I'm using
%@filewrite[%hfile,%[count]^t%@eval[%count - %OKcount]] in a BTM and it's giving, for example, 220^t51. Is that WAD? It's no big deal; @CHAR[9] works.