[XSL-LIST Mailing List Archive Home] [By Thread] [By Date]

Re: [xsl] Timezone concept broken in XPath 2.0?


Subject: Re: [xsl] Timezone concept broken in XPath 2.0?
From: Michael Ludwig <milu71@xxxxxx>
Date: Mon, 10 Nov 2008 23:54:12 +0100

Deborah Pickett schrieb am 10.11.2008 um 22:56:30 (+1100):
> Consider [...] a network router with its config files stored as
> XML (my router does just that).  Such implementations need not
> even have a real-time clock.  For them, having to implement
> XPath 2.0 fn:adjust-dateTime-to-timezone() is an imposition, but
> not a huge one.

It has just occurred to me that for those systems, implementing
fn:current-dateTime() would probably be the huger imposition ...

And indeed :-)

  All minimally conforming processors 7must7 support [...] a minimum
  fractional second precision of 1 millisecond or three digits (i.e.,
  s.sss).
  -- XQuery 1.0 and XPath 2.0 F. & O., 10.1.1 Limits and Precision

http://www.w3.org/TR/xquery-operators/#date-time-duration-conformance

But then, your network router probably does have a reasonable clock.

By the way, if Kiribati decides to move further east of their already
imposterous +14:00, they'll be in trouble:

  A dynamic error is raised [err:FODT0003] if $timezone is
  less than -PT14H or greater than PT14H [...]
  -- ibidem, 10.7 Timezone Adjustment Functions on Dates and Time Values

So yes, +/-14:00, this is the extent to which the Recommendation has
(grudgingly) accepted the Random Real World (in which Kiribati has been
allowed to stretch the margins of Time and Reason), deterring by means
of dynamic errors from indulging in any further randomness.

And now for another Tequila Sunrise ...

Michael Ludwig


Current Thread
Keywords