[lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Richard Stallman]

Would someone please install this?

------- Start of forwarded message -------
Date: Sun, 25 Feb 2007 17:09:10 +0100
From: "Lennart Borgman (gmail)" <lennart.borgman@gmail.com>
MIME-Version: 1.0
To: emacs-pretest-bug@gnu.org
In-Reply-To: <45E190EF.7050104@gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Subject: Re: map-y-or-n does not use minibuffer-prompt face
X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=failed 
	version=3.0.4

Lennart Borgman (gmail) wrote:
> The function `map-y-or-n-p' does not use minibuffer-prompt face.
> 
> To show this start Emacs with
> 
>    emacs -Q
> 
> then do
> 
>    M-x customize-face RET minibuffer-prompt RET
> 
> Set Foreground to red and click "Set for Current Session". Just to check 
> that the prompt is colored do
> 
>    M-x
> 
> Cancel the prompt and then paste this code into the *Scratch* buffer and 
> eval it:
> 
>   (map-y-or-n-p
>    "Prompt should have minibuffer-prompt face, object=%s? "
>    (function
>     (lambda(object)
>       (message "object=%s" object)))
>    '("a" "b"))
> 
> The prompt will not be colored.


Here is a patch for this. I would be glad if someone installed it. As 
usual I do not want to touch for CVS at this time.

Index: map-ynp.el
===================================================================
RCS file: /cvsroot/emacs/emacs/lisp/emacs-lisp/map-ynp.el,v
retrieving revision 1.9
diff -u -r1.9 map-ynp.el
- --- map-ynp.el	21 Jan 2007 02:44:24 -0000	1.9
+++ map-ynp.el	25 Feb 2007 16:05:54 -0000
@@ -149,9 +149,14 @@
  		     ;; Prompt in the echo area.
  		     (let ((cursor-in-echo-area (not no-cursor-in-echo-area))
  			   (message-log-max nil))
- -		       (message "%s(y, n, !, ., q, %sor %s) "
- -				prompt user-keys
- -				(key-description (vector help-char)))
+                       (let ((prompt-with-face
+                              (format "%s(y, n, !, ., q, %sor %s) "
+                                      prompt user-keys
+                                      (key-description (vector 
help-char)))))
+                         (add-text-properties
+                          0 (length prompt-with-face)
+                          minibuffer-prompt-properties prompt-with-face)
+                         (message "%s" prompt-with-face))
  		       (if minibuffer-auto-raise
  			   (raise-frame (window-frame (minibuffer-window))))
  		       (while (progn


_______________________________________________
emacs-pretest-bug mailing list
emacs-pretest-bug@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-pretest-bug
------- End of forwarded message -------

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Kim F. Storm]

Richard Stallman <rms@gnu.org> writes:

> Would someone please install this?

I have installed a different patch.

Since then Lennart has pointed out that several other minibuffer
prompts also lack fontification, notably isearch, query-replace, and
y-or-n-p.

Shall we fix those too?

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Lennart Borgman (gmail)]

Kim F. Storm wrote:
> Richard Stallman <rms@gnu.org> writes:
> 
>> Would someone please install this?
> 
> I have installed a different patch.
> 
> Since then Lennart has pointed out that several other minibuffer
> prompts also lack fontification, notably isearch, query-replace, and
> y-or-n-p.
> 
> Shall we fix those too?


I think they are just simple bugs and I have sent some patches to Kim.

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Richard Stallman]

    Since then Lennart has pointed out that several other minibuffer
    prompts also lack fontification, notably isearch, query-replace, and
    y-or-n-p.

    Shall we fix those too?

Isearch, definitely.

However, now I have second thoughts about this for map-y-or-n.  The
main purpose of this face is to more clearly separate the prompt from
the input that follows it.  That purpose does not apply to y-or-n-p,
or to map-y-or-n.  However, to the extent that this face also helps
show the user that Emacs is asking some sort of question, it is useful
to use the face for other kinds of questions.

So, should we use it for y-or-n-p and map-y-or-n and query-replace,
or not?

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Lennart Borgman (gmail)]

Richard Stallman wrote:

> However, now I have second thoughts about this for map-y-or-n.  The
> main purpose of this face is to more clearly separate the prompt from
> the input that follows it.  That purpose does not apply to y-or-n-p,
> or to map-y-or-n.  However, to the extent that this face also helps
> show the user that Emacs is asking some sort of question, it is useful
> to use the face for other kinds of questions.
> 
> So, should we use it for y-or-n-p and map-y-or-n and query-replace,
> or not?


