mainframe date "incident"greenspun.com : LUSENET : TimeBomb 2000 (Y2000) : One Thread
After months of having documentation on unstable behavior by a panel integral to job function ignored or responded to with snide comments-satisfaction! Our division rushed in a new program just after roll over. After
began to complain that data entry was failing, the programmers researched the situation and found that the problem was date related. Allegedly, active fields with a mm 31 yy date caused all input to be rejected. Just admitted to today, but that didn't retract the poor performance reviews given to the employees who thought they had accomplished information input and nothing showing on system files! There were data entry problems before roll over which came nicely into 2000 intact and well. Be interesting to see if the date patch on the new format over-rides the glitches. Since the squirrel panel has also been newly linked to a formerly parallel but not direct linked file a whole new realm of possibilities is peeking over the event horizon. By the way, this system is an ancient IBM entity from the late 1970's that has been so modified, linked, overlaid and patched that it's probably evolved into the government equivalent of Hal. Most recently it was interfaced with a system of private contractors who use NT linked pc's and THAT interface program is buggier than an ant farm. I guess mainframe can be forgiven for being a little on the schizo side.
-- another government hack (email@example.com), February 02, 2000
Good post. Thanks much.
-- Dee (T1Colt556@aol.com), February 02, 2000.
Don't let those ants propogate too far. Let us know if the protocol code ever gets straightened out. Don't do much good to have a ant free NT network if it's forced to talk through a bull horn whose battery is run down, eh?
-- paul leblanc (firstname.lastname@example.org), February 02, 2000.