Re: compilation issue with font-name-font-size.ly

[Graham Percival]

On Sun, Jan 16, 2011 at 07:50:57PM +0100, Werner LEMBERG wrote:
> > With this change, I see a dramatic change in font size -- the
> > vertical length of the post-font-name "p" is less than half of the
> > vertical length of the pre-font-name "p".
> 
> Yes.  However, it doesn't matter IMHO.  On the console, lilypond
> reports the same font size, and this should be checked in the
> regression test, nothing else.

I see.  I think I misunderstood your description of the change --
when I compile the current regtest, I see no unusual output (at
least, not with a typical "lilypond font-name-font-size.ly"
invocation).

If the console output is the only thing that matters, could the
texidoc reflect this?  I don't want helpful bug volunteers trying
to figure out what's (apparently) wrong with the graphical output,
if in fact nothing is wrong.

> > Should it be an error instead?
> 
> An error indicating what?

in configure -- if fontconfig < 201005xy, die with an error,
instead of merely giving a warning.  I proposed this because I
thought that your system (which I'm guessing has fontconfig 201005
or higher) produced different graphical output than mine.
It seems that this isn't relevant, so ignore this suggestion.

Cheers,
- Graham