----- Original Message -----
From: "Ian Hickson" <[EMAIL PROTECTED]>
To: "Pentasis" <[EMAIL PROTECTED]>
Cc: <whatwg@lists.whatwg.org>
Sent: Wednesday, November 26, 2008 12:32 AM
Subject: Re: [whatwg] Issues relating to the syntax of dates and times
On Tue, 25 Nov 2008, Pentasis wrote:
But the way it is described now still creates a difference in *possible*
markup between:
The battle of waterloo was fought on <time datetime="1815-06-18">Sunday
18 June 1815</time>
and:
Julius Ceasar was assassinated on the ides of march in the year 44BC.
The spec allows us to use the time-element in the first case but not in
the second, while the type of information and semantics of the sentence
are the same in both cases and both dates are known dates and agreed
upon to be true. I can imagine this to be confusing for authors/users.
Could you, for both cases, give the precise number of seconds from the day
of the event in question until now? My impression is that we would not be
able to give the precise number of seconds since a date on the Roman
calendar. For example, were the years after 44BC regular or intercalary?
As I said before, I think before we start supporting other calendars (like
the Roman calendar, or even the more recent Julian calendar), we should
support non-Gregorian calendars that are in active use today. But I don't
propose to do this at this time, as that is an inordinately complicated
problem and we don't yet know if <time> is going to be widely used or not.
No, I understand. That was not the point I was making. I mean it the other
way around.
I am *not* saying that the second example (44BC) should be able to be
marked-up like this, but that -because we can't mark that one up- neither
shoudl we mark up the second example. In other words the spec should be
clear on the fact that it is not intended for this kind of use either.
Perhaps it should be more of a "time-stamp"? (like the address-element is
actually only used for the author of the article/page/site so this element
is like that?)
I hope I make myself clear?
Bert
