[talk-au] Deletion of informal paths by NSW NPWS

[Andrew Harvey]

On Mon, 2 Oct 2023 at 14:19, Ben Ritter <benjaminaritter at gmail.com> wrote:

> I agree with all of this. If the track exists on the ground, something
> should exist in OSM.
>
> This situation is not a novel one that requires a new tag prefix, I think
> it should be represented with:
>
>    - highway=* because it is clearly a track to a surveyor
>    - informal=yes because it is not maintained like the other paths
>    - access=no because the relevant authority says so
>
> I believe it's more nuanced than that.

If the point of the closure is to permanently remove the track and restore
it back to bush, and especially if there has been some work done like
placing branches or fallen tree trunks along the path, or if vegetation is
regrowing within the track, then it should use one of the "stages of decay"
lifecycle prefixes.

If the future status is unknown, but it's currently closed, then that's
where I'd leave the highway=* value intact and add access=no.



On Tue, 3 Oct 2023 at 23:33, Mark Pulley <mrpulley at iinet.net.au> wrote:

> A brief summary of the options for this type of situation (not just this
> particular edit, but similar edits in the past and probably future):
>
> 1. Revert the change sets (in the absence of more information)
> 2. Partial revert, with a change in tags
> 3. Leave the deletion as it is.
>
> For this particular example, the results would be:
> 1. Full revert - way will be marked informal=yes, but without access tags
> 2. Partial revert - could add access=no, or
> alternatively abandoned:highway=* or disused:highway=*
> 3. No reversion
>

I would opt for 2, leave the way in place, but with access=no, a lifecycle
prefix on the highway tag like abandoned:highway=*
or rehabilitated:highway=*.

If there is signage that says closed for rehabilitation, we should
capture the closure reason somewhere, so OSM data consumers can present
that reason for the closure to users, whether that be
via rehabilitated:highway=* or something like, access:reason=rehabilitation.



On Sun, 8 Oct 2023 at 13:55, Ewen Hill <ewen.hill at gmail.com> wrote:

> Hi all,
>   A fantastic thread and I feel it is important to assist those protecting
> the environment over ground truth mapping.
>
>  On lord Howe Island, currently over 70% of the island is off-limits for
> an outbreak of Myrtle Rust with the Island Board stating "The rust has the
> potential to change the way our mountains and forest looks, it may alter
> food webs and ecology, and potentially affect world heritage values,". In
> Western Australia, there is Phytophthora (dieback), now prevalent in the
> Stirling Ranges which is mainly carried long distances by human activity.
> In these and other more local instances,we should endeavour to assist
> protection.
>
> I feel the  lifecycle prefixes and access=no in most instances however it
> might be better to remove all highway tagging other than a note to protect
> fragile ecology so that no downstream map accidentally maps these.
>



On Sun, 8 Oct 2023 at 22:57, Ben Ritter <benjaminaritter at gmail.com> wrote:

> I think we can assist environmental maintenance without compromising the
> ground truth value. They are not actually in conflict with each other.
>

Exactly this. If we map the closure including the reason for the closure,
we can help inform park users about which areas to avoid and why they are
asked to avoid those areas. People are going to still see the path on the
Strava heatmap or they are still going to find it on the ground anyway.


>
> In fact, I think it is *more helpful* to keep the highway features with
> the addition of the access tag and/or the lifecycle prefix.
>
> Many OSM users are used to incomplete data, so if they saw an OSM map
> which didn't include tracks that they observe in the wild, they would
> likely assume the data is missing, not that there is a restriction on it.
>

Good point, we see this already with Overture maps which conflates OSM
buildings with AI generated buildings. I can see in the future map
providers might conflate OSM highway=* network with probe data like Strava,
I'm not saying we need to map all the negative space too but for paths
which may still get activity it may help to map these in OSM so that a
conflation won't pick up on it being missing in OSM.


>
> With the aim of ensuring as many maps as possible indicate the closure,
> the existing lifecycle tag should be used, which is
> https://wiki.openstreetmap.org/wiki/Key:disused:highway, instead of a new
> one.
>
> Anyone publishing maps using OSM data while ignoring the access tag is
> being reckless, and should stop it. Deleting those features is not a
> solution in any specific case (this thread is case in point), or in the
> long term for the reasons above.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-au/attachments/20231009/62995dc3/attachment.htm>