<div dir="auto">Hi Martin,<div dir="auto"><br></div><div dir="auto">So, it's your own route relation. Nice....myself, I would code something in python to match using the OSM API so that it is easily rerunnable. Then I would publish as geojson.</div><div dir="auto"><br></div><div dir="auto">(Also, this is similar to the Slow Ways being discussed again in another thread)</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Sat, 12 Dec 2020, 13:16 Martin - CycleStreets, <<a href="mailto:list-osm-talk-gb@cyclestreets.net" target="_blank" rel="noreferrer">list-osm-talk-gb@cyclestreets.net</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
<br>
Hi everyone,<br>
<br>
I'm currently looking at a dataset with a view to it being published, that <br>
defines a set of routes (various streets and paths), not physically marked <br>
on the ground, i.e. they are subjective rather than objective. The aim is <br>
to publish this as a dataset that then other people using OSM could then <br>
easily use in routing engines to favour the specific paths. In other words, <br>
have some way of referencing the preferred ways easily, avoiding the need <br>
for any kind of GIS-based map-matching.<br>
<br>
The data is currently just an overlay manually drawn over a digital map <br>
background, but the underlying map hasn't actually been used in any way to <br>
decide on the network or check things back home. There are no IP issues in <br>
my view - it's just a set of route preferences. The lines can be redrawn <br>
from scratch if necessary.<br>
<br>
Do people have any tips on how best to create and maintained a linked <br>
dataset?<br>
<br>
I've been considering a few options:<br>
<br>
- Load all the data in the area in QGIS using a Geofabrik extract, and <br>
manually remove everything that isn't relevant, leaving the desired network <br>
only, from which the OSM IDs can be extracted<br>
<br>
- Use some kind of QGIS process to match the locations with some kind of <br>
key/value filtering, with some kind of 10-20m buffer.<br>
<br>
Do people think it's better to publish a list of OSM IDs or as GeoJSON, <br>
which would obviously contain the IDs but also have the benefit of visually <br>
showing the routes?<br>
<br>
I'm aware obviously the data could become unmatched over time, as OSM <br>
changes, e.g. a way is split or paths added that add more detail.<br>
<br>
<br>
Martin, ** CycleStreets - For Cyclists, By Cyclists<br>
Developer, CycleStreets ** <a href="https://www.cyclestreets.net/" rel="noreferrer noreferrer noreferrer" target="_blank">https://www.cyclestreets.net/</a><br>
<br>
_______________________________________________<br>
Talk-GB mailing list<br>
<a href="mailto:Talk-GB@openstreetmap.org" rel="noreferrer noreferrer" target="_blank">Talk-GB@openstreetmap.org</a><br>
<a href="https://lists.openstreetmap.org/listinfo/talk-gb" rel="noreferrer noreferrer noreferrer" target="_blank">https://lists.openstreetmap.org/listinfo/talk-gb</a><br>
</blockquote></div>