[Talk-GB] “No path”

[Andy Townsend]

On 07/01/2023 16:03, Edward Catmur via Talk-GB wrote:
>
>
> On Sat, 7 Jan 2023 at 15:49, Dudley Ibbett <dudleyibbett at hotmail.com> 
> wrote:
>
>     Hi
>
>     Just seen this article in today’s Guardian:
>     https://www.theguardian.com/uk-news/2023/jan/06/hiking-app-alltrails-changes-route-rescue-three-walkers-lake-district
>
>     I’m not sure if this is relevant to the data in OSM but it does
>     seem to look like a footway/path on the main map if I have
>     correctly located it.   According to the article, the Mountain
>     rescue team describe it as “no path” so it would seem reasonable
>     to ensure the tagging is correct. Perhaps someone who has walked
>     this route can review the tagging.
>
>
> It looks like there's an ongoing edit war over 
> https://www.openstreetmap.org/way/1127488902 and adjacent ways. 
> They're tagged sac_scale=demanding_alpine_hiking in the most tricky 
> parts; what more can we do?

The problem here is that some app developers are including everything 
that OSM has as a "path" and adding it to general purpose maps without 
any clue as to their difficulty.  There was a recent forum discussion 
which included:

https://community.openstreetmap.org/t/feature-proposal-voting-highway-scramble/5228

and

https://community.openstreetmap.org/t/rfc-highway-scramble/2496

about the use of the tag value "highway=scramble" in place of 
"highway=path" in cases such as this, as ways with (in OSM terms) 
"sac_scale=demanding_alpine_hiking" aren't in any sense a "path".

There was some pushback in those discussions that "surely app developers 
can just just look at the sac_sale tag" before deciding to include a 
difficult-to-access way.  Unfortunately, some of them have not shown 
themselves capable of doing that, hence the suggestion to use a 
different tag that forces them to think before doing so.

If https://www.openstreetmap.org/way/1127488902/history really exists in 
some sense (and I'm certainly not going to argue with Wainwright) then 
it should be in OSM, but it shouldn't be suggested to people as "a walk 
in the park" without some clue as to what they're letting themselves in 
for.  One option is something like this:

https://map.atownsend.org.uk/maps/map/map.html#20/54.62859/-3.21546/H

there, the problematic section isn't shown until people explicitly turn 
on a layer showing paths that are difficult / of very limited visibility:

https://map.atownsend.org.uk/maps/map/map.html#20/54.62859/-3.21546/H/N

See

https://www.openstreetmap.org/user/SomeoneElse/diary/400548

for more on this.  Note that that also refers to

https://wiki.openstreetmap.org/wiki/United_States/Trail_Access_Project#Suggested_Tagging

which is interesting for a couple of reasons.  One is that it's an 
attempt by part of the OSM community (in the US) to get app providers to 
"show trails properly" and not lead people astray, which would have 
avoided this problem here.  The language on that page is US-centric, but 
if as I suspect some app developers such as at AllTrails may never have 
been closer to the Lake District than reading Beatrix Potter, it's 
probably speaking their language.

Hopefully the US Trails Access Project can work with developers 
including at AllTrails to get them to start using the richer information 
that OSM can provide them with.  There are a couple of caveats with that 
though - Jerry mentioned a similar recent issue near Causey Pike - see 
https://www.grough.co.uk/magazine/2023/01/05/lakeland-rescuers-urge-caution-when-using-mapping-apps-after-callouts-to-walkers# 
(that also mentions the Barf issue).  That was mapped as 
https://www.openstreetmap.org/way/85797718/history and has been deleted 
but in OSM never had a trail_visibility tag.  If there isn't really 
anything there (just a way of getting from A to B) then it probably 
shouldn't be in OSM, but anything that is in OSM that isn't very visible 
really should have a trail_visibility tag on it.  App developers can't 
be expected to not show a path based on trail_visibility grounds if that 
tag is not in OSM.

Another "interesting" thing is (quoting from the Grough article) - that 
they claim there is an issue with the "perils of relying on digital 
mapping software".  I'd actually be extremely careful when relying on 
commonly-used NON-digital options in some areas.  For example, if you 
compare OSM data at

https://map.atownsend.org.uk/maps/map/map.html#15/54.3577/-1.1581/H

with OS Explorer mapping at

https://www.bing.com/maps?osid=0bd8c7ba-cf68-424a-a5db-9545f5493774&cp=54.355109~-1.153462&lvl=15&style=s&v=2&sV=2&form=S00027

you'll notice quite a few "public rights of way" missing from OSM for 
the reason that there is no evidence of them existing on the ground!

The OS shows non-PRoW tracks and some non-PRoW paths, but on Explorer / 
Landranger it doesn't tend to show dotted path lines underneath PRoWs so 
it's difficult to tell what sort of thing they think should exist there.

Best Regards,

Andy


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-gb/attachments/20230107/61d12214/attachment-0001.htm>