Thanks, Doug. That's a clear and convincing explanation, and I accept it. The point that these are machine-readable identifiers rather than gazetteer names, and that language- or government-specific aliases would set a bad precedent (and impair simple name matching, as the geonames example shows), settles the question. Please consider my 14 October v2 patch withdrawn; I won't pursue localized Zone aliases. The fix belongs in consumers like Grafana, not in tzdb.
Thanks also to Paul for the follow-through: removing the dubious backzone entries and opening the Grafana PR (#134996), which is now merged. That resolves the downstream picker problem that prompted the request. No further asks from me. Thanks again for the quick responses and the cleanup. Best regards, Khikmet Kerimov On Sat, Oct 17, 2026 at 12:00 AM Doug Ewell <[email protected]> wrote: > This type of request would be no more appropriate if the Zones in question > were in active use. > > Creating aliases for Zones to give them “improved,” language-specific or > government-approved names is not a good precedent. These are > machine-readable identifiers, like programming-language variables or API > endpoints; they are not gazetteer entries or map labels. > > If identifiers like “Europe/Zaporozhye” are exposed to end users (and we > know they are), that is unfortunate, but that is not a reason to get tzdb > into the business of providing localized names for cities, regions, > continents, and oceans. The theory file already addresses this (“prefer > Europe/Rome to Europa/Roma”). > > While many installations use the full tz database and code, and will > resolve aliases as needed, there are many others that simply refer to the > name of the Zone. For instance, the geonames database ( > https://www.geonames.org) shows my location as having a “timezone” value > of America/Denver. Adding more additional names for this zone would create > confusion and would impair or prevent matching. > > Removing the existing listings may have been a step forward in this > isolated case, but let’s also try to discourage these requests directly. > > -- > Doug Ewell, CC, ALB | Lakewood, CO, US | ewellic.org > > >
