asterisks disappear

asterisks disappear

[Katsumi Yamaoka]

Hi,

`gnus-treat-emphasize' is 50000 by default and Gnus users cannot
see asterisks which surround words.  Although I know what a bold
text means, all users aren't so.  I am perplexed when telling
someone "Try it on the *scratch* buffer".  WDYT?

Re: asterisks disappear

[Reiner Steib]

On Fri, Nov 05 2004, Katsumi Yamaoka wrote:

> `gnus-treat-emphasize' is 50000 by default and Gnus users cannot
> see asterisks which surround words.  Although I know what a bold
> text means, all users aren't so.  I am perplexed when telling
> someone "Try it on the *scratch* buffer".  WDYT?

I agree.  How do other clients that treat emphasis character handle
this (e.g. Mozilla, Thunderbird, ...)?

I have used the following for quite some time:

;; keep emphasis characters around after emphasis
(require 'gnus-art)
(setq gnus-emphasis-alist ; <u8zbs8p6iue.fsf@blackbird-2k.MITRE.ORG>
      (mapcar (lambda (x)
		(list (car x) (nth 2 x) (nth 2 x) (nth 3 x)))
	      gnus-emphasis-alist))
;; don't do strikethru
(setq gnus-emphasis-alist ; <m3u1lu95u7.fsf@ID-87814.user.dfncis.de>
      (delq nil (mapcar
		 (lambda (x)
		   (if (eq (car (last x)) 'gnus-emphasis-strikethru) nil x))
		 gnus-emphasis-alist)))

(strikethru is already switched off by default now, IIRC)

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: asterisks disappear

[Katsumi Yamaoka]

>>>>> In <v9zn1w36uj.fsf@marauder.physik.uni-ulm.de> Reiner Steib wrote:

> On Fri, Nov 05 2004, Katsumi Yamaoka wrote:

>> `gnus-treat-emphasize' is 50000 by default and Gnus users cannot
>> see asterisks which surround words.  Although I know what a bold
>> text means, all users aren't so.  I am perplexed when telling
>> someone "Try it on the *scratch* buffer".  WDYT?

> I agree.  How do other clients that treat emphasis character handle
> this (e.g. Mozilla, Thunderbird, ...)?

Oh, Mozilla doesn't but Netscape 7.1 emphasize it as a bold text
without hiding asterisks.  It may be better that Gnus follows it.

> I have used the following for quite some time:

> ;; keep emphasis characters around after emphasis
> (require 'gnus-art)
> (setq gnus-emphasis-alist ; <u8zbs8p6iue.fsf@blackbird-2k.MITRE.ORG>
>       (mapcar (lambda (x)
> 		(list (car x) (nth 2 x) (nth 2 x) (nth 3 x)))
> 	      gnus-emphasis-alist))
> ;; don't do strikethru
> (setq gnus-emphasis-alist ; <m3u1lu95u7.fsf@ID-87814.user.dfncis.de>
>       (delq nil (mapcar
> 		 (lambda (x)
> 		   (if (eq (car (last x)) 'gnus-emphasis-strikethru) nil x))
> 		 gnus-emphasis-alist)))

> (strikethru is already switched off by default now, IIRC)

Thanks.  It does just what I wrote.  However, it will be hard to
spread this to all users, especially newcomers.

Re: asterisks disappear

[Katsumi Yamaoka]

>>>>> In <b9yvfcklczj.fsf@jpl.org> Katsumi Yamaoka wrote:

> Oh, Mozilla doesn't but Netscape 7.1 emphasize it as a bold text
> without hiding asterisks.  It may be better that Gnus follows it.

I mistook.  It was an act of Gmane, not Netscape. ;-)

http://article.gmane.org/gmane.emacs.gnus.general/59051

Re: asterisks disappear

[Daniel Pittman]

On 5 Nov 2004, Reiner Steib wrote:
> On Fri, Nov 05 2004, Katsumi Yamaoka wrote:
>
>> `gnus-treat-emphasize' is 50000 by default and Gnus users cannot
>> see asterisks which surround words.  Although I know what a bold
>> text means, all users aren't so.  I am perplexed when telling
>> someone "Try it on the *scratch* buffer".  WDYT?
>
> I agree.  How do other clients that treat emphasis character handle
> this (e.g. Mozilla, Thunderbird, ...)?

