Re: Timezone concept broken in XPath 2.0?

Michael Ludwig wrote:
> Why aren't the functions specified to
> use a timezone database and then work like timezone-aware localtime() in
> Perl and C?

At a guess, I'd wager that it's the sheer complexity of maintaining a
timezone database that can never be completely future-proofed.  Imagine
if every implementation of XSLT 2.0, including embedded ones, had to
grok timezones all around the world.  Then imagine when $government
decides to change the rules for $region next year...

I agree that the XPath zone-conversion functions that you mention are
limited, and arguably misnamed, but they do have a use: the functions in
XPath 2.0 are enough for you to implement your own zoneinfo database, or
even to parse the Olson database file format.  It's possible to code a
pure userland XSLT 2.0 implementation of zoneinfo, but XSLT
implementations which have no need for it don't have to carry the baggage.