On Tue, Nov 17, 2026 at 1:10 AM James Bellaire <[email protected]> wrote:
> On 11/16/2026 2:35 PM, Mark Davis Ⓤ via tz wrote: > > > CLDR has bypassed the issue by using "PDT", "MDT" and "CDT" year round. > Even though it is odd to have a xDT without an xST. > > That is just a short-term fix for CLDR. Governments really should give a > much longer lead time for timezone changes, especially when it involves the > creation of new timezones. > > The benefit is that "PDT" is easily understood as "one hour ahead of PST" > (Los Angeles, Seattle, etc). Getting people to accept that "PDT" could be > valid year round and not a mistake is a challenge. Especially where common > usage for "standard time" has grown to year round use that includes > daylight time when the time is advanced. (The standard time in Los Angeles > in the month of June is UTC-07, called and abbreviated "PDT" but still > considered the standard time.) > > Woe to Arizona where "MST" really means "Mountain Standard Time" (UTC-07) > year round, with distant contacts making the mistake of assuming Arizona > springs forward. But people seem to accept the concept of MST being UTC-07 > year round (in Mexico, Arizona, northeast British Columbia and Yukon) > easier than PDT being UTC-07 year round. CST, EST and AST are also valid > year round in North America. > > My personal opinion (not speaking for the Unicode Consortium) is that the > best naming convention for the new no-daylight zones that would maintain > continuity with previous names as much as possible and yet be unique (which > is important*), would be to build on "PCT" for both the acronym and the > name: > > Pacific Canada Time (PCT) > Mountain Canada Time (MCT) > Central Canada Time (CCT) > Eastern Canada Time (ECT) > Atlantic Canada Time (NCT) > Newfoundland Canada Time (NCT) — this doesn't have a uniqueness issue, and > could just be Newfoundland Time (NT) > > I am assuming you meant (ACT) for Atlantic time. (NT) is a non-starter for > tzdata since three letter abbreviations are required (otherwise "PT" could > have been used). > Yes, typo! > * The problem with just using "Pacific Time (PT)" and others like it, is > that those are already in widespread use for wall time, for zones that do > still have daylight shifts. > > The problem also includes the three letter minimum in the standard. > > The abbreviations are generally an acronym for the name, not the name > being a backronym for the abbreviation (such as “Common Locale Data & > Rules” replacing “Common Locale Data Repository”. "Pacific Canada Time" is > a creative solution, better than "Canadian Pacific Time" which makes one > think of a major international railroad (now CPKC) or "CP time" which has a > slang meaning. But in tzdata it should be an abbreviation for actual on the > ground use, not something created externally. If one can get people to > commonly use a name for an area's time one is a step closer to having a > valid abbreviation. > Note that CLDR acronyms/abbreviations are *not* the same as the TZDB acronyms/abbreviations — they may sometimes be the same, but often are not. I'm only talking about the CLDR usage, not the TZDB usage. I agree that both the names and the acronyms should be driven by usage. In CLDR we do supply names for the zones, but using "metazone" names which can be common across multiple countries, and are more likely to be understandable. For acronyms/abbreviations, we don't generally supply the t hem except in the (often very few) locales (ie, languages+countries) where they are likely to be understood. For example, only a handful of locales have defined acronyms/abbreviations for what was North American Pacific Time, and there are none for China (see link below, for example). https://www.unicode.org/cldr/charts/49/by_type/timezones.north_america.html#8b673ce436b6213 How well has the public (in general) embraced the government's suggestion > (not requirement) to use the "PCT" abbreviation for "PaCific Time" (or > "PacifiC Time")? I would accept that the legislated "Pacific Time" is used, > but are people also using "British Columbia Time" or "Vancouver Time" for > the area to distinguish the area from Los Angeles and Seattle? Is the use > of "Pacific Time" more common than other alternatives? > CLDR cannot use "Pacific Time" because that would be far too ambiguous and confusing for users. There is still the existing "Pacific Time" which represents 2 offsets at different times of the year, not just 1. If we used the same name, someone reading "Pacific Time" would not know whether it meant UTC-8 or UTC-7, for many months of the year. The same goes for "PT". If Canada adopts distinctive names (like British Columbia Time or Vancouver Time or Pacific Canada Time, or just UTC-7), then we would of course follow them. Otherwise we would probably append a qualifier to the name, such as: - Pacific Time (UTC-7) - Pacific Time (BC) - Pacific Time (Canada) Are people who are seeing "PDT", "MDT" and "CDT" on their devices (thanks > to the use of CLDR) accepting of those abbreviations? Or is the project > being flooded with requests for "PCT", "NWT", "ABT" and "MBT"? >
