This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.

Bug 11616 - UTC began at 1961-01-01T00:00Z, and there is no such thing as proleptic UTC. Therefore the Nero example is ill-defined.
Summary: UTC began at 1961-01-01T00:00Z, and there is no such thing as proleptic UTC. ...
Alias: None
Product: HTML WG
Classification: Unclassified
Component: LC1 HTML5 spec (show other bugs)
Version: unspecified
Hardware: Other other
: P3 normal
Target Milestone: ---
Assignee: Ian 'Hixie' Hickson
QA Contact: HTML WG Bugzilla archive list
Depends on:
Reported: 2010-12-30 03:01 UTC by contributor
Modified: 2011-08-04 05:02 UTC (History)
7 users (show)

See Also:


Description contributor 2010-12-30 03:01:56 UTC

UTC began at 1961-01-01T00:00Z, and there is no such thing as proleptic UTC. 
Therefore the Nero example is ill-defined.

Posted from:
Comment 1 Tab Atkins Jr. 2010-12-30 05:13:26 UTC
ISO 8601 disagrees.
Comment 2 Aryeh Gregor 2010-12-30 21:01:24 UTC
Proleptic UTC is logically problematic, given that you wouldn't know which years had leap seconds.  So assuming you had some unambiguous way to identify a particular second in history, there might not be a well-defined algorithm to associate it with a UTC timestamp.  But this seems like a rather academic objection, because nobody is tracking things back to Nero's birthday with second- or minute-level accuracy.
Comment 3 Ian 'Hixie' Hickson 2011-01-24 04:27:36 UTC
That there's no such thing as proleptic UTC is tautological  "proleptic" means we're extrapolating it to a time before it existed.

But I guess I can change it to refer to UT1 instead of proleptic UTC. If I do that I'll try to make sure it's still obvious what it means.
Comment 4 Warner Losh 2011-01-31 16:53:07 UTC
Actually, when UTC started is a lot more complicated than a simple date.

UTC as we know it today started on 01-Jan-1972T00:00:00Z.  This is when whole seconds were inserted into UTC to keep in closely in sync with UT1.  Prior to that date, frequency offsets and small steps of tens or hundreds of milliseconds were inserted to keep them in sync.  That practice does date to 1961.  However, prior to that, the time scale now known as TAI ticked from 1958 forward.  Prior to that, well, time of day was based on a number of different things.  UT1 is just a special variation of UT that was convenient to keep since approximately 1900 (although many different definitions of UT1 existed), so speaking of 'UT1' back 2000 years likely is a bit unwise as well.

Generally, as implemented by most time libraries, proleptic UTC doesn't do anything close to a project back into the past of when leap seconds happened or any of that crazy nonsense.  Instead, it tends to be implemented as more of a proleptic TAI, where time ticks back in time without regard for leap seconds at all.  But the idea is that these seconds that tick aren't SI seconds of a fixed duration, but rather mean solar seconds of whatever era the time stamps are from.  This matches historical practice and gets away from implementations need to have complicated tables of numbers for the period of time we have good data on these small differences (basically since 1600 or so), plus, there's no standard for projecting time back like this: everybody agrees to paper over these fine-details and go with what people used back in the era that time information comes from.

If you are going to fix the Nero example, I'd say that it is midnight universal time, and then explain that in prior eras time keeping was based not on fixed SI seconds like today, but rather were 1/86400 of a mean solar day which has varied throughout history by a few milliseconds.  Times expressed from this era are 'universal' times.  But as a concept, universal times didn't really exist prior to the various treaties that established Greenwich as the basis for the longitude for the globe.

Of course, there is also the problem that timezones didn't exist prior to the railroads promulgating them in the mid 19th century (in the US and UK, other times elsewhere).  Prior to that, all times were either local solar time, or whatever the big clock tower near where you lived said it was.  Each observatory had its own time from about the 1500's forward, just to complicate things further.

So putting this all together, the time example you gave for nero's birth most likely should be interpreted as midnight local time on the date of his birth had be been born somewhere near London.
Comment 5 Ian 'Hixie' Hickson 2011-02-25 07:46:01 UTC
EDITOR'S RESPONSE: This is an Editor's Response to your comment. If you are satisfied with this response, please change the state of this bug to CLOSED. If you have additional information and would like the editor to reconsider, please reopen this bug. If you would like to escalate the issue to the full HTML Working Group, please add the TrackerRequest keyword to this bug, and suggest title and text for the tracker issue; or you may create a tracker issue yourself, if you are able to do so. For more details, see this document:

Status: Accepted
Change Description: see diff given below
Rationale: Proleptic TAI makes no more sense than proleptic UTC, it really is proleptic UT1 that we're talking about.

Anyway I've added a paragraph to the spec that discusses this in far more detail than anyone else is ever going to care about. If it confuses anyone, though, I'll probably remove it again and not worry about it.
Comment 6 contributor 2011-02-25 07:53:12 UTC
Checked in as WHATWG revision r5913.
Check-in comment: Change what was non-normatively implied to be proleptic UTC to explicitly and normatively be proleptic UT1.
Comment 7 Michael[tm] Smith 2011-08-04 05:02:48 UTC
mass-moved component to LC1