[OSM-talk] Timezones (was: Deleting data)
toby.murray at gmail.com
Mon Oct 21 18:32:17 UTC 2013
Yes, in that small fraction of cases there would be duplication of
positional data. But in some cases where you think this is the case, it
might actually not be. My county border was defined by a river. Now part of
the river is a reservoir and the other part has shifted over time and
through floods. The border was not moved with these changes and still
follows the original course of the river even though there is no way to see
this path on the ground at this point. So is the border really defined by
the river? Or was it defined by the river at a certain point in time which
may or may not be the same as it is now.
Don't get me wrong, I'm not suggesting an edit to nuke all admin boundaries
tomorrow. They are a part of the core OSM database and will likely continue
to be for some time. I'm just pointing out that OSM is not good at them and
that there might be a better way to handle such things. Perhaps this
approach could be tried with time zone data since we don't already have a
large body of them in the database.
On Mon, Oct 21, 2013 at 9:54 AM, Philip Barnes <phil at trigpoint.me.uk> wrote:
> It is not always possible to separate admin boundaries from real world
> features. Rivers, roads or even hedges often define a boundary.
> Phil (trigpoint)
> Sent from my Nokia N9
> On 21/10/2013 15:41 Toby Murray wrote:
> On Mon, Oct 21, 2013 at 6:29 AM, Colin Smale <colin.smale at xs4all.nl>wrote:
>> Back on topic: how do you phrase an objective rule, or at least
>> well-worded guidelines, which allow admin boundaries but disallow time zone
>> boundaries? I wonder where the UK ceremonial counties, fire department
>> areas, national parks etc will end up. My point is, gut feelings aside,
>> that it is not reasonable to single out TZ boundaries for this deprecation.
>> Having edited over a thousand of them, I would not be sad to see admin
> boundaries removed from the general OSM database. I think Russ is on to
> something with his "ClosedStreetMap" concept although that is some terrible
> branding so we need another name :) But at the end of the day, we are
> terrible at maintaining such boundaries and very good at breaking them in
> OSM, mostly because they are usually hard/impossible to spot on the ground
> and verify. So people see random lines going through the area they are
> trying to map and either don't pay attention when they touch them or just
> delete them outright. Essentially what we need is the concept of layers. If
> all the admin/timezone boundaries were in their own "layer" and didn't
> interact with roads, rivers, etc in OSM then they would be much easier to
> keep up to date from external sources.
> Yes, OSM *can* contain just about anything. But if we are terrible at it
> and there are other datasets available that aren't terrible then why should
> we try to poorly duplicate others efforts?
> Some of my opinion may be due to some problems with the way they were
> imported here in the US but I suspect it isn't all that different in most
> other places.
> talk mailing list
> talk at openstreetmap.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the talk