Re: pdfeTeX error when compiling lilypond.texi with accents in node names

[Oleg Katsitadze]

On Thu, Jul 26, 2007 at 05:03:24PM -0500, Karl Berry wrote:
> Adding @thinspace{} sounds ok to me.  I think I agree with using
>   and U+200A.

Ok, but see below.

> >   
> > http://en.wikipedia.org/wiki/Space_%28punctuation%29#Space_characters_and_digital_typography
> 
> That page makes a point out of distinguishing a "hair space" around
> em-dashes and a "thin space".

Ah, I was wondering about the weird name.

On Thu, Jul 26, 2007 at 09:46:30PM +0200, John Mandereau wrote:
> why not use " "?  We already use it in LilyPond
> HTML manual. " " is unbreakable in Mozilla Firefox.

Indeed, it's unbreakable in Mozilla Firefox, but breakable in Opera
(I've just checked).  Also, according to the link above, it _is_
breakable.  There is U+202F "Narrow No-Break Space", but it is much
wider than U+2009 aka " " aka "Thin Space" (about twice as
wide on my Firefox in "sans-serif" font, whatever that is).

Googling brought up this page:

  http://www.cs.tut.fi/~jkorpela/html/french.html

It says, with regard to the thin unbreakable spaces with guillemets:

 "It's even theoretically unclear what one should use. Version 3.0 of
  the Unicode standard said, in the discussion of language-based usage
  of quotation marks (p. 151–152):  "Of these languages, at least
  French inserts space between text and quotation marks. In the French
  case, U+00A0 NO-BREAK SPACE can be used for the space that is
  enclosed between quotation mark and text; this choice helps
  line-breaking algorithms." Yet, Figure 6-1 after that statement
  displays «French» example with no spacing between the word and the
  quotation marks! As regards to THIN SPACE, it is a compatibility
  character, with the SPACE character as its compatibility
  decomposition. According to the Unicode line breaking rules, THIN
  SPACE, being in line breaking class BA, allows a line break after
  it, and this means that one would need something extra to prevent
  such line breaks."

>From the same page I gather that the only reliable way to get a thin
unbreakable space in HTML is to "use no-break spaces and use some
formatting techniques, such as the CSS word-spacing property, to
reduce the width of the spaces".  HTML output already includes some
CSS definitions, so maybe we can use this trick?

But even if we do that in HTML, I still don't see how we can get thin
unbreakable spaces in the Unicode output.  I'm at a loss now.  Seems
so ridiculous that Unicode does not provide for such an important
language as French (with the stated goal to provide for _any_
language, even the defunct ones).

Oleg