[Talk-GB] OS Boundaries

Lester Caine lester at lsces.co.uk
Sun Apr 25 15:29:06 BST 2010

Dave F. wrote:
> Lester Caine wrote:
>> Dave F. wrote:
>>> Lester Caine wrote:
>>>> But well mapped rivers don't have ways down their middle
>>> Really?
>>> Care to expand on that please?
>> MOST rivers are now being mapped fully and so are areas rather than a
>> line with some arbitrary width. So there is no 'way' corresponding to
>> some arbitrary mid point to the river ...
> I contest your assertation that 'most' is accurate, but that's another
> point.

Where new mapping is being carried out ... most of that work is to provide river 
data in the now prefered format ....

> But...
> http://wiki.openstreetmap.org/wiki/Tag:waterway%3Driverbank
> There should be a way to indicate both the direction of flow, bit also
> the route, where applicable, for boat routes. To show that it goes
> through a lock rather than a weir for example.
> & also for boundaries, of course.

The way associated with a river provides the direction of flow only. There is no 
requirement that it should follow a boat route, or boundary. THOSE need to be 
mapped with additional ways, since the river details can not be assumed to be 
correct in either case?

>>>> Even more important, we need a way to maintain historic information
>>>> such as '1995 boundary' where later boundaries are different.
>>> Why do we need to do that?
>>> I delete out of date data.
>>> Please explain why you think we should keep it?
>> Just because YOU are not using the data does not entitle you to delete
>> it!
> And because you might want to use it doesn't mean it should be kept in
> the database of a *current* map.
> I wasn't suggesting that because I do something a certain way it was
> correct, just that I do it.
> However, is this being done by others? I've yet to come across it in use.
> Do you have a link to a wiki page?

http://wiki.openstreetmap.org/wiki/Key:start_date ( and end_date )

>> The whole reason *I* am interested in OSM is as a base for documenting
>> my genealogical data. Being able to check a location at some point in
>> time is important and while many of the attempts to get time data
>> properly tagged have not been accepted, simple information like
>> 'constructed=1980' would at least allow maps to be rendered to provide
>> a view in a particular year. ONCE that is possible, then the related
>> boundary information is also important.
>>> If a footpath gets moved do you think I should still show a way & mark
>>> it as 'this is where it used to go'?
>> 'closed=2007' makes perfect sense to me. People then coming back to an
>> area that they walked 30 years ago would then see why they can't
>> follow the same route today?
> How far back do you suggest going? AFAIS, we are up to our necks in
> current data let alone trawling through OoD data.
> I live in a old Roman city, if I had the patience & time to go back that
> far, the database would be unreadable & unusable with so much info
> layered on top of each other.

The current unorganized tangle of data would only be improved by adding the 
correct start_date ( and end_date ) tags to that data. In the UK we are 
fortunate to have fairly accurate data on when roads were created, and features 
added. 'start_date' should be a part of every historic object placed on the map. 
The problem is that the FORMAT for that simple data differs from tag to tag, 
rather than providing a single standard that higher level tools can use :(

>> Just like 'micromapping', historic information may not be of interest
>> to everybody, but moving forward, why would you NOT want to maintain
>> data that has already been mapped. We just need agreement on how it is
>> maintained - since the 'history' of object edits is simply no
>> substitute for mapping historic data.

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//
Firebird - http://www.firebirdsql.org/index.php

More information about the Talk-GB mailing list