I think so because there is no other way to show that Emacs is asking 
some sort of question.

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Miles Bader]

Richard Stallman <rms@gnu.org> writes:
>     Since then Lennart has pointed out that several other minibuffer
>     prompts also lack fontification, notably isearch, query-replace, and
>     y-or-n-p.
>
>     Shall we fix those too?
>
> Isearch, definitely.

FWIW, isearch's prompt certainly _looks_ like it's using a face to me!

> So, should we use it for y-or-n-p and map-y-or-n and query-replace,
> or not?

Definitely we should, I think -- it just looks weird if some prompts
use the minibuffer face and others are do not.

-Miles
-- 
"Suppose we've chosen the wrong god. Every time we go to church we're
just making him madder and madder." -- Homer Simpson

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Kim F. Storm]

Richard Stallman <rms@gnu.org> writes:

>     Since then Lennart has pointed out that several other minibuffer
>     prompts also lack fontification, notably isearch, query-replace, and
>     y-or-n-p.
>
>     Shall we fix those too?
>
> Isearch, definitely.
>
> However, now I have second thoughts about this for map-y-or-n.  The
> main purpose of this face is to more clearly separate the prompt from
> the input that follows it.  That purpose does not apply to y-or-n-p,
> or to map-y-or-n.  However, to the extent that this face also helps
> show the user that Emacs is asking some sort of question, it is useful
> to use the face for other kinds of questions.
>
> So, should we use it for y-or-n-p and map-y-or-n and query-replace,
> or not?

To easier see the impact of this, I set the bold attribute on the
minibuffer-prompt face -- actually quite nice effect IMO.  It makes
the _questions_ in the minibuffer stand out from ordinary _messages_
in the echo area.

With this setting, it seems rather odd that SOME questions are
asked in the minibuffer-prompt face, and other are not.

E.g. doing M-% a RET b RET

Asks three questions:

1) Query replace:
2) Query replace a by:
3) Query replace a by b (? for help):

The first two are in minibuffer-prompt, the third is not,
so it does not stand out as a question.  IMO that's a bug.

Similar for y-or-n-p questions ...  they require user attention
so they should be in the minibuffer-prompt face.

Otherwise, we need some other face for such questions (e.g.
a `question' face.) which are for y-or-n-p and yes-and-no-p
It could by default be derived from minibuffer-prompt face.

But for now, I think minibuffer-prompt face is sufficient.

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

RE: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Drew Adams]

> > So, should we use it for y-or-n-p and map-y-or-n and query-replace,
> > or not?
>
> I think so because there is no other way to show that Emacs is asking
> some sort of question.

Those are blocking questions. You can't do anything but answer them. Sooner
or later, you will notice the question.

That is not an argument not to make the question more noticeable, of course;
it's just a response to the statement that you can't know that Emacs is
asking a question.

I would leave the face of minibuffer prompts and messages alone for now. Or
change the color only slightly (e.g. to medium blue), and use the face
consistently for all prompts. (Personally, I thought DarkBlue was fine, but
I don't use black for minibuffer text.)

After the release, we can always discuss whether there should be different
classes of prompts, with different faces, and so on.

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[David Kastrup]

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

>> > So, should we use it for y-or-n-p and map-y-or-n and query-replace,
>> > or not?
>>
>> I think so because there is no other way to show that Emacs is asking
>> some sort of question.
>
> Those are blocking questions. You can't do anything but answer them.

Of course you can, particularly with focus-follows-mouse or even
mouse-autoselect-window.

And if you do so, you retain a dormant and unhappy pending minibuffer.

> Sooner or later, you will notice the question.

By which time it might no longer apply...

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Lennart Borgman (gmail)]

Drew Adams wrote:
>>> So, should we use it for y-or-n-p and map-y-or-n and query-replace,
>>> or not?
>> I think so because there is no other way to show that Emacs is asking
>> some sort of question.
> 
> Those are blocking questions. You can't do anything but answer them. Sooner
> or later, you will notice the question.


Yes, as soon as you happen to type "y" or "n" you answer the questions. 
And rather later than sooner you will notice that you did. ;-)

I think the prompt colors should be the same. That will be the base for 
making them more noticeable.

Then as I said before I want them to be noticeable. The user should know 
that there Emacs is expecting some special input and that Emacs will do 
something special with that input.

