Thank you very much for your answer. I understand the concern to minimize the count of identifiers where possible. However, now the question arises if the rules of outdated Asia/Hanoi will be updated, too, if for example the rules of Asia/Ho_Chi_Minh might change in the future. Will the rules of backzone-identifiers be adjusted to future changes or not? I would take an answer to this question into consideration to decide if I include the backzone-file into my future tz-distributions or not.
Meno Hochschild wrote:
Should there be a link between Asia/Hanoi and Asia/Ho_Chi_Minh in TZDB? ... (see: https://github.com/ThreeTen/threetenbp/issues/57) Since the same problem was reported for Asia/Yekaterinburg, I don't see why it matters whether Asia/Hanoi is in backzone or in some other file. Anyway, we're not likely to create unnecessary names merely because of improperly configured or buggy software; otherwise we'd need entries for every misspelling of every city on the planet. Asia/Hanoi is in backzone because it is beyond the scope of the tz database: Hanoi's time history since 1970 is identical to that of Asia/Bangkok, the entry that zone1970.tab uses for North Vietnam.