) is correct: there is no frame
resizing and no font resizing when you show/hide the tool bar. And the
initial frame has nothing to do with it. When I show/hide the tool bar, the
number of displayed lines decreases/increases. That's all.

Re: Toolbar mode on/off

[Richard Stallman]

A clarification, in case what I said before was not clear.

Disabling the toolbar should reduce the frame height, assuming you are
using an X toolkit.  And enabling the toolbar should increase the
frame height by the same amount.  Thus, repeatedly toggling an even
number of times should not alter the height.

RE: Toolbar mode on/off

[Davis Herring]

>> "Drew Adams" <drew.adams@oracle.com> writes:
>>
>>   >> at the current point of time, enabling the toolbar and disabling it
>>   >> again is not a noop, but rather causes shrinking.
>>
>> That is true.  On an instance with a tool bar, M-x tool-bar-mode RET
>> makes the frame smaller and when tool-bar-mode is reenabled, larger.
>> The command does not change the font size.
>
> That looks very authoritative: `"Drew Adams" <drew.adams@oracle.com>
> writes:', but it's just plain wrong.

It's not actually wrong: at the very top of the message available at
http://lists.gnu.org/archive/html/emacs-devel/2007-09/msg02264.html, you
did in fact write the following two lines without what is now their
left-most 3 angle brackets.  (I have adjusted the spacing to preserve the
line break positions.)

Anyone who understands quoting conventions at all knows that a header like
"X writes:" means that what follows the header comes from X if it begins
with a > and that the (one!) > is not part of the quoted text.  In fact,
if we are all following the reply-below idea, the first line of nearly
every message will begin with a >, and so when quoted will have two and
will look like what you didn't like.

Of course, replying to your message and then omitting everything you typed
and replying to what David actually said is odd, and perhaps even (as you
later said) confusing, but it's not (as you later stated) inaccurate or
unethical.

Davis

-- 
This product is sold by volume, not by mass.  If it appears too dense or
too sparse, it is because mass-energy conversion has occurred during
shipping.