[Tagging] Feature Proposal - RFC - Validity of Route Relations

Warin 61sundowner at gmail.com
Fri Oct 13 21:22:41 UTC 2017


On 14-Oct-17 04:11 AM, Michael Reichert wrote:
> Hi Andrew,
>
> Am 12.10.2017 um 22:45 schrieb Andrew Davidson:
>> Do you want data users to consume this tag? Or is intended for other
>> mappers to know when something needs to be updated?
> It is their decision what they consume and what not. The primary users
> are mappers and validators. But if data consumers see any benefit, they
> should not be hindered to use this tag.
>
>> There does appear to be a strong push against using tags that indicate
>> something has gone from the map:
>>
>> https://wiki.openstreetmap.org/wiki/Key:end_date
> end_date=* is used on disused objects or objects whose last day of
> operation/service is known. Common examples are disused railway tracks
> or shops which announced the last day they will be open. The latter
> example is mainly used in areas with a high density of active mappers. :-)
>
> The wiki page you linked to discourages the use of end_date=* because
> historic information is kept in OSM only until a certain limit. An
> object with highway=trunk + end_date=2016-05-21 should be tagged
> highway=disused + disused=trunk (or lifecycle prefix instead) and maybe
> with end_date=* because data consumers are not used to look on secondary
> tags which invert the meaning of the main tag.
>
> I did not choose end_date=YYYY-MM-DD because that tag would imply that
> the service on the given public transport line will end YYYY-MM-DD but
> that's wrong. timetable:valid_until=* is intended to be an expiry tag
> (i.e. you don't have to check this OSM object until YYYY-MM-DD.

timetable:end_date=* would be better than a new tag of valid_until=* ?


>
>> If this is for other mappers then maybe:
>>
>> 1. note:valid_until=*
> note=* is not intended to be machine-readable. Why should any note:*=* be?
>
>> 2. source=* source:date=YYYY-MM (ie: more than 12 months ago needs to be
>> updated).
> Unfortunately, sometimes timetables or operators change not only in
> December but also in June. New (railway) lines or stations are not
> always opened at the great timetable change in December. There are
> frequent cases when (smaller) stations are opened in the course of the year.
>
The unscheduled changes will also occur with timetable:end_date=*, or similar.

Personally I don't enter timetable data, not something I expect the map to deal with. .




More information about the Tagging mailing list