[tz] Preparing to fork tzdb

Russ Allbery eagle at eyrie.org
Thu Sep 23 04:52:26 UTC 2021

dpatte <dpatte at relativedata.com> writes:

> But I have yet to get a clear answer as to whether we can request
> modifications to backzone records even if they only affect pre1970
> clocks, and whether we can add new backzones for regions that differ
> before 1970.

Obviously, I'm not Paul and can't give you an authoritative answer, but I
have read the archives of this mailing list back to its inception and I
see absolutely no reason to believe that you can't contribute
modifications to backzone records or that those wouldn't be accepted.
Other people certainly have contributed such entries in the past and they
were accepted.

For example, from just last year:


Or in 2019:


Or in 2017:


Of course, they would need to come with supporting references for why they
are correct, and there are some time practices that aren't representable
or are not reasonable to track, which might be exceptions.  For instance,
theory.html mentions:

    In particular, the tz database's LMT offsets should not be considered
    meaningful, and should not prompt creation of timezones merely because
    two locations differ in LMT or transitioned to standard time at
    different dates.

so presumably that specific change would not be welcome.  But in general,
I see no reason why contributions to backzone would not be welcome.  I
just refreshed my memory by reading the thread in 2014 that originally
introduced backzone and the clear intent at the time was, among other
things, to capture such contributions, including by adding new zone
identifiers when necessary.

There does seem to be some reluctance expressed in theory.html to take on
the full number of entries that would be required to fully express
historical time, and I know there have been some ambiguous past
discussions about this, but in practice I don't think this will be
relevant.  Given the long history of the project and past experience, I
think we can predict with a fair amount of confidence that the number of
people willing to research and contribute such entries will be modest and
are unlikely to overwhelm the project with data.

Anyway, obviously Paul (or Tim) would be canonical here; I'm just an
observer.  But my clear impression from reading the mailing list over the
years is that the reason for the low volume of contributions to that data
is lack of interest among contributors and the difficulty of the effort,
not that the maintainers are declining such submissions.  Few people care
about such data enough to do the work to track it down and document it,
but there doesn't seem to be any reluctance to incorporate it if someone
does do that work and provides it in the form of a patch.  (It obviously
gets somewhat lower priority than changes to times after 1970 if multiple
things are competing for maitnainer attention.)

Russ Allbery (eagle at eyrie.org)             <https://www.eyrie.org/~eagle/>

More information about the tz mailing list