Re: improving function "mode-line-unbury-buffer" in bindings.el

Re: improving function "mode-line-unbury-buffer" in bindings.el

[Richard Stallman]

Could you explain what problem this change is meant to fix?
How is the current behavior of mode-line-unbury-buffer less than ideal?
I cannot figure that out from the code that you sent.

Re: improving function "mode-line-unbury-buffer" in bindings.el

[Toru Tsuneyoshi]

I'm sorry that the explanation may have been insufficient.

 Somehow, "bury-buffer <f>" doesn't display a buffer which is not
displayed still once, when there are several buffers which is displayed
before.  It is explaned in example 3 of spec of function "bury-buffer"
called by "mode-line-bury-buffer" in file "report.txt" which was
attached before. When there are buffers "A" (now displayed), "*scratch*"
(displayed before), and "*Messages*" (not displayed still once), this
function never displays a buffer "*Messages*".
 But, this function has displayed a buffer which is not displayed still
once, when there is only another buffer which is displayed.  For
example, when there are buffers "*scratch*" (now displayed) and
"*Messages*" (not displayed still once), this function has displayed
a buffer "*Messages*".

 While, current mode-line-unbury-buffer has displayed a buffer (or
buffers) which is (or are) not displayed still once, at all times.

 I felt the non-symmetry nature about the behavior of current these two
functions. Do you also feel that? I avoid that by my new code.


From: Richard Stallman <rms@gnu.org>
Subject: Re: improving function "mode-line-unbury-buffer" in bindings.el
Date: Sun, 14 Jul 2002 09:22:27 -0600 (MDT)
Message-ID: <200207141522.g6EFMRW20484@aztec.santafe.edu>

> Could you explain what problem this change is meant to fix?
> How is the current behavior of mode-line-unbury-buffer less than ideal?
> I cannot figure that out from the code that you sent.
> 
> 

Re: improving function "mode-line-unbury-buffer" in bindings.el

[Richard Stallman]

     Somehow, "bury-buffer <f>" doesn't display a buffer which is not
    displayed still once, when there are several buffers which is displayed
    before.

I think you have probably found a real bug, i'm having a lot of
trouble parsing your explanation.  Do you think you could explain the
problem to handa@etl.go.jp in Japanese?  Then he can tell us what's
wrong.

Re: improving function "mode-line-unbury-buffer" in bindings.el

[Kim F. Storm]

Richard Stallman <rms@gnu.org> writes:

>      Somehow, "bury-buffer <f>" doesn't display a buffer which is not
>     displayed still once, when there are several buffers which is displayed
>     before.
> 
> I think you have probably found a real bug, i'm having a lot of
> trouble parsing your explanation.  Do you think you could explain the
> problem to handa@etl.go.jp in Japanese?  Then he can tell us what's
> wrong.

As I read the bug report, it's quite simple:

Here is what is CLAIMED to happen with bury-buffer:

When you use bury-buffer to bury the buffer in the selected window,
the NEXT buffer to be shown in the selected window is chosen with a
preference to show a buffer which has previously been shown in a
window.

So if you have buffers xyz (in the selected window), *scratch* (shown
before), and *Messages* (not shown before), then:

- doing bury-buffer once will bury xyz and show *scratch*, and
- doing bury-buffer again will bury *scratch* and show xyz.

I.e. *Messages* is not shown as a result of bury-buffer since it has
not been shown before.

The CLAIM is then that (mode-line-)unbury-buffer should not choose to
show *Messages* either, but it may do that.


However, using CVS emacs, I cannot even get bury-buffer to work as
reported, so I'm still not sure what the exact problem is.



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

Re: improving function "mode-line-unbury-buffer" in bindings.el

[Richard Stallman]

    Here is what is CLAIMED to happen with bury-buffer:

    When you use bury-buffer to bury the buffer in the selected window,
    the NEXT buffer to be shown in the selected window is chosen with a
    preference to show a buffer which has previously been shown in a
    window.

What I see in the code of Fother_buffer is that it prefers the buffers
that have been shown in this frame before.  However, burying a buffer
removes it from the frame's buffer list (this is a recent bug fix).
So as you cycle in this way, the frame's buffer list will empty out
and you will indeed get all the non-hidden buffers.

    - doing bury-buffer once will bury xyz and show *scratch*, and
    - doing bury-buffer again will bury *scratch* and show xyz.

I get *Messages* the third time now.

Re: improving function "mode-line-unbury-buffer" in bindings.el

[Kim F. Storm]

Richard Stallman <rms@gnu.org> writes:

>     Here is what is CLAIMED to happen with bury-buffer:
> 
>     When you use bury-buffer to bury the buffer in the selected window,
>     the NEXT buffer to be shown in the selected window is chosen with a
>     preference to show a buffer which has previously been shown in a
>     window.
> 
> What I see in the code of Fother_buffer is that it prefers the buffers
> that have been shown in this frame before.  However, burying a buffer
> removes it from the frame's buffer list (this is a recent bug fix).
> So as you cycle in this way, the frame's buffer list will empty out
> and you will indeed get all the non-hidden buffers.
> 
>     - doing bury-buffer once will bury xyz and show *scratch*, and
>     - doing bury-buffer again will bury *scratch* and show xyz.
> 
> I get *Messages* the third time now.

So do I (which is why I said "CLAIMED to happen" in my explanation).

So with the recent change to bury-buffer, unbury-buffer now does the
right thing I guess.

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

Re: improving function "mode-line-unbury-buffer" in bindings.el

[Toru Tsuneyoshi]

From: Richard Stallman <rms@gnu.org>
Subject: Re: improving function "mode-line-unbury-buffer" in bindings.el
Date: Thu, 18 Jul 2002 08:56:05 -0600 (MDT)
Message-ID: <200207181456.g6IEu5W25120@aztec.santafe.edu>

> What I see in the code of Fother_buffer is that it prefers the buffers
> that have been shown in this frame before.  However, burying a buffer
> removes it from the frame's buffer list (this is a recent bug fix).

Do you use emacs of which version?
(I use emacs-21.2 downloaded from ftp://ftp.gnu.org/gnu/emacs/.)
Will you tell me how to get emacs of newer vesion or it's reference
site?

> So as you cycle in this way, the frame's buffer list will empty out
> and you will indeed get all the non-hidden buffers.
> 
>     - doing bury-buffer once will bury xyz and show *scratch*, and
>     - doing bury-buffer again will bury *scratch* and show xyz.
> 
> I get *Messages* the third time now.
> 

By the way, for example, if you have only one buffer *Messages*, bury-buffer
makes a buffer *scratch*. How about unbury-buffer?

Re: improving function "mode-line-unbury-buffer" in bindings.el

[Richard Stallman]

    > What I see in the code of Fother_buffer is that it prefers the buffers
    > that have been shown in this frame before.  However, burying a buffer
    > removes it from the frame's buffer list (this is a recent bug fix).

    Do you use emacs of which version?

I am using the latest development sources, which is where that fix
was made.

    Will you tell me how to get emacs of newer vesion or it's reference
    site?

You can get it through CVS from savannah.gnu.org, project "Emacs".

    By the way, for example, if you have only one buffer *Messages*, bury-buffer
    makes a buffer *scratch*. How about unbury-buffer?

Thanks.  I made it handle that.