On Wed, 29 Feb 2012 09:47:03 -0600, Paul Gilmartin wrote: >On Wed, 29 Feb 2012 09:12:20 -0600, Bill Godfrey wrote: >> >>http://publibz.boulder.ibm.com/cgi-bin/bookmgr_OS390/BOOKS/BPXZB5A0/E.6.2 >> >>The value in the word returned by this routine is in seconds-since-1/1/1970, >>but, unlike the value returned by the C library time() function and expected >>by localtime(), it includes leap seconds, so for current dates it is 24 >>seconds too high. >> >No, it's correct; POSIX is the one that's wrong here. (Those >who suspect me of being invariably biased toward UNIX, >please note.) >
Maybe some day I'll write a little test program that takes the STCK value for 1999 from the Principles of Operation table, run it through the logic, see what gmtime() returns. I'm pretty confident it will be 22 seconds past midnight because of the leap seconds. Maybe somebody else will do that before I do, and prove me right or wrong. >>The value is also used in the Rexx code at these links: >> >>ftp://ftp.boulder.ibm.com/s390/zos/tools/wjsip/wjsigamu.txt >> >>ftp://public.dhe.ibm.com/s390/zos/tools/wjsigshl/wjsigshl.txt > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
