[OSM-talk] Bicycle GPS traces - more opendata

Paul Johnson baloo at ursamundi.org
Sat Apr 30 09:14:43 UTC 2016

On Wed, Apr 27, 2016 at 6:03 AM, john whelan <jwhelan0112 at gmail.com> wrote:

> Now that is an interesting idea.  I think they're date stamped as well.

If they're uploaded with the Trackable or Identifiable permission, which is
advantageous for OSM usage but has potential privacy issues for the
paranoid.  This should fall under the category of "duh", and truly paranoid
people probably wouldn't be uploading GPX to start with, but...it is what
it is.


> My only thought was not to clutter up OSM with lots more but there is a
> treasure trove sitting there it just needs some tools to analysis the data.

It's a two edged sword.  The more GPX points in one area, the longer it
takes to download, the longer it takes to render, and (self inflicted
gunshot wound here given how prolific I was with uploading GPX during my
field service years, density increasing the closer you get to Tulsa
International Airport) the more memory and unweildly it gets.

However, if your system has the memory to deal with it (or if we could get
a native editor comparable to JOSM that didn't have to deal with Java's
horrible memory management, or if I understood how to edit OSM in
Merkaartor or QGIS with the level of proficiency and flexibility I have
with JOSM), a high abundance of identifiable GPX can spot potential turn
restrictions (is there lots of traffic making various movements but rarely
to never a specific one?), possible one-ways (are all of the traces in one
direction only?), traffic signals (do some of the traces slow to a stop and
wait before going, while others sail through without slowing down?), yield
signs (same as traffic lights, but does the priority direction almost
always stay at speed?), a rough idea of what the speed limit is (are
highest speed movements roughly the same?), and stop signs (do all the
traces making a specific movement slow to a stop before proceeding?).

Granted, for some of these, one has to assume that the OSM users uploading
in the area in question are, at worst, only bending the rules and not just
hooning it (granted, a couple of my GPX on the road do have rediculously
high speeds, like when I was riding with an Amtrak Cascades Thruway driver
to Canada that thought 110 MPH was appropriate for driving between Seattle
and the border, or a Uber driver deciding 135 was a perfectly prudent speed
for Interstate 44 through midtown Tulsa, or when my buddy decided (on a
completely empty Interstate 40) to find out what the top speed of a Fiat
500 Abarth actually is...), so one does have to take GPX with a grain of
salt.  Though even these outliers can provide some insight to what the
actual usage is on the ground, as well.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20160430/ecbddabc/attachment-0001.html>

More information about the talk mailing list