Han-Wen Nienhuys wrote:
[EMAIL PROTECTED] writes:
Han-Wen Nienhuys wrote:
I want the syntax to reflect what Lily is doing inside. If you can figure out a way to unify both types of properties internally, I will unify the syntax.
Ideally, the user shouldn't need to know anything about the structure of the implementation.
No, but they should know the "specification" of what LilyPond input means. I want to change implementation matches closely the (mental model) of the input language. That's also why I'm doing some major brain surgery on the Translation part, so we can have more consistency and have
\context Score \with { .. }
I think that
If we need the separation, we also have to be able to explain it in a way that's obvious to a person who has never done any programming.
misses the point. If we fail to explain things logically, then the implementation is at fault.
I agree completely on this point of view!
Also, one problem from the users side, even when you have realized the difference between translator properties and backend properties, is that it's not always obvious if a certain feature belongs to the translator or to the backend, when you start searching for a property.
When you unify both, the search space doubles in size - that doesn't exactly help the users.
Yes! And, regarding unification, one of the main points of the backend is that many different output objects share the same properties. If we didn't have this nice feature, the search space wouldn't just double, it would explode!
/Mats
_______________________________________________ Lilypond-devel mailing list [EMAIL PROTECTED] http://mail.gnu.org/mailman/listinfo/lilypond-devel