They show the characters that caused the emphasis as well as the style
change.  This is, I believe, vastly superior behavior.

I, also, have modified this forever and like it much better.

   Daniel
-- 
I am for an art of teddy bears and guns and decapitated rabbits,
exploded umbrellas, raped beds, chairs with their brown bones broken,
burning trees, firecracker ends, chicken bones, pigeon bones and boxes
with men sleeping in them.
        -- Claes Oldenburg

Re: asterisks disappear

[Miles Bader]

Daniel Pittman <daniel@rimspace.net> writes:
> They show the characters that caused the emphasis as well as the style
> change.  This is, I believe, vastly superior behavior.
>
> I, also, have modified this forever and like it much better.

FWIW, I absolutely hate the "emphasize-but-don't-hide" behavior -- it
just looks sucky (more sucky with underlines that with bold/asterisks
though).

-Miles
-- 
=====
(^o^;
(()))
*This is the cute octopus virus, please copy it into your sig so it can spread.

Re: asterisks disappear

[Katsumi Yamaoka]

>>>>> In <874qk2dix8.fsf@tc-1-100.kawasaki.gol.ne.jp> 
>>>>>	Miles Bader <miles@gnu.org> wrote:

> FWIW, I absolutely hate the "emphasize-but-don't-hide" behavior -- it
> just looks sucky (more sucky with underlines that with bold/asterisks
> though).

The present default behavior is very good if all asterisks,
underlines and slashes which surround words are used to intend
to emphasize them.  I raised the problem on which Gnus users may
misunderstand buffer names, e.g., "*Messages*", "*scratch*", etc.

Re: asterisks disappear

[Miles Bader]

Katsumi Yamaoka <yamaoka@jpl.org> writes:
>> FWIW, I absolutely hate the "emphasize-but-don't-hide" behavior -- it
>> just looks sucky (more sucky with underlines that with bold/asterisks
>> though).
>
> The present default behavior is very good if all asterisks,
> underlines and slashes which surround words are used to intend
> to emphasize them.  I raised the problem on which Gnus users may
> misunderstand buffer names, e.g., "*Messages*", "*scratch*", etc.

Yes, I fully understand this.  Reading messages about Emacs with missing
asterisks is annoying -- but that represents only something like 1% of
the messages I read.

An easy-to-use per-group customization that would cause _only_ asterisks
to be retained might be a good thing (underlines are both more common,
and more adversely affected by the "retaining" behavior).  You might
even try to have this default to "true" on common Emacs-related
newsgroups.

-Miles
-- 
"1971 pickup truck; will trade for guns"

Re: asterisks disappear

[Katsumi Yamaoka]

>>>>> In <87vfchcmrj.fsf@tc-1-100.kawasaki.gol.ne.jp> Miles Bader wrote:

> Katsumi Yamaoka <yamaoka@jpl.org> writes:

>> I raised the problem on which Gnus users may
>> misunderstand buffer names, e.g., "*Messages*", "*scratch*", etc.

> Yes, I fully understand this.  Reading messages about Emacs with missing
> asterisks is annoying -- but that represents only something like 1% of
> the messages I read.

I have many opportunities to write buffer names in gnu.emacs.gnus
and so forth.  In the latest example, the buffer name "*trace of
SMTP session to XXX*" might not have been understood correctly by
a recipient.

> An easy-to-use per-group customization that would cause _only_ asterisks
> to be retained might be a good thing (underlines are both more common,
> and more adversely affected by the "retaining" behavior).  You might
> even try to have this default to "true" on common Emacs-related
> newsgroups.

It is not my problem.  People who began to use Emacs and Gnus
will not know it.  In case I reply to their question, I might be
unable to tell buffer names correctly.  For instance, "Haven't
any warning messages remained in the *Messages* buffer?".

