On Mon, 11 Aug 2014 17:50:07 -0400, John Gilmore wrote: >... note that what >John Eells in fact wrote was > ><begin extract> >It is often better to use 32,760 for load libraries than to use any >smaller block size, and it's never worse. ></end extract> > >As I read this it disparages a block size B < 32760, i.e. 2^15 - 1 = >32767 rounded down to the nearest fullword multiple. > ITYM "doubleword".
>It is silent about half-track blocks for any "current" DASD geometry, >which are in fact alluded to favorably elsewhere in the same post. > You correctly and entirely quoted John Eells. The "alluded to favorably elsewhere" was a citation of Barry Merrills text, with which John E. was (slightly) disagreeing. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
