The offset indexer currently identifies offsets between the start of the piece and the last observation of whatever series have been passed to it plus the the step (i.e. the quarterLength setting value). Often at the end of a piece, however, the last values are longer than average and our method looses these terminal repeats. If we use each part stream's highestTime attribute (like the duration indexer now does) we could do this properly for the entire piece.
Though it is already started, an offset indexer rewrite will have to wait until after VIS 3.0 when (and if) I have time. This rewrite will also allow for a revisiting of this indexer in general, and also a thorough integration of the new "dynamic" setting.
The offset indexer currently identifies offsets between the start of the piece and the last observation of whatever series have been passed to it plus the the step (i.e. the quarterLength setting value). Often at the end of a piece, however, the last values are longer than average and our method looses these terminal repeats. If we use each part stream's highestTime attribute (like the duration indexer now does) we could do this properly for the entire piece.
Though it is already started, an offset indexer rewrite will have to wait until after VIS 3.0 when (and if) I have time. This rewrite will also allow for a revisiting of this indexer in general, and also a thorough integration of the new "dynamic" setting.