Re: lilypond-book refactor patch

[John Williams]

On Mon, 4 Apr 2005, Han-Wen Nienhuys wrote:
> address@hidden writes:
> > This patch looks large, but mainly it is just moving code around.
> >
> > It refactors the compose_ly procedure, making it into a method of
> > the Lilypond_snippet class.  The option processing is broken off
> > the start of compose_ly and put into a do_options method.
> >
> > This allows the fully parsed options to be accessed by other methods
> > of the class.  This is needed for some of the patches which follow this.
> >
> > This patch should change no existing functionality, but it makes doing the
> > rest of the patches I am submitting a lot easier.
>
> just to be sure: have you verified that
>
>      make web
>
> in the lilypond source tree still works?

No.  Unfortunately, I just don't get excited about being on the bleeding
edge of every package like you guys do.  I'm working off the binary rpm
version of lilypond.  (Which means these patches are against the
post-path-substituted version of lilypond-book.)

If you could point me at an rpm of mftrace, it would help a lot.  (I'm
running fedora core 3, if it matters.)  I did a lilypond-2.4 compile on
Jan 31, and that was worst thing to compile.  This was a non-root
installation, so I had to compile _everything_ to get versions lilypond
wanted.  Tex was easy; ghostscript was easy; mftrace was an unpleasant
experience.  For example, I ended up compiling both autotrace and potrace,
because mftrace would not compile with the first one I tried. (I think I
tried autotrace first, but I don't remember for sure.)

Anyhow, the html functions in lilypond-book seem a bit neglected, so I
thought you guys might appreciate some help in that area.

~ John Williams