bug#14382: 24.3.50; Strange display bug

bug#14382: 24.3.50; Strange display bug

[Richard Stallman]

I visited the file free-sw.html, which I've edited.
I typed C-x v v as if to check in the changes.
It displayed *vc-log* in the bottom window.

In the *vc-log* buffer, I typed aaaa C-a C-o.
This made the buffer start with a blank line.

The bug is, that line is displayed in inverse video
just like the mode line above it.

I can reproduce the bug only in *vc-log*.
In *scratch*, in free-sw.html, in this mail buffer,
if I go to the beginning and type C-o, the bug does not occur.



In GNU Emacs 24.3.50.1 (mips64el-unknown-linux-gnu, GTK+ Version 2.20.1)
 of 2013-05-01 on chiefs-gnewsense
Bzr revision: 112434 juri@jurta.org-20130501081012-n3c351r92cr17lu5
System Description:	Debian GNU/Linux 6.0.6 (squeeze)

Configured using:
 `configure CFLAGS=-g -O0'

Important settings:
  value of $LANG: en_US.UTF-8
  locale-coding-system: utf-8-unix
  default enable-multibyte-characters: t

Major mode: Log-Edit

Minor modes in effect:
  gpm-mouse-mode: t
  tooltip-mode: t
  mouse-wheel-mode: t
  tool-bar-mode: t
  menu-bar-mode: t
  file-name-shadow-mode: t
  global-font-lock-mode: t
  font-lock-mode: t
  auto-composition-mode: t
  auto-encryption-mode: t
  auto-compression-mode: t
  line-number-mode: t
  transient-mark-mode: t

Recent input:
C-x C-f w w w / p h i l TAB f r e e - s w . h t m l 
RET C-x v v a a a a C-a C-o ESC x r e p o r t SPC e 
m a c s SPC b u g RET

Recent messages:
("emacs")
For information about GNU Emacs and the GNU system, type C-h C-a.
Mark set
Press C-c C-c when you are done editing.
Enter a change comment.  Type C-c C-c when done
Source file `/home/rms/emacs-bzr/trunk/lisp/gnus/message.el' newer than byte-compiled file

Load-path shadows:
None found.

Features:
(shadow sort nadvice gnus-util mail-extr emacsbug message format-spec
rfc822 mml mml-sec mm-decode mm-bodies mm-encode mail-parse rfc2231
mailabbrev gmm-utils mailheader sendmail rfc2047 rfc2045 ietf-drums
mm-util mail-prsvr mail-utils log-edit easymenu easy-mmode ring
pcvs-util add-log vc vc-dispatcher parse-time vc-cvs sgml-mode t-mouse
time-date tooltip ediff-hook vc-hooks lisp-float-type mwheel x-win x-dnd
tool-bar dnd fontset image regexp-opt fringe tabulated-list newcomment
lisp-mode register page menu-bar rfn-eshadow timer select scroll-bar
mouse jit-lock font-lock syntax facemenu font-core frame cham georgian
utf-8-lang misc-lang vietnamese tibetan thai tai-viet lao korean
japanese hebrew greek romanian slovak czech european ethiopic indian
cyrillic chinese case-table epa-hook jka-cmpr-hook help simple abbrev
minibuffer loaddefs button faces cus-face macroexp files text-properties
overlay sha1 md5 base64 format env code-pages mule custom widget
hashtable-print-readable backquote make-network-process dbusbind inotify
dynamic-setting system-font-setting font-render-setting move-toolbar gtk
x-toolkit x multi-tty emacs)

-- 
Dr Richard Stallman
President, Free Software Foundation
51 Franklin St
Boston MA 02110
USA
www.fsf.org  www.gnu.org
Skype: No way! That's nonfree (freedom-denying) software.
  Use Ekiga or an ordinary phone call

bug#14382: 24.3.50; Strange display bug

[Eli Zaretskii]

> Date: Fri, 10 May 2013 09:31:27 -0400
> From: Richard Stallman <rms@gnu.org>
> 
> I visited the file free-sw.html, which I've edited.
> I typed C-x v v as if to check in the changes.
> It displayed *vc-log* in the bottom window.
> 
> In the *vc-log* buffer, I typed aaaa C-a C-o.
> This made the buffer start with a blank line.
> 
> The bug is, that line is displayed in inverse video
> just like the mode line above it.

If you go to that line and type "M-x describe-text-properties RET",
does Emacs show that it deliberately put some face there that is
displayed as inverse video?

IOW, is the bug that this line uses some face that shouldn't be there,
or is the bug that the display engine displays the line in inverse
video even though it has a different (like default) face?

bug#14382: 24.3.50; Strange display bug

[Dmitry Gutov]

Richard Stallman <rms@gnu.org> writes:

> I visited the file free-sw.html, which I've edited.
> I typed C-x v v as if to check in the changes.
> It displayed *vc-log* in the bottom window.
>
> In the *vc-log* buffer, I typed aaaa C-a C-o.
> This made the buffer start with a blank line.
>
> The bug is, that line is displayed in inverse video
> just like the mode line above it.