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Mathias Dahl]

[-- Attachment #1: Type: text/plain, Size: 696 bytes --]

> Then as I said before I want them to be noticeable. The user should know
> that there Emacs is expecting some special input and that Emacs will do
> something special with that input.

I agree that some different colors might be more noticeable, but it
don't seem to be much more so, IMO. Check out the attached screenshots
for an example from emacs -q, where I customized the prompt face
manually on one of them.

What I believe catch my attention more than the change in color is the
blinking prompt. I think that if the whole minibuffer changed
background color, then it might work. Might. One have to play with
overlays (or The Gimp) to know for sure.

Hmm, how about those elephants... :)

[-- Attachment #2: blueonwhite.png --]
[-- Type: image/png, Size: 21804 bytes --]

[-- Attachment #3: yellowonred.png --]
[-- Type: image/png, Size: 21815 bytes --]

[-- Attachment #4: Type: text/plain, Size: 142 bytes --]

_______________________________________________
Emacs-devel mailing list
Emacs-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/emacs-devel

RE: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Drew Adams]

> I think that if the whole minibuffer changed
> background color, then it might work. Might.

That's what I do, with a standalone minibuffer frame. And the colors are
different for isearch and for recursive minibuffer levels. The changes are
slight, but noticeable.

With buffer-local background colors (after the release), we could do
something like this for a minibuffer window (not just a frame).

Changing background color this way should be optional, of course.

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Lennart Borgman (gmail)]

Mathias Dahl wrote:

> What I believe catch my attention more than the change in color is the
> blinking prompt. I think that if the whole minibuffer changed
> background color, then it might work. Might. One have to play with
> overlays (or The Gimp) to know for sure.


A little bit more difficult to implement. Needs some testing I think.


> Hmm, how about those elephants... :)


I heard a rumour they just left from the savannah. Do you know they hit 
harder at difficult changes? And that in the late pretest season the big 
mail elephants may even hit at bug fixes?

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Miles Bader]

Miles Bader <miles.bader@necel.com> writes:
>> Isearch, definitely.
>
> FWIW, isearch's prompt certainly _looks_ like it's using a face to me!
>
>>     Since then Lennart has pointed out that several other minibuffer
>>     prompts also lack fontification, notably isearch, query-replace, and
>>     y-or-n-p.

Ok, now I looked a bit closer, and it looks like some of those
prompts _already_ use the minibuffer-prompt face:

          isearch:  already uses minibuffer-prompt
    query-replace:  already uses minibuffer-prompt (except in loop; should)
         y-or-n-p:  already uses minibuffer-prompt
     map-y-or-n-p:  does not use minibuffer-prompt (but should)

-Miles

-- 
97% of everything is grunge

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Lennart Borgman (gmail)]

Miles Bader wrote:
> Miles Bader <miles.bader@necel.com> writes:
>>> Isearch, definitely.
>> FWIW, isearch's prompt certainly _looks_ like it's using a face to me!
>>
>>>     Since then Lennart has pointed out that several other minibuffer
>>>     prompts also lack fontification, notably isearch, query-replace, and
>>>     y-or-n-p.
> 
> Ok, now I looked a bit closer, and it looks like some of those
> prompts _already_ use the minibuffer-prompt face:
> 
>           isearch:  already uses minibuffer-prompt
>     query-replace:  already uses minibuffer-prompt (except in loop; should)
>          y-or-n-p:  already uses minibuffer-prompt
>      map-y-or-n-p:  does not use minibuffer-prompt (but should)


Sorry, but there is a misunderstanding. What you should look at is 
rather minibuffer-prompt-properties. That specifies which font to use. 
The default is to use minibuffer-prompt face.

I think those places above never got changed when 
minibuffer-prompt-properties were introduced.

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Miles Bader]

"Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:
>>           isearch:  already uses minibuffer-prompt
>>     query-replace:  already uses minibuffer-prompt (except in loop; should)
>>          y-or-n-p:  already uses minibuffer-prompt
>>      map-y-or-n-p:  does not use minibuffer-prompt (but should)
>
> Sorry, but there is a misunderstanding. What you should look at is
> rather minibuffer-prompt-properties. That specifies which font to
> use. The default is to use minibuffer-prompt face.

What exactly are you saying?  That you think those places should use
`minibuffer-prompt-properties' instead of using a hard-wired set of
prompt properties [e.g.  '(face minibuffer-prompt)]?

I'd agree with that, but it's hardly clear from this thread that you
mean that or not.

-Miles
-- 
x
y
Z!

