RFC Errata
Found 1 record.
Status: Rejected (1)
RFC 1609, "Charting Networks in the X.500 Directory", March 1994
Source of RFC: osids (app)
Errata ID: 2696
Status: Rejected
Type: Technical
Publication Format(s) : TEXT
Reported By: Michael Deckers
Date Reported: 2011-01-30
Rejected by: Peter Saint-Andre
Date Rejected: 2011-06-27
Section 2 says:
The integer value is the number of seconds, excluding leap seconds, after midnight UTC, January 1, 1970.
It should say:
The integer value is 63 072 000 when UTC was 1972-01-01; it increases by 1 for every second of UTC, excluding positive leap seconds.
Notes:
At the time when UTC was 1970-01-01, TAI was 1970-01-01 + 8.000 082 s,
according to [ftp://maia.usno.navy.mil/ser7/tai-utc.dat]. (The rate of
UTC was slower than the rate of TAI at the time but there have not been any
leap seconds in UTC between 1970 and 1972-01-01.) The original wording could be
taken to imply that the "integer value" was 63 072 001 when UTC was 1972-01-01
and TAI was 1972-01-01 + 10 s, and reached the value 63 072 002 just 82 µs
later. However, UNIX practice is to assign the value 63 072 000 to
the instant when UTC was 1972-01-01. The proposed wording makes it clear
that seconds of UTC are counted, not any seconds.
-- Regarding negative leap seconds (which have not occurred and probably
never will): "excluding" them would be wrong because, when they occur,
the phase of UTC increases by 2 s (and so must the time_t value) while the
phase of TAI only increases by 1 s. The proposed wording simply does not deal with the case, while the original would do it incorrectly.
--VERIFIER NOTES--
The quoted text does not appear in RFC 1609. However, it does appear in
RFC 4049 (!). If the reporter wishes to file an erratum against RFC 4049,
he will need to file a new report with the correct RFC number.