Hello. Generally speaking, landcover tags like natural=wood should NOT go on ways and/or multipolygons that define cadstre or property boundaries. The "woods" don't start and end at the property boundary. If you wish to do landcover mapping, it should be defined as its own way that is algined to where the tress actually ends. Sounds reasonable? Thanks.

Mashin, can you point to any other mappers in CT or elsewhere in New England/NY that are actively removing accurate names from paths/footways to support your assertion: "quite common way of mapping hiking routes". I have been involved in multiple threads over multiple years of people complaining about your practice of removing names. I map a lot in MA/VT/NH and see no one else doing this. CT is the land of nameless trails on Strava/AllTrails/Gaia etc.

I completely understand (and technically agree with) your argument and wish we could just rely on name in the route relation. Unfortunately, OSM is far from having a perfect data model schema nor are the popular downstream data users completely adequate in their data parsing. For the sake of real-world practicality, and to adhere to the OSM tenet of "community cohesion over data perfection", I request you abandon your practice of removing valid names from path/footway ways. Thank you.

Hello. Thanks for pointing this out. I had mapped a lot of this a few years ago when I was a less experienced mapper. Some insight:

Indeed, the previous multipolygon for "Royalston State Forest" included both DCR and private CR parcels. On the MassGIS OpenSpace data for the private CRs, there is this piece of metadata: FORMAL_SITE_NAME = Royalston State Forest. I erroneously assumed the private CRs were also known as "Royalston State Forest". I didn't realize there are signs that identify the private CRs as separate things.

However, the multiple adjacent private CRs are not all known as "Beals & Fleet National Bank CR". Your recent edit labeled all as such because I previously mapped the private CRs as a single multipolgon. Before, there was a multipolgon for the DCR parcels and a single multipolygon for the private CRs. And then a bigger multipolygon that had all the above parcels named "Royalston State Forest". Could be confusing!...especially if you're a new mapper.

I just redefined things so "Royalston State Forest" is JUST the DCR owned parcels. I will go back soon to redefine the smaller private CRs and name them per what I see in MassGIS OpenSpace layer.

Please let me know if you have any on-the-ground info that may differ from the MassGIS OpenSpace data...I find it's sometimes not up-to-date. I'm happy to collaborate to ensure this beautiful part of the state is properly mapped. Cheers.

Hello. What supporting data do you have to call this road access=private? The town's tax map shows there is a recreational easement on this.

https://www.hartford-vt.org/DocumentCenter/View/8849/Tax-Map---13-PDF

Please provide some more information.

Thanks.

Hello. There is ongoing debate about the use of footway versus path. Most downstream renders will treat it the same. However, there is some local consensus that highway=path should be used for "hiking trails".

osm.wiki/Massachusetts/Conservation

Footways are used in more urban areas. (It's a British-ism).

Happy mapping...

Hello. I see you made one of the ways a cycleway, but then added a bicycle=permissive. Was that intentional? Seems "cycleway" would mean bikes are most welcome there. Thanks for mapping.

Hello. The natural=wood should not go on the way or multipolygon that defines the property boundary. Land cover mapping should be done with separate ways as the trees usually don't automatically stop at the invisible boundary. Also, the multipolygon for the state forest has cutouts for the lakes and parking lots. Presumably, this was done so trees would not render over them. This is incorrect because the water bodies and parking lots ARE part of the property so should not be "cut out". Makes sense? Thanks for mapping.

I appreciate the difficulty of large polygons. However, I disagree it's "necessary". Ideally, OSM would have layers where the parcel boundaries couldn't "talk" to the land cover layer. But it doesn't, so we have to manually enforce some mapping best practices. Philosophically, where the trees end and a parcel boundary demarcation has nothing to do with each other. What sometimes happens is a future mapper will want to adjust the trees but they inadvertently also adjust the parcel boundary because of these connections

Please stop attaching natural=wood to existing parcel boundary ways. This is an incorrect way of doing landcover mapping and creates headaches. I reached out previously in another changeset and have yet to hear back. Please respond. Thanks.

Hello. Why are you aligning natural=wood to cadastre boundaries e.g. the closed way that indicates the State Forest? The "woods" do not begin/start at parcel boundaries. It creates lots of headaches.

I think the "access=no" was added by mistake? Else, that overrides the bicycle=yes, etc.

Good to see you back on here. I see you had review requested checked. The new way you added was unconnected to at one end, and where it intersected another existing way, it didn't connect there also. No biggie. Cheers.

Hello. I see you've created some new "paths" that are unconnected to other ways. I see some evidence on lidar, but there is nothing on heatmap. It's highly unusual for unconnected ways to exist in the middle of the forest. Do you have any additional data to support the creation of these paths in OSM? e.g. that they are presently navigable or even apparent in-person. It's likely just old skid trails that are probably very overgrown. Thanks.

Hello. I just went ahead and reverted this changeset. I inspected the area and all those raw points and tracks are very non-standard way of doing this. I usually wait for a response to not make people feel disenfranchised...but I think there a clearly much better ways to accomplish what you're trying to do.

osm.org/changeset/146012609

Again, I (and I'm sure others) are very happy to assist. Please be in touch.

Cheers.

Looks like you're new to OSM so let me say: welcome!

Generally speaking, we do not upload raw GPS/GPX data without further smoothing as it looks like what you did. Uploading multiple single points is not really "correct". Furthermore, the tagging on these, and the other ways you added, does not conform to standard practice per the OSM wiki and local mapping norms. I'd recommend reverting this whole changeset. (I can help with that if you like).

This area already has a lot trails mapped. Looking at Strava Heatmap data and lidar, it DOES look like many existing mapped trails can be substantially improved. I think referencing these data sources will be much more fruitful than a single GPS trace. Strava data has many multiple traces aggregated so the "average" path is easier to discern.

I'm an experienced mapper and happy to collaborate with you on the mission of your Wayfinding Master Plan. Cheers.

I retraced this with lidar. Should be really accurate now. Cheers.

Thanks for the note. I was curious and looked at this region on lidar. I think it's just bad data from the original import. I trimmed significantly to just what I think is the driveway (however, I have no local knowledge). Cheers.

Hello. I'm mapping more in Death Valley lately. I'm curious about your switch from track to highway=unclassified on the Saline Valley Road. Are there any OSM Death Valley working group conventions or something like that?

On the NPS AGOL Roads dataset, Saline Valley Road, and roads around it like Hunter Mountain Road, have a NPS road class (RDCLASS) of 4WD. I propose that RDCLASS=4WD maps to highway=track in OSM and the NPS RDCLASS=local maps to highway=unclassified (like Racetrack Valley Road) for consistency. Thoughts on any of this? Thanks!

(BTW, I'm active on the OSM Slack if you're inclined to collab there.)

You're welcome. And some more detail: I think they've always been service, and I recently asked the community on Slack, and a few experienced mappers confirmed. That's what I meant by "discussion". Thanks.

Per typical OSM mapping norms, "roads" inside cemeteries should be mapped as "service" as they are not part of the regular road network. If you disagree, please provide more information to support your recent change. There has been OSM community discussion on this topic recently. Thanks.