[OSM-talk] The Return of the Highway tags and other junk
guy at graviles-reynolds.org
guy at graviles-reynolds.org
Mon Dec 18 19:24:08 GMT 2006
Quoting Ben Robbins <ben_robbins_ at hotmail.com>:
> In tagging it like this, data has been lost, or incorrectly mapped. The
> highway is a footway. The tracks type is unclear. What the railway was is
>
> unclear, and is nessesery as some people may wish to know. Gates and
> cattle_grids are not infinitly small features, so should not be nodes, as
> this fails to give information about them.
>
> I object to the idea that a person cant tag valid relevent information
> becasue others don't wish to tag it themselves. If that was the case, I
> would have no highway=motorway tag, just cause I don't need it. Others
> would then be limited.
>
The problem with this view point is that of where to you stop with the detail,
since nothing is infinitely small, do you measure and plot the dimensions of
each fence post or why aren't we plotting the actual edges of each road rather
than placing a segment roughly down the centre, or the segments the same length
as the seconds in a concrete highway?
The accuracy of the map, being as it is based in the most part on GPS tracks,
is not good enough to warrent the capture of details such as Cattle Grids and
Gates as anything other than nodes unless they are bigger than the margin of
error in their placement on the map. Having a map at a scale that has items
rendered a couple of metres by few centimetres in the case of a gate or few
metres by a fews metres in the case of a grid both clearly visible and
dimensionally accurate is somewhat pointless if they are potentially displaced
by 8 or 10 metres (or 150m in the case of some of the tracks in the system) in
any direction.
Given the limitations on the accuracy of the OSM data, storing CattleGrids and
Gates as nodes is quite appropriate.
Turning to the issue of the viaduct the current tag node:railway=viaduct is
clearly inappropriate since the whole reason for differentiating a viaduct from
a bridge is its sheer length, also viaducts are not limited to carrying
railways. You would thus be justified and supported in requesting a change in
or the total removal of the tag. In my opinion a viaduct could adequately be
represented by a segment the length of the viaduct being tagged a bridge=yes,
or a series of segments if the viaduct is curved.
On the issue of a dissued railway now used as a track, we have a number of ways
locally that were once, railways, and roads that now have an alternate use such
as a bus route or cycleway. Since they clearly need to be tagged with their
current use, and not both their current and historic using standard tags as the
create a myriad of rendering issue as to which should be rendered, this prompts
the requirement of a tag such as old_way, which allows the inclusion of useful
historic information, again of proposed this tag would be supported. As a
result your viaduct would be a way tagged:
highway=track
bridge=yes
old_way=abandoned_railway
name=somename viaduct
or in the case of Wallington Road adjacent to Clothall Common in Baldock:
highway=cycleway
oldway=tertiary_highway
old_name=Wallington Road
Guy
More information about the talk
mailing list