Currently -fallow-overlapping-instances only allows overlapping instances if one is a strict subset of the other. This is good (determinate), but sometimes you really need two instances that partially overlap.
From a type-theoretic viewpoint, instance declarations are relations between (sets of) trees (= elements of the respective data types). So one needs representations of such relations with effective decidability of overlapping-ness, most-sepcific-ness and such.
Unsurprisingly, there are representations with nice decidabilities, but they are not very expressive; and on the other hand there are representations with undecidable properties (see -fallow-undecidable-instances) This is an active area of research in the term rewriting/tree automata community, see for example the chapters on "automata and n-ary relations" (as well as "tree set automata") in the TATA book. http://www.grappa.univ-lille3.fr/tata/ Well, at least my impression is that these topics *are* closely related, and *both* should certainly benefit from a closer investigation of their interrelations. (this is what I'll be attempting in one part of my thesis.) best regards, -- -- Johannes Waldmann ---- http://www.informatik.uni-leipzig.de/~joe/ -- -- joe@informatik.uni-leipzig.de -- phone/fax (+49) 341 9732 204/252 --