Therefore, I think it is better to display at least asterisks by
default.  I will probably make them invisible, though.

Re: asterisks disappear

[Ted Zlatanov]

On Mon, 08 Nov 2004, yamaoka@jpl.org wrote:

>>>>> In <87vfchcmrj.fsf@tc-1-100.kawasaki.gol.ne.jp> Miles Bader wrote:
> 
>> Katsumi Yamaoka <yamaoka@jpl.org> writes:
> 
>>> I raised the problem on which Gnus users may
>>> misunderstand buffer names, e.g., "*Messages*", "*scratch*", etc.
> 
>> Yes, I fully understand this.  Reading messages about Emacs with missing
>> asterisks is annoying -- but that represents only something like 1% of
>> the messages I read.
> 
> I have many opportunities to write buffer names in gnu.emacs.gnus
> and so forth.  In the latest example, the buffer name "*trace of
> SMTP session to XXX*" might not have been understood correctly by
> a recipient.

How about turning asterisk highlight off if the word "buffer" is in
the message?

Ted

Re: asterisks disappear

[Katsumi Yamaoka]

>>>>> In <4nis8gfvr7.fsf@lifelogs.com> Ted Zlatanov wrote:

> How about turning asterisk highlight off if the word "buffer" is in
> the message?

That's a bright idea!  I need to try it on later.  Thanks.

Re: asterisks disappear

[Miles Bader]

Katsumi Yamaoka <yamaoka@jpl.org> writes:
>> How about turning asterisk highlight off if the word "buffer" is in
>> the message?
>
> That's a bright idea!  I need to try it on later.  Thanks.

I think it cause a lot of confusion if turned on by default, though, as
it would likely miss a lot of cases, and maybe make the astericks appear
to be hidden randomly!

Maybe it's best to make underlines default to "hidden" and astericks
default to "shown" and simply make sure that's fairly convenient to
change these defaults, especially per group (for my personal usage, it's
probably sufficient to simply makes astericks shown in any group whose
name matchs ".*emacs.*", and hidden otherwise).

-Miles
-- 
`Life is a boundless sea of bitterness'

Re: asterisks disappear

[Katsumi Yamaoka]

>>>>> In <4nis8gfvr7.fsf@lifelogs.com> Ted Zlatanov wrote:

>>> How about turning asterisk highlight off if the word "buffer" is in
>>> the message?

> Katsumi Yamaoka <yamaoka@jpl.org> writes:

>> That's a bright idea!  I need to try it on later.  Thanks.

>>>>> In <87zn1sldkv.fsf@tc-1-100.kawasaki.gol.ne.jp>
>>>>>	Miles Bader <miles@gnu.org> wrote:

> I think it cause a lot of confusion if turned on by default, though, as
> it would likely miss a lot of cases, and maybe make the astericks appear
> to be hidden randomly!

Yes, it may be so.

> Maybe it's best to make underlines default to "hidden" and astericks
> default to "shown" and simply make sure that's fairly convenient to
> change these defaults, especially per group (for my personal usage, it's
> probably sufficient to simply makes astericks shown in any group whose
> name matchs ".*emacs.*", and hidden otherwise).

I still think it is better to show all text by default.  It
doesn't mean abolishing emphasis but, for example:

* scratch *  => *<bold>scratch</bold>*

It might look indeed ugly.  However, there is likely to be this
custom (i.e. hiding asterisks) only in Gnus and others don't
seem to do so.

Re: asterisks disappear

[Miles Bader]

Katsumi Yamaoka <yamaoka@jpl.org> writes:
> I still think it is better to show all text by default.  It
> doesn't mean abolishing emphasis but, for example:
>
> * scratch *  => *<bold>scratch</bold>*

I understand what you mean.  For bold, it's not so bad.  But in the case
of underlining, it's _really_ ugly -- underlined+non-hidden text looks
_worse_ than simply doing no processing at all.

[BTW, if you're going to retain the emphasis characters, apply the
emphasis to them too.  <bold>*foo*</bold> looks less weird than
*<bold>foo</bold>*.]

