Re: Can we use +nan.0?

[David Kastrup]

David Kastrup <address@hidden> writes:

> Han-Wen Nienhuys <address@hidden> writes:
>
>> On Mon, Apr 4, 2011 at 3:15 PM, David Kastrup <address@hidden> wrote:
>>>
>>> Hi, I am trying to come up with an alteration value that does not match
>>> "normal" alterations but is neither positive nor negative (since
>>> otherwise a "change" will cause naturals to be printed).
>>>
>>> +nan.0 appears fine for that.  But can we rely on it being available on
>>> supported platforms?  I don't find it in the sources of Lilypond, though
>>
>> It sounds a bit fishy especially since alterations are normally
>> rational numbers.  Why not use a non-number value to signify this?
>> Maybe #f ?
>
> Because alterations are compared using `=' which bombs out on `#f'
> (which was actually my first attempt).

Actually, we already have code to deal with accidentals marked as 'tied.
So probably one can just mark everything as 'tied.

Which is a bit inaccurate.  Maybe instead mark tied accidentals also as
#f?

-- 
David Kastrup