[Fixed] uneven summary lines

[Fixed] uneven summary lines

[Katsumi Yamaoka]

Hi,

I am trying to use the following settings in .gnus.el file:

(setq gnus-use-correct-string-widths t)
(setq gnus-summary-line-format
      "%C%U%R%z[%5N]%I%S%-60= %(%[%-16,16n%]%)\n")

I expected Gnus shows summary lines orderly as follows:

O. [38006]Is there a new agent bug?                          [Harry Putnam    ]
O. [38007]    Re: Is there a new agent bug?                  [Kai Großjohann  ]
O. [38021]        Re: Is there a new agent bug?              [Harry Putnam    ]
O. [38010]    Re: Is there a new agent bug?                  [Paul Stodghill  ]
O. [38022]        Re: Is there a new agent bug?              [Harry Putnam    ]
O. [38008]Re: X-Face and depth                               [Andreas Büsching]
O. [38011]    Re: X-Face and depth                           <Lars Magne Ingeb>

However, authors were not in a column if a subject contains
Japanese wide characters.  So, I have modified the function
`gnus-spec-tab' to deal with wide characters when the option
`gnus-use-correct-string-widths' is non-nil, in Gnus CVS.

By the way, why is the default value of
`gnus-use-correct-string-widths' t only for XEmacs?
The value t is fit for me irrespective of the Emacs version, and
nil is fit for people irrespective of the Emacs version, either.
-- 
Katsumi Yamaoka <yamaoka@jpl.org>

Re: [Fixed] uneven summary lines

[Lars Magne Ingebrigtsen]

Katsumi Yamaoka <yamaoka@jpl.org> writes:

> By the way, why is the default value of
> `gnus-use-correct-string-widths' t only for XEmacs?
> The value t is fit for me irrespective of the Emacs version, and
> nil is fit for people irrespective of the Emacs version, either.

I thought what we were trying to fix with
`gnus-use-correct-string-widths' was the tendency of `format' under
Emacs and XEmacs to return different lengths.

Emacs:

(length (format "%10s" "你好"))
=> 8

XEmacs:

(length (format "%10s" "你好"))
=> 10

However, there may be other subtle issues in the gnus-spec code that
introduces other problems that `gnus-use-correct-string-widths' fixes,
but my feeling is that these problems are probably bugs in Gnus in the
Emacs case...

I'll do some testing.

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: [Fixed] uneven summary lines

[Daniel Pittman]

On Sat, 19 Jan 2002, Lars Magne Ingebrigtsen wrote:
> Katsumi Yamaoka <yamaoka@jpl.org> writes:
> 
>> By the way, why is the default value of
>> `gnus-use-correct-string-widths' t only for XEmacs?
>> The value t is fit for me irrespective of the Emacs version, and
>> nil is fit for people irrespective of the Emacs version, either.
> 
> I thought what we were trying to fix with
> `gnus-use-correct-string-widths' was the tendency of `format' under
> Emacs and XEmacs to return different lengths.

[...]

> However, there may be other subtle issues in the gnus-spec code that
> introduces other problems that `gnus-use-correct-string-widths' fixes,
> but my feeling is that these problems are probably bugs in Gnus in the
> Emacs case...

My recollection, and why the code got written initially, was the
incorrect width used for `cut' forms, pretty much, and for padding in
the summary buffer.

        Daniel

-- 
One likes to believe in the freedom of music,
But glittering prizes and endless compromises
Shatter the illusion of integrity.
        -- Rush, _The Spirit of Radio_, 1980

Re: [Fixed] uneven summary lines

[Lars Magne Ingebrigtsen]

Daniel Pittman <daniel@rimspace.net> writes:

> My recollection, and why the code got written initially, was the
> incorrect width used for `cut' forms, pretty much, and for padding in
> the summary buffer.

Let's see...

(length "日本語")
=> 3

(gnus-correct-length "日本語")
=> 6

Indeed.  I've now defaulted `gnus-use-correct-string-widths' to t
again. 

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: [Fixed] uneven summary lines

[Daniel Pittman]

On Sun, 20 Jan 2002, Lars Magne Ingebrigtsen wrote:
> Daniel Pittman <daniel@rimspace.net> writes:
> 
>> My recollection, and why the code got written initially, was the
>> incorrect width used for `cut' forms, pretty much, and for padding in
>> the summary buffer.
> 
> Let's see...
> 
> (length "日本語")
> => 3
> 
> (gnus-correct-length "日本語")
> => 6
> 
> Indeed.  I've now defaulted `gnus-use-correct-string-widths' to t
> again. 

Things fail if `gnus-use-correct-string-widths' is *not* set (== nil)
under XEmacs. IIRC, Gnus works correctly without that under Emacs,
though.

Anyway, keeping it that way for (featurep 'xemacs), at least, would be
nice.

        Daniel

-- 
The most potent weapon in the hands of the oppressor 
is the mind of the oppressed.
        -- Steve Biko

Re: [Fixed] uneven summary lines

[Lars Magne Ingebrigtsen]

Daniel Pittman <daniel@rimspace.net> writes:

> Things fail if `gnus-use-correct-string-widths' is *not* set (== nil)
> under XEmacs. IIRC, Gnus works correctly without that under Emacs,
> though.

No -- this is under Emacs:

(length "日本語")
=> 3

This is, of course, correct, although not right for Gnus' purposes.

However:

(string-width "日本語")
=> 6

Which is right.  So perhaps I should just conditionalize it to use
`string-width' if that exists and `gnus-use-correct-string-widths' is
not set, and default `gnus-use-correct-string-widths' to `(featurep
'xemacs)' again?

-- 
(domestic pets only, the antidote for overdose, milk.)
   larsi@gnus.org * Lars Magne Ingebrigtsen

Re: [Fixed] uneven summary lines

[Daniel Pittman]

On Sun, 20 Jan 2002, Lars Magne Ingebrigtsen wrote:
> Daniel Pittman <daniel@rimspace.net> writes:
> 
>> Things fail if `gnus-use-correct-string-widths' is *not* set (== nil)
>> under XEmacs. IIRC, Gnus works correctly without that under Emacs,
>> though.

[...]

> Which is right.  So perhaps I should just conditionalize it to use
> `string-width' if that exists and `gnus-use-correct-string-widths' is
> not set, and default `gnus-use-correct-string-widths' to `(featurep
> 'xemacs)' again?

That sounds fine to me, though I should note that the
length/string-width stuff got exactly the same results for me with:

"21.5  (beta4) \"bamboo\" XEmacs Lucid"

        Daniel

-- 
It is easier to build strong children than to repair broken men.
        -- Frederick Douglas