STL Time Machine Report #44 - Tuesday 04 May 1999 (1999-05-04) The boys & girls have been busy. 9/9/99 won't be any problem. I will be busy this month, because the third-party database upgrade to LE/MVS looks like a sure thing. I have six production and two certification CICS 4.1 regions to upgrade by the end of May. These upgrades have been waiting on the database LE/MVS installation for a couple of months. It seems like the hardware and OS upgrades for our Unix network computers will be completed closer to June 1 than July 1. The Unix guys tell me this is almost completely finished. Looks like the last sandbag will be in place by June. Testing will continue through the end of 1999. It seems some end users are requesting some special Y2k test setups. This is a touchy subject, because resources are fully committed. There's no slack in the schedule, and programmers are not sitting around boosting their minesweeper scores. I think you have to handle this kind of thing on a case-by-case basis. You want to keep your users happy, and if it's a test that more than one may want to do, it's probably going to get added. The workload is not winding down yet. I heard an interesting rumor. I have no source for it, it may be a hoax, but supposedly somebody is maintaining a closed Kroger store somewhere just for Y2k testing. They're running electronic cash registers, check verification, credit cards, making sure you can still rent a video during the end of the world. Anybody else heard this story? I'd like to know if it's true. Speaking of grocery stores, I talked with a geek in the business. He tells me they have big supplies (> 3 days) of candy, soft drinks, snack chips, cheese, frozen food, and dog food. The other products have to be replenished more frequently. He says there's a cave in Warrenton, Missouri, where a million frozen turkeys are stored due to over-production. Outside of the family stuffing holidays, Thanksgiving and Christmas, I doubt there's a huge market for frozen turkeys. SIGNS, OMENS, PORTENTS Monday the DJIA closed above 11,000. Surely the end must be near. California debates Y2K price controls. The Senate debates Y2k immunity. We still haven't had a programmer draft. The 250-day milestone passed without any comment. Only the dealers seem to be making money on Gold. My new PC was advertised as Y2K Compliant, but it runs Microsoft Windows 98. First computer I've ever owned with the correct date & time right out of the box. But the time zone was set wrong, maybe I should call a lawyer. The traffic volume on comp.software.year-2000 has definitely fallen off by almost 50% since January. We still haven't seen a fortune 500 failure, and I don't think we will until after rollover, if at all. The Y2k testers seem to fall in the pollyanna camp. I haven't heard anybody say, screw the bonus money and hotel room, I'm buggin' out. SO WHAT'S THE BIG DEAL WITH 1999-07-05? The issue with this date is that in one system I support, we have an online process that adds records with a precalculated purge date 180 days into the future, stored as an IBM Julian date in YYDDD format. 1999-07-05 is 99186. 99186 + 180 days = 00001. Every night the batch job purges records if the calculated purge date is less than the current date. 00001 < 99186? Ooops. This problem was identified in 1996. The code was windowed in 1997 and Time-machine tested in 1998. This is no longer a problem. But on July 5th, some lucky rabbit gets to stay up all night and verify that records with purge dates of 00001 are correctly interpreted as 2000/001 > 1999/186. As of 1999-04-17, my countdown now reads: 62 days until 1999-07-05 (First Fail Date) 128 days until 1999-09-09 (Another contingency date) 242 days until 2000-01-01 (Rollover) 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!"