Re: possible bad output in mensural-ligatures.ly

[Reinhold Kainhofer]

Am Donnerstag, 22. September 2011, 18:07:42 schrieb David Kastrup:
> Graham Percival <address@hidden> writes:
> > On Thu, Sep 22, 2011 at 02:11:53PM +0300, Dmytro O. Redchuk wrote:
> >> On Thu 22 Sep 2011, 13:01 Janek Warchoł wrote:
> >> > However, according to Benko Pal we shouldn't worry about it:
> >> > 
> >> > 2011/9/19 Benkő Pál <address@hidden>:
> >> > > this is an impossible ligature, LilyPond should (but never did)
> >> > > refuse it (with an explanation).

I wouldn't say "refuse it", I would propose to print a warning and continue 
processing (like we do with all other kinds of wrong input)....

> >> If lilypond should refuse it (with an explanation) -- please, can anyone
> >> prepare a tiny example?
> > 
> > If somebody says "this is a regression test", then you can skip
> > the tiny example and just add it directly.  All our developers
> > know what the regression tests are.  :)
> 
> I don't think we can currently employ regression tests that have
> _failure_ as the desired behavior.

We can. Look at input/regression/safe.ly

Cheers,
Reinhold

-- 
------------------------------------------------------------------
Reinhold Kainhofer, address@hidden, http://reinhold.kainhofer.com/
 * Financial & Actuarial Math., Vienna Univ. of Technology, Austria
 * http://www.fam.tuwien.ac.at/, DVR: 0005886
 * LilyPond, Music typesetting, http://www.lilypond.org