STL Time Machine Report #32 - Wednesday 21 October 1998 (1998-10-21) I visited the Y2K chat room at http://www.ntplx.net/~rgearity last Tuesday night. Lots of csy2k regulars were there: Rick Cowles, phystad, Frank Ney, GregS, cory hamasaki, BarbKnox, Declan, Jo Anne. It was fun. I entertained Frank with a brief account of my campaign debate, moderated by the League of Women Voters. As I have mentioned before, we plan to start beta testing our Time Machine around November 1. But we actually began last week. Monday we got the official word in a staff meeting - Time Machine testing will require Production level beeper coverage by the normal rotation personnel. I can tell you this is not being received joyfully, and some geeks may bail. We've already ordered extra beepers and cellphones. Our typical application group has a handful of senior programmers who rotate beeper coverage. If there's six in your group it's one week on and five weeks off. The big debate now is, do we attempt to cover production and Time Machine by the same person in the same week, or do we go to one week on production, one week on the Time Machine, and only four weeks off. Either way, the guys on the beepers are not happy. I don't think it's that bad. I expect Time Machine alerts to be less frequent than the normal run of production problems, and we're not finding new bugs, just the occasional setup problem. Language Environment for MVS. It's official, Tech Services will install LE/MVS as the default runtime environment (for batch only) next weekend. CICS will be handled in STEPLIB and DFHRPL later. This puts us one step closer to converting to COBOL for MVS and VM. We still have one batch system that's OS/VS COBOL. We've been waiting for months to get permission to install the Y2K-compliant COBOL II versions of the programs. That's now scheduled within the next couple of weeks. A new SVC 11 Bug. I'm on the rotation beeper this week, and we got burned by another SVC 11/TIME macro bug. An application with an assembler subprogram had some very strange problems last week. Certain files were being created with Tomorrow's date. This one is interesting. Our operating system, MVS/ESA 5.2.2, allows both local time and GMT time to be set during IPL (bootup). In the past, we have always set both times to local time. About ten days ago they installed a Sysplex Timer, which requires us to actually maintain local time separate from GMT time. The assembler program was using the TIME macro to get system date and time, and was coded in such a way that it always received GMT time. Since GMT time is no longer equal to local time, jobs that run after 7:00 PM (Central Time) are getting tomorrow's date written into the filename. Ooops. They're rushing a fix into production, but only users of that particular assembler subprogram are affected. Miscellaneous. I made my prediction a couple of weeks ago that the DOW would dip below 7000 by the end of October. Last week it went the wrong way. It's a good thing I don't have any bets going with -bks-, I'll probably be wrong whatever happens. As of 1998-10-21, My countdown now reads: 11 days until 1998-11-01 (Beta Test begins) 72 days until 1999-01-01 (External Testing begins) 437 days until 2000-01-01 (Rollover) Also, just 13 days until Election day. If anyone's interested, I'll report my vote totals, no matter how embarrassing. Previous Year 2000 Time Machine Reports are available at: http://home.att.net/~arnold.trembley/tmr.htm STANDARD DISCLAIMER: I am NOT an official corporate spokesperson. My opinions should not be held against my benevolent employer. -- Arnold Trembley http://home.att.net/~arnold.trembley/ "Y2K? Because Centuries Happen!"