Re: Calcul ate Tape B ytes to Tr acks‏

2010-05-03 Thread Staller, Allan
2**0 = 1 >32760 = 4096 x 8 4096 * 8 = 32768 No power of 2 ends in a ZERO (or an odd number). Whatever the real number is, it's greater than 27998. -- For IBM-MAIN subscribe / signoff / archive access instructions, send email

Munged Subject [Calcul ate Tape B ytes to Tr acks‏]

2010-05-02 Thread Thompson, Steve
-Original Message- From: IBM Mainframe Discussion List [mailto:ibm-m...@bama.ua.edu] On Behalf Of john gilmore Sent: Sunday, May 02, 2010 9:16 AM To: IBM-MAIN@bama.ua.edu Subject: Re: Calcul ate Tape B ytes to Tr acks‏ I have been watching subject lines get munged for a while. And so I

Re: Calcul ate Tape B ytes to Tr acks‏

2010-05-02 Thread john gilmore
Mea culpa maxima: 32760 = 4095 x 8. 32760 is the largest multiple of 8 that is less than or equal to 2^15 - 1 = 32767, the capacity of a signed halfword. The alignment of a buffer on a doubleword boundary does confer small performance advantages in some circumstances. More important, it

Re: Calcul ate Tape B ytes to Tr acks‏

2010-05-02 Thread Ted MacNEIL
>32760 = 4096 x 8 4096 * 8 = 32768 No power of 2 ends in a ZERO (or an odd number). Whatever the real number is, it's greater than 27998. - Too busy driving to stop for gas! -- For IBM-MAIN subscribe / signoff / archive access

Re: Calcul ate Tape B ytes to Tr acks‏

2010-05-02 Thread john gilmore
IBM is right. BLKSIZE=32760 is the maximal signed integer value that is 1) not greater than the capacity of a halfword (usually required to avoid control block overflow) and 2) a multiple of 8, 32760 = 4096 x 8 (required for doubleword buffer alignment). John Gilmore, Ashland, MA 01721