[OSM-talk] Timezones (was: Deleting data)

Lester Caine lester at lsces.co.uk
Sat Oct 19 12:42:18 UTC 2013


Frederik Ramm wrote:
> Time zones, in contrast, are
>
> * orders of magnitude less useful than admin boundaries (most countries will
> simply have one time zone for the whole country)
> * not useful for mappers
> * readily available in shape file form
>
> so why map them?

Except that the currently available 'shape' files are totally useless near 
boundaries putting whole towns in the wrong time zone! The 'boundaries' we are 
looking for ARE in many cases the same as the political boundaries, so SIMPLY 
adding a timezone tag to these boundaries is all that is required ... except 
where there are specific variations in which case these NEED to be mapped. What 
we are currently pushing is that OSM provides the base for that information 
against the TZ database.

The current arguments on the TZ database are that variations in time prior to 
1970 are pointless and so should not even be recorded! Another area where 
historic data is being purged, but many of the details relate specifically to 
locational data which can only be accurately recorded as a map. Yes these are 
also political boundaries in many cases, but it is information that needs to be 
'surveyed' by access to historic material rather than simply looking on the 
ground. This is an area that OHM has been set up to complete the coverage, but 
in a large majority of cases, this material is still current, and so should be 
in OSM first ...

The major problem I've pointed out several times is where a higher level tag 
like 'timezone' is then attached to every element below the boundaries that it 
relates to. We need to develop the boundary relations in a manor that a search 
for a locations 'details' return all of the relevant higher level tags - if 
requested! - and where a higher level tag is applied to several adjacent areas, 
a single relation is created using the boundary of the whole area ... that is if 
one does not already exist. I've come to the conclusion that ADDING areas is 
where a lot of this is now breaking down when one needs to use the edges of 
those areas for other disjointed tagging. At a lower level fields and field 
boundaries where one has three ways all overlaying because two of them are parts 
of areas! Moving up to residential area and other small political boundaries 
which have been drawn as areas but need to be pulled apart to use existing ways 
in much the same way that nodes are automatically shared?

-- 
Lester Caine - G8HFL
-----------------------------
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk



More information about the talk mailing list