I wonder if having a 9!:49]_1 which would enable x., y., ... and also erase x, y, ... (if local definitions exist) would help people with this migration issue.
This would have a performance penalty on every explicit function call (which, if documented, might be sufficient to discourage its use in cases where it does not address a real problem). -- Raul ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
