The 3-digit time zones are stable they change only when there is a change in jurisdiction (.e.g. the break-up of the Soviet Union, Yugoslavia, Ethiopia, etc.) led to the creation of new jurisdictions. One key attribute of these jurisdictions is that they now become the authority/source for establishment/declaration of time zones, switches to daylight saving time, etc. Consequently, changes in 3-digit numbers represent either additional objects in the same object class or significant changes to existing objects in the object class. The two and three alphas are merely human interface attributes of each instance of an object. These human interface attributes can and do change depending on the will of the authority/source as do the associated names. However, the assignment of numeric codes is linked to international recognition of changes in jusrisdictions and handled through the United Nations. On this there is a liaison from the UN Statistical Office to DIN the registration authority for ISO 3166. In short, the following rule-base applies, (1) changes to the numeric codes in ISO3166-1 occur only when agreed to by the UN where its member bodies collectively recognize the addition of a new jurisdiction and/or changes in existing jurisdictions particularly where these involve physical boundaries. (2) Time zones are physical boundary oriented and one needs to have unambiguous identification and referencing in a linguistically neutral and IT-enabled manner of of the authoritative sources stating the rules for each disrcete time zones (3) ISO 3166-1 countries identified through their numeric codes as assigned via the UN and disseminated through DIN provide such unambiguous identification and referencing (4) ISO 3166-1 countries have the authority within this schema, i.e. UN + ISO 3166 rules, to attach whatever names they wish to their entity including 2 and 3-alpha codes as long as these have not already been taken by other entities, i.e. instances of objects in the same class. (5) ISO 3166-1 countries have the authority to identify and label their 3166-2 level object as they see fit including delegated authority, if they so wish, with respect to jurisdiction over time zones in the physical territory referenced through an 3166-2 code. Personally, I find it surprising and a bit disconcerting that those who are taking so much time and effort in identification of time zones, their properties, etc. would take such a cavalier attitude to the unambiguous identification and referencing of the authorities/sources of the same. If we are ever to have a referenceable "database" on date/timezones as Marcus Kuhn is thinking and I see David Keegel also sees as a long term objective, we should take this matter a bit more seriously, i.e. the IT-enabled and IT-interface API versus the multiple human interface equivalents which it may have. Enough said - Jake Knoppers <mpereira@istar.ca>