Charles Mills wrote: > I'm engaged in an internal dispute about best practice for the TOD clock: > should it be set to universal time (or whatever GMT is now called) with the > appropriate local offset, or alternatively, to local time (with an offset of > zero)? > > The argument for local time I guess is that it's more simple and convenient > -- why complicate life? You don't set your clocks at home to GMT and > remember what the offset is to local time. > > I'm on the other side of the argument. My gut feeling is that GMT is what > the platform designers had in mind -- but as I have said before in this > forum, I have no credentials as an ops guy, and I am getting nowhere with > this argument. > > Can anyone venture a concise statement on why it is best practice to do it > one way or the other? > > One specific question: which time goes into a DB2 program bind timestamp? If > one is compiling on one machine and executing on another, and the machines > are in two different local time zones, would setting the hardware clock on > both machines to GMT make the error messages go away?
there was a tod timer task force in the early 70s (that ran on & off for period of something like 3 months). the two big issues i remember being discussed was what was the first day of the century (and how to explain it to people) and what to do about leap seconds. one of the problems was that MVT(?) started shipping with the epoch at 1970 instead of start of the century (this was initial 370s which were announced and shipped with TOD, a couple new instructions, but virtual memory hadn't been announced yet). gmt was taken as a given. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
