Re: Once again I don't understand...

[Han-Wen Nienhuys]

address@hidden writes:
> Han-Wen Nienhuys wrote:
> 
> > BTW, maybe we should change the naming from interface to something
> > else, since BarNumber doesn't "implement an interface". An "interface"
> > is just a group of functions and variables.  A few of those are
> > combined to form the BarNumber, so it's really the otehr way around:
> > side-position-interface implements BarNumber.
> 
> Yeah, I am used to the Java-namings...
> But then...
> If it is the interfaces that contain the variables and setting i.e.
> NoteHead.font-name affects a variable (font-name) in the font-interface,

no: the "interface" is a group of functions (typically implemented in
C++), which read and write variables.  The "interface description"
(scm/interface-description.scm) contains a list of variables read and
written by those functions.  Defaults are in Grob descriptions
(scm/grob-description), but any Grob may have non-default values for
variables (using Grob::set_grob_property (), or Scheme
ly-set-grob-property). The grob provides a shared namespace, so
setting direction for a ^ or V accent both influences the glyph
printed (Script::brew_molecule) as the direction wrt to the stem
(Side_position::aligned_side)

> > What's a better name for this?  Something like "building block"
> 
> How about "behaviour", "conduct" or "conductor"?

Hmm. Behavior sounds better; any other suggestions?


-- 

Han-Wen Nienhuys   |   address@hidden    | http://www.cs.uu.nl/~hanwen/