Value range compatibility is currently implemented as a filter, which is clearly sub-optimal. Options:
- Bring in the contain/containedIn joiners from CS. This has its own downsides; joiners would have to get
solutionView argument, and that'd be a breaking change. It would also make CS joiners and MS joiners further diverge, at which point the indexer logic might just split off and become full code duplication.
- Bring in precompute from CS and do a relatively complex join which would still run the filter for which values are in which value range, but that information would be precomputed and therefore "free".
- ... ?
Value range compatibility is currently implemented as a filter, which is clearly sub-optimal. Options:
solutionViewargument, and that'd be a breaking change. It would also make CS joiners and MS joiners further diverge, at which point the indexer logic might just split off and become full code duplication.