proposed tz patches for Mexico, Nunavut, Vostok, etc.
gjb at gbch.net
Mon Apr 2 06:50:27 UTC 2001
| ] Sorry, I haven't had time to think through the issue of Australian
| ] time zone abbreviations yet. If I recall correctly, we have so far:
| ] * one well-researched proposal to change from EST/EST etc to AEST/AEDT etc;
| ] * one second.
| ] * one dissenting vote to leave it at is.
| ] * my own brief research indicating that Australians themselves are
| ] not consistent
| ] Right now we have to get the Mexican patches out, so I'll leave the
| ] Australian abbreviations alone for now, but I'd like more votes and/or
| ] research if possible.
| Thanks for leaving EST/EST vs AEST/AEDT in the pending queue.
| I think that's the best place for it at the moment.
| We don't want a repeat of the LHI thing last year where we changed
| tzdata and then changed it back again. If we change it, we should
| first confirm that the proposed change is right *and* confirm that
| the existing tzdata is wrong.
| I could say that AEST/AEDT is `right' in the sense that is it fairly
| widely used and understood in Australia. But I think the jury is
| still out on whether EST/EST is wrong (I'd say EST/EST is also widely
| used and understood).
In the absence of legislative prescription, there can be no
positively `right' answer for Australia; but there are sensible
and less sensible answers.
Although those of us who are humans can manage quite well when
we are trying to interpret Australian time zones, the task is
more difficult for software. Far too many mail clients just use
the `EST' or whatever string they find and this leaves other
programs in the invidious position of having to determine which
`EST' is meant: is it US Eastern Standard Time; is it Australian
Eastern Standard Time; or is it Australian Eastern Summer Time?
When those unfortunate pieces of software are asked to sort a
mailbox by date, how do they disambiguate these? Obviously, if
the original mailer does the right thing and puts `+1000' or
whatever in the date field, there is no problem; but far too
many mailers do not do this. We can't fix all those mailers,
nor can we stop people from using them. So what we need to do
is provide a mechanism that is unambiguous to assist all the
users who have to work with Australian time zones. Even though
the choice of `AEST/AEDT' can be criticised on various grounds,
the important thing it has going for it is that it disambiguates
these Australian zones from each other and from US `EST'. This
is clearly a Good Thing and nobody has shown any harmful effects
that it might cause. Nor has anybody shown any better mechanism
to do the disambiguating.
Further research is not going to help here; we need a solution
that does something useful and the `AEST/AEDT' proposal is the
only one that stands a chance of being acceptable for that.
 The lack of legislation is, I think, a Good Thing. Our
political masters have never demonstrated any intelligence
when it comes to issues like this and they would be almost
certain to invent some truly dreadful `solution' to this.
More information about the tz