Rollin Hill wrote: > '| t1:take 2', > '| Join', > '| j:Juxtapose', > '| t2:take 4', > '| s:Specs', > ' Outstream 0', > ' 1.18 1', > ' write', > ' Outstream 1', > ' 19.18 1', > ' write', > ' Outstream 2', > ' 37.18 1', > '| Join * / /', etc.
Nice, that eliminates the extra JUXTAPOSE from my solution (though we still need a separate pipeline for each header record). I'd still put the TAKE 4 and CHOP 0 *before* the alternate input of JUXTAPOSE, rather than sticking the headers onto all the data records only to have to untangle them again. ¬R