home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!think.com!hsdndev!ncar!ico!dirtydog.ima.isc.com!karl
- From: karl@ima.isc.com (Karl Heuer)
- Newsgroups: comp.std.c
- Subject: Re: Change to time() fn?
- Message-ID: <1992Sep13.184530.27370@ima.isc.com>
- Date: 13 Sep 92 18:45:30 GMT
- References: <1992Sep11.194827.19500@iqsc.COM> <1992Sep12.042747.26773@ima.isc.com> <1992Sep12.140423.23672@iqsc.COM>
- Sender: usenet@ima.isc.com (news)
- Organization: Interactive Systems, Cambridge, MA 02138-5302
- Lines: 21
-
- In article <1992Sep12.140423.23672@iqsc.COM> matt@iqsc.COM (Matt Reedy) writes:
- >And you're right, the number of seconds from that point to today is roughly
- >60*60*24*365*92 which is 2.9 billion.
- >
- >The interesting thing about Microsoft's doc is that their example program
- >calls this function, and does a printf("The number of seconds since 1/1/70:").
- >So, which one is right, the comment or the sample program???
-
- Assuming they implemented time_t as a signed 32-bit integer, didn't you just
- finish agreeing with me that the documentation *cannot* be correct?
-
- Of course, if you have access to this implementation, you could run the
- sample program and do the arithmetic on the output value, and verify that
- it does correspond to the 1970 Epoch. My guess is that someone in charge
- of documentation decided that an American audience would have trouble
- understanding "00:00" and so attempted to change it to "midnight" (meaning
- 24:00) of the previous day; and this person managed to get 1899 when
- subtracting one from 1970. (Possibly through a typo, or by blindly assuming
- that the "1970" in the original *must* have been a typo for 1900.)
-
- Karl W. Z. Heuer (karl@ima.isc.com or uunet!ima!karl), The Walking Lint
-