> It might look indeed ugly.  However, there is likely to be this
> custom (i.e. hiding asterisks) only in Gnus and others don't
> seem to do so.

I don't think it's so important what other clients do, I think it only
matters that what Gnus does (1) `makes sense', and (2) doesn't screw up.

In the case of _underlining_, Gnus' heuristics work very well -- there
seem to be extremely few false matches -- but for *bold*, Emacs buffer
names and common-lisp global-variables provide a source of frequent
false matches.

Since underlining+non-hidden looks _really_ bad, but bold+non-hidden is
not that horrible, I think this means that they deserve independent
defaults.

-Miles
-- 
Quidquid latine dictum sit, altum viditur.

Re: asterisks disappear

[Katsumi Yamaoka]

>>>>> In <87oei8kl0m.fsf@tc-1-100.kawasaki.gol.ne.jp> Miles Bader wrote:

> Katsumi Yamaoka <yamaoka@jpl.org> writes:
>> I still think it is better to show all text by default.  It
>> doesn't mean abolishing emphasis but, for example:
>>
>> * scratch *  => *<bold>scratch</bold>*

> I understand what you mean.  For bold, it's not so bad.  But in the case
> of underlining, it's _really_ ugly -- underlined+non-hidden text looks
> _worse_ than simply doing no processing at all.

> [BTW, if you're going to retain the emphasis characters, apply the
> emphasis to them too.  <bold>*foo*</bold> looks less weird than
> *<bold>foo</bold>*.]

I agree.  I'm going to modify the default behavior of emphasis
bold during this week.  In addition to this, I'd like to improve
the customization type for the `gnus-emphasis-alist' variable so
that it may be easier to understand.  Probably, to show actual
regexps and `Match group' subsections will be unnecessary.

>> It might look indeed ugly.  However, there is likely to be this
>> custom (i.e. hiding asterisks) only in Gnus and others don't
>> seem to do so.

> I don't think it's so important what other clients do, I think it only
> matters that what Gnus does (1) `makes sense', and (2) doesn't screw up.

Indeed.  I only felt uneasy since I wrote buffer names in some
articles recently.

> In the case of _underlining_, Gnus' heuristics work very well -- there
> seem to be extremely few false matches -- but for *bold*, Emacs buffer
> names and common-lisp global-variables provide a source of frequent
> false matches.

> Since underlining+non-hidden looks _really_ bad, but bold+non-hidden is
> not that horrible, I think this means that they deserve independent
> defaults.

I agreed with you extensively.  Thank you.

Re: asterisks disappear

[Romain Francoise]

Katsumi Yamaoka <yamaoka@jpl.org> writes:

>> [BTW, if you're going to retain the emphasis characters, apply the
>> emphasis to them too.  <bold>*foo*</bold> looks less weird than
>> *<bold>foo</bold>*.]

> I agree.  I'm going to modify the default behavior of emphasis
> bold during this week.  In addition to this, I'd like to improve
> the customization type for the `gnus-emphasis-alist' variable so
> that it may be easier to understand.

I see you committed your changes a few hours ago to both HEAD and the
v5-10 branch.  I agree that you have a point about new users being
confused by this, but I don't think it is appropriate to change the
default behavior of a stable release in this fashion...

FWIW, I like the way bold emphasis works in previous Gnus versions, and
I'd customize it back if the new way becomes the new default.

Cheers,

-- 
Romain Francoise <romain@orebokech.com> | I am the walrus. I am the
it's a miracle -- http://orebokech.com/ | walrus. I am the walrus.

Re: asterisks disappear

[Katsumi Yamaoka]

>>>>> In <87oei5u9u1.fsf@orebokech.com> Romain Francoise wrote:

> Katsumi Yamaoka <yamaoka@jpl.org> writes:

>> I'm going to modify the default behavior of emphasis
>> bold during this week.  In addition to this, I'd like to improve
>> the customization type for the `gnus-emphasis-alist' variable so
>> that it may be easier to understand.

> I see you committed your changes a few hours ago to both HEAD and the
> v5-10 branch.  I agree that you have a point about new users being
> confused by this, but I don't think it is appropriate to change the
> default behavior of a stable release in this fashion...

