A few days ago, this snippet passed my way, which I'm sure the members of the tz list are already familiar with:
While today the Knesset passed the initial proposal to reduce DST by some three weeks, a new compromise is being worked out between Minister of Justice Meir Sheetrit and Minister of Interior Eli Yishai to revert to standard time for a period of 48-96 _hours_ (sic) around the Yom Kippur fast day (September 15-16) and then go *back* to DST until the end of October. The details of the proposal have yet to be worked out, but the second and final readings of the bill have until July 24 to pass.
That announcement prompted this reflection: I've occasionally wondered how much the deployment of our favorite tz package was slowed (back before it was as widespread as it is now, that is) by a perception that it was overkill. "Yes, there have been some historical quirks with time zones and DST in the past, and we used to be at the mercy of lawmakers who didn't know better, but now that everyone knows what a bad idea irregular times are, surely there won't be any more of them going forward, so we can use a minimal, compiled-in table of the historical quirks, and avoid all of these dynamically-loaded tables and associated overhead." If the 48-96 hour version of the Israeli bill passes, I'm thinking that Arthur David Olsen ought to get the Turing Award.