Re: [PATCH] Porrectus: third try

[Juergen Reuter]

On Sat, 6 Oct 2001, Han-Wen Nienhuys wrote:

> address@hidden writes:
> > 
> > * Added auto-properties property for automatic determination of grob
> >   properties add-stem and stem-direction from musical context.
> 
> That is puzzling me: why should this be property? What good does it do
> when it's switched off ?
> 

There are different forms of the porrectus (without stem, with upwards
stem, with downwards stem, solid, outlined, ...).  The musical meaning of
these forms changed dramatically between, let's say, the 13th and 16th
century.  In mensural music of the 16th century, you can determine the
shape of the porrectus from the durations of the notes.  Earlier music
partially uses the same porrectus shapes, but with different meaning,
hence you do not want to apply these rules if you are typesetting
music of a different epoche that still uses the same shape.

Ok, you can argue, that property style should not take values
"mensural" and "vaticana", if mensural style porrectus can be used even
for pre-mensural music.  But that is a rather pragmatically decision,
since the shape is typical for mensural epoche (but not limited to).  And
the same pragmatic decision has already been in lily for years according
to mensural style heads, which also appear in earlier music.

What might be considerable for the future is to replace property
auto-properties with a property that holds a scheme function that
determines the properties from the musical context (durations and pitch of
adjacent notes, current meter, etc.).  Ideally, you would specify such a
function for each of the most important kinds of notation such as
gregorian chant square neumes, ars antiqua, ars nova, mannered notation
(a variation of ars  nova), modal notation, mensural notation,
etc.  However, at the moment I have not enough musical knowledge to design
a general framework for such a function that considers all flavours of
notation between 11th and 16th century.  But I may reconsider this some
time in the future.

Another alternative is to introduce "auto" as property
value.  stem-direction, for example, then could have one of the three
values "up", "down", and "auto".  However, I find the possibility of
setting add-stem to "auto" and, at the same time, fixing stem-direction
e.g. to "down" not very attractive (both, from the user's perspective of
mudela semantics, and from the perspective of writing a proper
implementation).

The bottom line is that you would like to switch property auto-properties
off when you are using mensural-style porrectus ligatures, but do not want
to apply mensural rules, because you are (for example) typesetting music
of an earlier epoche, e.g. as in mannered notation.  In this case you are
on the safe side if you have full control over the properties that control
the stem.  But you surely want to switch automatic property settings on
for convenience, if you are typesetting mensural music of the 16th
century.

Anyway, did you ever encounter a situation when you wanted to switch off
the auto focus of your photo camera (assuming you have a camera)?  You
always want to be able to switch off automatic functions because there
will often be situations where the automatic function just will give you a
result other than you want, right? :-)

Greetings,
           Juergen