Yes, I broke intentionally the rule for the v5-10 branch which
forbids changes other than bugfixes.  If they don't produce
actual harm even if functions are inferior to No Gnus, I won't
improve them.  However, the conventional behavior which strips
asterisks from buffer names has continued preventing from
exchanging of messages with Gnus users who don't have full
knowledge of them.  New Gnus which new Emacs contains is very
attractive, and it will fascinate many new users.  I thought
that the change is required because of them.  It is not for the
sophisticated users, of course including you.

> FWIW, I like the way bold emphasis works in previous Gnus versions, and
> I'd customize it back if the new way becomes the new default.

I will probably also do the same.  For that purpose, I made it
easy to change the default without altering the Lisp form of the
gnus-emphasis-alist variable.

Re: asterisks disappear

[Reiner Steib]

On Tue, Nov 09 2004, Katsumi Yamaoka wrote:

> I'm going to modify the default behavior of emphasis bold during
> this week.  In addition to this, I'd like to improve the
> customization type for the `gnus-emphasis-alist' variable so that it
> may be easier to understand.  Probably, to show actual regexps and
> `Match group' subsections will be unnecessary.

After customizing "Show start and end patterns: On" for
"gnus-emphasis-underline", not all underscore chars are visible:

| (_one_) _two_

The underscores around "one" are visible, while those around "two"
aren't.

(See also <news:v9oei57d76.fsf@marauder.physik.uni-ulm.de> in
gmane.test.)

Bye, Reiner.
-- 
       ,,,
      (o o)
---ooO-(_)-Ooo---  |  PGP key available  |  http://rsteib.home.pages.de/

Re: asterisks disappear

[Katsumi Yamaoka]

>>>>> In <v93bzglbh3.fsf@marauder.physik.uni-ulm.de> 
>>>>>	Reiner Steib <reinersteib+gmane@imap.cc> wrote:

> After customizing "Show start and end patterns: On" for
> "gnus-emphasis-underline", not all underscore chars are visible:

> | (_one_) _two_

> The underscores around "one" are visible, while those around "two"
> aren't.

I see.  The emphasis rule for underline text is divided into two
parts.  One is the element which you customized, the other is
the last element of gnus-emphasis-alist.  So, you needed to
change last one as well.  I have not succeeded in unifying them,
sorry.

Re: asterisks disappear

[Jesper Harder]

Miles Bader <miles@gnu.org> writes:

> In the case of _underlining_, Gnus' heuristics work very well -- there
> seem to be extremely few false matches -- but for *bold*, Emacs buffer
> names and common-lisp global-variables provide a source of frequent
> false matches.

FWIW, I have

((gnus-emphasis-alist
  (cdr gnus-emphasis-alist)))

in the group parameters for comp.lang.lisp (bold happens to be the
first element in the alist).

-- 
Jesper Harder                                <http://purl.org/harder/>

Re: asterisks disappear

[Ted Zlatanov]

On Mon, 08 Nov 2004, miles@gnu.org wrote:

> Katsumi Yamaoka <yamaoka@jpl.org> writes:
>>> How about turning asterisk highlight off if the word "buffer" is in
>>> the message?
>>
>> That's a bright idea!  I need to try it on later.  Thanks.
> 
> I think it cause a lot of confusion if turned on by default, though, as
> it would likely miss a lot of cases, and maybe make the astericks appear
> to be hidden randomly!

I think it's an improvement on the current situation.

> Maybe it's best to make underlines default to "hidden" and astericks
> default to "shown" and simply make sure that's fairly convenient to
> change these defaults, especially per group (for my personal usage, it's
> probably sufficient to simply makes astericks shown in any group whose
> name matchs ".*emacs.*", and hidden otherwise).

In addition, I would have a customizable list of words whose
appearance in a message tells us to turn off highlighting:

"buffer"
"scratch"
"Messages"
etc.

Then I would make the highlight default as you suggest, with the
additional possibilities of 'smart-detect (what I proposed), t (the
current state), and, of course, nil.

Ted