Sounds like the header separator, look at `log-edit-font-lock-keywords',
last item.

Maybe we should add a check that the header is not empty, and only show
the separator in that case.

bug#14382: 24.3.50; Strange display bug

[Glenn Morris]

Dmitry Gutov wrote:

> Sounds like the header separator, look at `log-edit-font-lock-keywords',
> last item.

I mainly find this a visual annoyance.
Mail modes have been separating headers from bodies for ages.
Has any of them used such a separator?

> Maybe we should add a check that the header is not empty, and only show
> the separator in that case.

bug#14382: 24.3.50; Strange display bug

[Stefan Monnier]

> I visited the file free-sw.html, which I've edited.
> I typed C-x v v as if to check in the changes.
> It displayed *vc-log* in the bottom window.

Was this buffer empty?  If so, that's arguably the source of the problem.
It should normally start with a content along the lines of "Summary:\nAuthor:\n\n".

> The bug is, that line is displayed in inverse video
> just like the mode line above it.

The *vc-log* buffer nowadays uses an rfc822-style format, so the
reverse-video line you saw was the representation of the separator
between the header and the body.


        Stefan

bug#14382: 24.3.50; Strange display bug

[Stefan Monnier]

>> Sounds like the header separator, look at `log-edit-font-lock-keywords',
>> last item.
> I mainly find this a visual annoyance.
> Mail modes have been separating headers from bodies for ages.
> Has any of them used such a separator?

I'm not wedded to the current representation.

Especially since even in GUI frames I currently see a "full height"
thick line despite the (:height 0.1 :inverse-video t) face property.

If someone could look into this bug#9787 I'd appreciate it.

> Maybe we should add a check that the header is not empty, and only show
> the separator in that case.

We could, tho normally the header should never be empty.


        Stefan

bug#14382: 24.3.50; Strange display bug

[Richard Stallman]

I think you guessed it.  

    If you go to that line and type "M-x describe-text-properties RET",
    does Emacs show that it deliberately put some face there that is
    displayed as inverse video?

There are text properties here:
  face                 (:height 0.1 :inverse-video t)
  fontified            t

Maybe this is a feature.

-- 
Dr Richard Stallman
President, Free Software Foundation
51 Franklin St
Boston MA 02110
USA
www.fsf.org  www.gnu.org
Skype: No way! That's nonfree (freedom-denying) software.
  Use Ekiga or an ordinary phone call

bug#14382: 24.3.50; Strange display bug

[Dmitry Gutov]

Stefan Monnier <monnier@IRO.UMontreal.CA> writes:

>> Maybe we should add a check that the header is not empty, and only show
>> the separator in that case.
>
> We could, tho normally the header should never be empty.

It starts empty and a user may temporarily put an empty line at the
beginning of the buffer for whatever reason, like in this bug's
scenario.

Figuring out what's this black line for and how to go above it ("my
cursor disappeared!") is not effortless.

> It should normally start with a content along the lines of
> "Summary:\nAuthor:\n\n".

Huh?

bug#14382: 24.3.50; Strange display bug

[Stefan Monnier]

>>> Maybe we should add a check that the header is not empty, and only show
>>> the separator in that case.
>> We could, tho normally the header should never be empty.
> It starts empty

That's the problem that needs fixing.  Can someone check why it doesn't
start with "Summary:\nAuthor:\n\n".


        Stefan

bug#14382: 24.3.50; Strange display bug

[Dmitry Gutov]

On 11.05.2013 18:54, Stefan Monnier wrote:
>>>> Maybe we should add a check that the header is not empty, and only show
>>>> the separator in that case.
>>> We could, tho normally the header should never be empty.
>> It starts empty
>
> That's the problem that needs fixing.  Can someone check why it doesn't
> start with "Summary:\nAuthor:\n\n".

It doesn't seem to have ever done that.

`log-edit' inserts those lines when its second argument is t.

`vc-log-edit' passes it nil in the second argument (and has been doing 
that since before rev 87525).

Having the "Author" header present by default seems to encourage the 
user to leave it empty (if they're the author). `vc-git-checkin', for 
example, doesn't look like it's going to handle the empty value.

bug#14382: 24.3.50; Strange display bug

[Dmitry Gutov]

Dmitry Gutov <dgutov@yandex.ru> writes:
> `log-edit' inserts those lines when its second argument is t.

Ah, and also when `log-edit-setup-invert' is t, which is not the default value.

bug#14382: 24.3.50; Strange display bug

[Richard Stallman]

    That's the problem that needs fixing.  Can someone check why it doesn't
    start with "Summary:\nAuthor:\n\n".

