Re: texinfo-6.8.90 pretest

[Patrice Dumas]

On Thu, Oct 20, 2022 at 05:15:41PM +0200, Robert Pluim wrote:
> >>>>> On Thu, 20 Oct 2022 16:42:45 +0200, Patrice Dumas <pertusus@free.fr> 
> >>>>> said:
>     Patrice> Previously there were no check at all on @w content.  I assumed 
> that @w
>     Patrice> could only contain "simple text" which meant all the inline 
> @-commands
>     Patrice> except for 'titlefont', 'anchor', 'footnote', 'verb', 'xref', 
> 'ref',
>     Patrice> 'pxref', 'inforef'.
> 
>     Patrice> Actually, for this specific example it seems to me that the code 
> put on
>     Patrice> @w is too long to allow for correct display.  I checked in HTML, 
> and
>     Patrice> indeed, the non breakable part seems to me to be too long to 
> allow for
>     Patrice> good rendering, unless the browser width is within a specific 
> range.
>     Patrice> Wouldn't you obtain the effect you want or even something better 
> by
>     Patrice> simply using @w for the arguments of the @pxref instead, 
> possible adding
>     Patrice> a third argument in @w?
> 
> @pxref often 'expands', if that is the right word, to something that
> is considerably shorter that the source, in this case in info it
> produces:
> 
>     (*note (elisp)Yanking Media::).

I was under the impression that the @w would be more for printed output,
to avoid line breaking at the wrong place.  I can't see clearly why it
would be better, in Info, to avoid a line break within a @pxref, I can
only imagine ugly results (not breaking which means that text would not
wrap in an ugly way), if anything, given how line breaking is done in Info.

> so putting @w on the arguments wouldnʼt help. In html it produces
> 
>     (see Yanking Media in The Emacs Lisp Reference Manual)
> 
> which, whilst longer, fits within the width of any reasonable browser
> window.

The issue is not that it does not fit, but that it forces an ugly break
with a large (but not unreasonably so) browser width.

> I guess we could use @w for "The Emacs Lisp Reference Manual"
> there (but I never read emacs docs in html, so Iʼm the wrong person to
> offer an opinion).

I think that it would avoid more unwanted break in HTML, though there
could still be.  But I think that the main issue is whether, in a
printed manual, not breaking in "The Emacs Lisp Reference Manual" gives
the wanted results, that triggered using @w in the first place.

-- 
Pat