On 2015/08/26 11:26:56, hanwenn wrote:
"preparation of .. " : deriving from Grob is not a feature but a risk.
In fact,
it would even be nice if Item and Spanner could disappear, but that
seems to be
intractable.
I spent a giant amount of energy distangling formatting logic from C++
type
hierarchy, so behaviors can be mixed and matched at runtime, and this
is taking
a step in the opposite direction.
Thanks for the feedback. David had the same concern about the rationale in the review comment but accepted the change to the code apart from it. Will you also be content if I limit the rationale in the commit message to the fact that nobody currently instantiates a base Grob? Or do you object to the change on the grounds that someone might want to instantiate a base Grob? https://codereview.appspot.com/260810043/ _______________________________________________ lilypond-devel mailing list [email protected] https://lists.gnu.org/mailman/listinfo/lilypond-devel
