[whatwg] Dates BCE

Sam Kuper sam.kuper at uclmail.net
Thu Jul 30 09:47:49 PDT 2009


2009/7/30 David Singer <singer at apple.com>:
> Quite.  We've had this debate before and Ian decided that it might be
> confusing to apply a dating system to days when that dating system was not
> in effect on those days, I think.

If by "confusing" you mean "sufficiently confusing that it needs to be
avoided, then the proleptic Gregorian calendar would not be suitable
for use in HTML5. Yet it has been adopted for HTML5. So either the
confusion is tolerable or the reasoning has been inconsistent. I
assume the former, and actually I think that using the proleptic
Gregorian calendar *decreases* confusion by creating a mutually-agreed
neutral vocabulary for dates that other calendars can be translated
from and to, thus reducing the total number of mappings needed between
calendars if all calendars are to be mappable to each other.

> Against that, one has to realize that
> "the label of the day before X" is well-defined for the day before the
> introduction of the Gregorian calendar, and iteratively going back to year
> 1, year 0, year -1, and so on.  And it would be nice to have a standard way
> of labelling dates in historical documents so that they are comparable; I am
> reminded of Kilngaman's book in which he has parallel chapters for China and
> Rome in the first century CE
> <http://www.amazon.com/First-Century-Emporers-Gods-Everyman/dp/0785822569/ref=sr_1_1?ie=UTF8&s=books&qid=1248970679&sr=8-1>.
> It would be nice if one could determine that two events in separate
> documents were essentially contemporary, despite being labeled in the
> original text in different ways.

It's not simply "nice", it's a necessity for accurate automated
processing of historical or other non-Gregorian temporal information.

> However, whether the spec. formally blesses using <time> like this may not
> be very relevant, as it can be done textually with or without the blessing.

By "textually", do you mean manually? If so, many exciting
possibilities in online historical research would be rendered quite
impractical (as they are currently) simply because of the massive
amount of time that would be required to manually process each date
conversion. This is a *very* real problem.


More information about the whatwg mailing list