Update: I removed unsupported natural=peak nodes (slope points, ridgeline clusters without distinct apex) based on the discussion above. See changeset/181079175

I removed unsupported natural=peak nodes (slope points, ridgeline clusters without distinct apex) based on the discussion above. See changeset/181078710

Hi Peetang618,

I noticed that this changeset adds a large number of natural=peak nodes, including features that appear to be derived from DEM/SRTM local maxima (e.g. points along slopes or ridgelines without a distinct summit).

I’ve started a discussion with proposed guidelines for mapping peaks in Thailand, along with a planned cleanup of this data:
https://community.openstreetmap.org/t/natural-peak-mapping-guidelines-for-thailand-proposal-cleanup-notice/142734

The goal is to align on clearer mapping practices and improve overall data quality.

Your input would be very welcome.

Thanks!

Hi,

I noticed that this changeset adds a large number of natural=peak nodes, including features that appear to be derived from DEM/SRTM local maxima (e.g. points along slopes or ridgelines without a distinct summit).

I’ve started a discussion with proposed guidelines for mapping peaks in Thailand, along with a planned cleanup of this data:
https://community.openstreetmap.org/t/natural-peak-mapping-guidelines-for-thailand-proposal-cleanup-notice/142734

The goal is to align on clearer mapping practices and improve overall data quality.

Your input would be very welcome.

Thanks!

Hi Stef,
Thank you for taking the time to explain your background and methodology, I really appreciate the detailed response.
It’s clear you have long-term regional experience, and your involvement with humanitarian work and HOT mapping is commendable. That context helps a lot. My concern was not about your intentions, but about data reliability and the specific sensitivity of this border stretch.
From imagery alone, especially in forested terrain, it can be very difficult to distinguish between:
* temporary logging tracks
* erosion scars
* historical trails no longer usable
* patrol paths not meant for public access
In this context, adding routable highway=* features, even if visible, can unintentionally create real-world navigation risks.
I’m not advocating for a “fog of war,” nor for leaving areas blank indefinitely. But for sensitive border regions, especially where routing engines may direct people, we need to apply a higher verification threshold than in standard HOT-style building tracing.
Regarding your suggestion of tagging access=no: that could be one compromise, but it still assumes the physical existence and permanence of the feature. If uncertainty remains about usability or continuity, sometimes it’s safer to avoid mapping as routable infrastructure until confirmed on the ground.
Thanks again for engaging constructively. And if I’m ever in Siem Reap, I’ll take you up on that coffee 🙂

Thanks for your contributions, StefDeGreef.

Could you clarify the source and level of local verification for the paths and tracks you added near the Thailand–Cambodia border? Several of these appear to be traced from satellite imagery alone, without on-the-ground confirmation. This is generally not recommended, as imagery can be outdated or misaligned, and routing based on non-existent or incorrect paths can cause real issues.

Additionally, this is a sensitive border area with a recent conflict history. Adding or modifying potential trails in such locations without local knowledge or verification feels particularly risky and inappropriate.

Please let me know how these edits were verified. If I don’t hear back, I’ll likely proceed with cleaning up the data on the Thailand side.

Thanks for your understanding.

osm.wiki/How_We_Map

Thanks for your contributions, Waldhüter.

Could you please share the URL to the HOT OSM task you’re working from? A number of the paths and tracks you added appear to be traced from imagery alone, without on-the-ground verification, which generally isn’t recommended. Imagery can be outdated or misaligned, and agricultural tracks in particular can be seasonal or disappear entirely.

It would be unfortunate to route someone onto a path that doesn’t actually exist. Could you clarify your sources or local knowledge for these edits? If I don’t hear back, I’ll likely clean up the data on the Thailand side.

Thanks for understanding.

Thank you for your contributions! Please remember to upload your traces publicly to OSM so they appear in the OSM Public GPS Traces layer, enabling verification and future alignment or improvements. Thank you!
---

Published using OSMCha: https://osmcha.org/changesets/175995891

Thanks a lot for your contributions! Just a friendly tip: try to keep your changes smaller, like focusing on one country at a time.
osm.wiki/Changeset#Geographical_size_of_changesets

Thanks a lot for your contributions! Just a friendly tip: try to keep your changes smaller, like focusing on one country at a time.
osm.wiki/Changeset#Geographical_size_of_changesets

Thanks a lot for your contributions! Just a friendly tip: try to keep your changes smaller, like focusing on one country at a time.
Also adding a good changeset description will help a lot others understand what you changed.

osm.wiki/Changeset#Geographical_size_of_changesets
osm.wiki/Good_changeset_comments

Thanks a lot for your contributions! Just a friendly tip: try to keep your changes smaller, like focusing on one country at a time.
osm.wiki/Changeset#Geographical_size_of_changesets

reverted for conflicting classification changes and lack of response from original mapper

changeset/174799745

i GrabTH_Thamolvan, thanks for your contributions. Could you please share what source or wiki definition you used to justify these classification changes? These roads belong to an Agriculture and Development Center. Best, Julien

Thanks for your contributions! While you added some new paths, many existing forestry double tracks were downgraded to “path.” These tracks are regularly used by 4WDs and ATVs, and although they can become temporarily impassable during the rainy season, as a rule, any trails 2 m wide or more is considered a track, regardless of its condition. I will revert your track > path changes. Thanks for your understanding. Julien
---

Published using OSMCha: https://osmcha.org/changesets/173563873

Please follow this link and make sure Visibility is set to Public. Happy mapping!

osm.org/traces/new

Thanks for your contributions! Please upload your GPX traces publicly so they can be added to the OSM GPS layer and help improve map accuracy.

Thanks a lot for your contributions! Just a friendly tip: try to keep your changes smaller, like focusing on one country at a time.
osm.wiki/Changeset#Geographical_size_of_changesets

Hi NateeT, earlier this year I removed some path segments I had added based only on Strava or public GPS traces, without ground verification, after receiving reports about non-existent, abandoned, or sensitive sections. Please feel free to re-add any missing segments you have firsthand knowledge of — their trace should still be visible in the public traces layer in iD. If not, just PM me the location and I can look them up on Strava.

Thanks a lot for your contributions! Just a friendly tip: try to keep your changes smaller, like focusing on one country at a time.
osm.wiki/Changeset#Geographical_size_of_changesets