I have never seen it start out as anything but empty.
What is "Summary:\nAuthor:\n\n" meant for, and why
should it appear in *vc-log* for a gnu.org web page?

-- 
Dr Richard Stallman
President, Free Software Foundation
51 Franklin St
Boston MA 02110
USA
www.fsf.org  www.gnu.org
Skype: No way! That's nonfree (freedom-denying) software.
  Use Ekiga or an ordinary phone call

bug#14382: 24.3.50; Strange display bug

[Stefan Monnier]

>> That's the problem that needs fixing.  Can someone check why it doesn't
>> start with "Summary:\nAuthor:\n\n".
> It doesn't seem to have ever done that.
> `log-edit' inserts those lines when its second argument is t.
> `vc-log-edit' passes it nil in the second argument (and has been doing that
> since before rev 87525).

Ah, I see.  We should change it, then to always insert such lines.

> Having the "Author" header present by default seems to encourage the user to
> leave it empty (if they're the author). `vc-git-checkin', for example,
> doesn't look like it's going to handle the empty value.

No, empty header lines get stripped before passing anything to the backend.
What the "Author:" does is encourage people to indicate who's the real
author of the code.


        Stefan

bug#14382: 24.3.50; Strange display bug

[Dmitry Gutov]

On 11.05.2013 21:41, Stefan Monnier wrote:
> What the "Author:" does is encourage people to indicate who's the real
> author of the code.

I don't like this approach. Most of the time, I'm both the author and 
the committer, so the empty line will just stick out like a sore thumb.

Not to mention that VC can also be used with projects other than Emacs, 
where accepting outside contributions as patch files can be a much rarer 
occurrence.

--Dmitry

bug#14382: 24.3.50; Strange display bug

[Richard Stallman]

    What the "Author:" does is encourage people to indicate who's the real
    author of the code.

If the idea is to propose a general format convention for VC change
log entries, it seems to me that we should have a broader discussion
before adopting it.  If the convention is not limited to Emacs, that
discussion should not be limited to Emacs either.

-- 
Dr Richard Stallman
President, Free Software Foundation
51 Franklin St
Boston MA 02110
USA
www.fsf.org  www.gnu.org
Skype: No way! That's nonfree (freedom-denying) software.
  Use Ekiga or an ordinary phone call

bug#14382: 24.3.50; Strange display bug

[Dmitry Gutov]

On 12.05.2013 1:44, Richard Stallman wrote:
>      What the "Author:" does is encourage people to indicate who's the real
>      author of the code.
>
> If the idea is to propose a general format convention for VC change
> log entries,

Not at all. The "Format" header value is translated into the first line 
of the commit message. The "Author" header value is translated into the 
"--author" argument passed to the VCS program.

So, having the "Author:" header present in a new log-edit-mode buffer is 
meant to encourage users to fill in its value, i.e. specify the actual 
author of the code they are checking in.

bug#14382: 24.3.50; Strange display bug

[Richard Stallman]

    Not at all. The "Format" header value is translated into the first line 
    of the commit message. The "Author" header value is translated into the 
    "--author" argument passed to the VCS program.

I start to understand now.

Is there a difference between "the first line of the commit message"
and "whatever I would previously have written on the first line
of *vc-log*?

-- 
Dr Richard Stallman
President, Free Software Foundation
51 Franklin St
Boston MA 02110
USA
www.fsf.org  www.gnu.org
Skype: No way! That's nonfree (freedom-denying) software.
  Use Ekiga or an ordinary phone call

bug#14382: 24.3.50; Strange display bug

[Dmitry Gutov]

On 12.05.2013 23:39, Richard Stallman wrote:
>      Not at all. The "Format" header value is translated into the first line
>      of the commit message. The "Author" header value is translated into the
>      "--author" argument passed to the VCS program.
>
> I start to understand now.
>
> Is there a difference between "the first line of the commit message"
> and "whatever I would previously have written on the first line
> of *vc-log*?

Probably not.

To be more clear, before the commit happens, the value of "Summary", if 
any, is inserted after the headers and, together with the text already 
present there, is used as commit message. See `log-edit-extract-headers'.

bug#14382: bug#15979: 24.3.50; Bad display in *vc-log* buffer

[Lars Ingebrigtsen]

Richard Stallman <rms@gnu.org> writes:

> I visited the file free-sw.html, which I've edited.
> I typed C-x v v as if to check in the changes.
> It displayed *vc-log* in the bottom window.
>
> In the *vc-log* buffer, I typed aaaa C-a C-o.
> This made the buffer start with a blank line.
>
> The bug is, that line is displayed in inverse video
> just like the mode line above it.

(I'm going through old bug reports that unfortunately weren't resolved
at the time.)

Skimming this bug report, it seems like this is working as designed, so
I'm closing this bug report.

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