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Lennart Borgman (gmail)]

Miles Bader wrote:
> "Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:
>>>           isearch:  already uses minibuffer-prompt
>>>     query-replace:  already uses minibuffer-prompt (except in loop; should)
>>>          y-or-n-p:  already uses minibuffer-prompt
>>>      map-y-or-n-p:  does not use minibuffer-prompt (but should)
>> Sorry, but there is a misunderstanding. What you should look at is
>> rather minibuffer-prompt-properties. That specifies which font to
>> use. The default is to use minibuffer-prompt face.
> 
> What exactly are you saying?  That you think those places should use
> `minibuffer-prompt-properties' instead of using a hard-wired set of
> prompt properties [e.g.  '(face minibuffer-prompt)]?
> 
> I'd agree with that, but it's hardly clear from this thread that you
> mean that or not.
> 
> -Miles


Sorry, I thought I just said that those places should use 
`minibuffer-prompt-properties'.

And I wonder what made my answer hard to understand. If you want to you 
may very well tell me, off list perhaps. (It is the second time I tried 
to say that it looks to me like minibuffer-prompt-properties is the 
variable that is meant to specify the face for prompt.)

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Richard Stallman]

    To easier see the impact of this, I set the bold attribute on the
    minibuffer-prompt face -- actually quite nice effect IMO.  It makes
    the _questions_ in the minibuffer stand out from ordinary _messages_
    in the echo area.

    With this setting, it seems rather odd that SOME questions are
    asked in the minibuffer-prompt face, and other are not.

That is a good argument.

Please fix the other questions to use minibuffer-prompt.

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Kim F. Storm]

"Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:

> Sorry, I thought I just said that those places should use
> minibuffer-prompt-properties'.

Maybe you just said that in the mails you sent to me without
copying the list...?

We should clearly fix the prompts which lack the minibuffer-prompt face
completely (I will fix query-replace).

But at this time I don't see a big need to fix this generally by
replacing all places that set (face minibuffer-prompt) explicitly
rather than using minibuffer-prompt-properties.

I know Lennart disagree, but WDOT?

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Lennart Borgman (gmail)]

Kim F. Storm wrote:
> "Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:
> 
>> Sorry, I thought I just said that those places should use
>> minibuffer-prompt-properties'.
> 
> Maybe you just said that in the mails you sent to me without
> copying the list...?

Maybe. Sorry.

> We should clearly fix the prompts which lack the minibuffer-prompt face
> completely (I will fix query-replace).
> 
> But at this time I don't see a big need to fix this generally by
> replacing all places that set (face minibuffer-prompt) explicitly
> rather than using minibuffer-prompt-properties.
> 
> I know Lennart disagree, but WDOT?

Yes, I disagree and I think I will repeat the arguments for clarity. The 
reason is that I want it to be simple to toggle the face used by the 
minibuffer prompts.

As I said in many other places we make it easy to toggle the face used 
by having a variable with the same name as the face. This variable is 
used when Emacs decides which face to use. By default it points to the 
face with the same name. Then the user can easily toggle between two 
different faces.

This is what I want the user to be able to do in this case too. There is 
however no variable named minibuffer-prompt. Instead 
minibuffer-prompt-properties should be used here. Currently in some 
places it is used to select the face, in other places not.

Kim, I think I sent you patches for all cases where I think 
minibuffer-prompt-properties should be used. Didn't I do that? (I am 
currently using those patches and they work well for me.)

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Kim F. Storm]

"Lennart Borgman (gmail)" <lennart.borgman@gmail.com> writes:

> Kim, I think I sent you patches for all cases where I think
> minibuffer-prompt-properties should be used. Didn't I do that? (I am
> currently using those patches and they work well for me.)

You sent me some patches, yes, and I'll review then again later today.

Whether they correspond to what you think, I don't know.  Sadly, it
seems M-x read-mind-remote is currently broken, and I doubt we can
find a trivial fix for it, that can be installed before the release.

-- 
Kim F. Storm <storm@cua.dk> http://www.cua.dk

Re: [lennart.borgman@gmail.com: Re: map-y-or-n does not use minibuffer-prompt face]

[Lennart Borgman (gmail)]

Kim F. Storm wrote:

> Whether they correspond to what you think, I don't know.  Sadly, it
> seems M-x read-mind-remote is currently broken, and I doubt we can
> find a trivial fix for it, that can be installed before the release.

Too bad. My inability to express myself clearly have made me severly 
dependent on it.