[Tagging] Mapping time zones as geometries (relations)
colin.smale at xs4all.nl
Mon Mar 6 17:59:07 UTC 2017
On 2017-03-06 17:51, Paul Johnson wrote:
> On Mon, Mar 6, 2017 at 10:42 AM, Colin Smale <colin.smale at xs4all.nl> wrote:
> On 2017-03-06 17:21, Paul Johnson wrote:
> On Mar 5, 2017 18:30, "Frederik Ramm" <frederik at remote.org> wrote:
> I would like to start a discussion about the mapping of time zones.
> What do you think?
> I'm generally opposed to mapping timezones in OpenStreetMap unless the tzdata maintainers are 100% on board. Since timezones are a royal pain to keep track of, often changing 100+ times a year, on as little as a few hours notice in some cases.
Do you mean territories switch to another timezone 100+ times a year, or
that the details of the timezone (DST dates etc) change 100+ times a
year? If it's the latter, it won't be a problem for OSM if we only store
the TZ name. Looking at the archives of the tz-announce mailing list,
there are only 1-2 actual changes per month which would affect OSM. I
think there are more changes to admin boundaries than to timezones.
No, the maddening number of changes unrelated to DST. Computerphile
rather accurately describes the situation . I only have to deal with
it tertiarily in my line of work and it's still caused a lot of problems
for us as a company.
Interesting video! Nothing new there for me though. The number of times
a territory changes to a new time zone is very limited. His example of
Libya where DST was cancelled with basically no notice does not put
Libya into a new TZ - it changes the definition of the TZ. So this would
not require any changes to OSM. Cyprus on the other hand would require
maintenance in OSM (last year its TZ split so TRNC now follows mainland
Turkey and can therefore be an hour different to the rest of the
island). But changes like these are "relatively rare" and well within
the capabilities of OSM and its mappers to maintain with minimal effort.
All we have in OSM is the name of the timezone, e.g. Europe/London.
>> In what way would the tzdata maintainers need to be "on board"?
> Considering that group most closely tracks this, it would be best if we worked very closely together on getting this data right. A lack of close collaboration would rather render having this data in OpenStreetMap outdated very quickly.
There is a mailing list for updates to tzdata -
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Tagging