Right. "park", in OSM speak, is more of an urban park like Boston Common.

OSM tagging has a long history, influenced by its British origins, and is messy and not always semantically self-consistent. In the US, as you know, was have quite the diverse array of "public lands": from city parks, to "private" non-profit land-trust conservation land in Lincoln, to Audubon, to state forests, etc. There has yet to be a fully articulated "right" way of tagging "public lands" in the US.

Here is a start:
osm.wiki/Proposed_features/Park_boundary

In New England, most non-manicured "parks" are tagged with boundary=protected_area and leisure=nature_reserve.

I updated Cold Spring park. I noticed its boundary was a bit out of alignment compared to the MassGIS parcel data.

In any event, a "natural=wood" should not be applied to the geometry that is defining the cadastre. "Land cover" should get its own geometry...and I've mostly stopped doing woods land cover in MA because almost everything is covered with trees anyways!

Check out parks/conservation land elsewhere in Belmont/Linclon/Concord etc for some examples. I've "been" just about everywhere in the state trying to make this all more consistent. It's an exercise in OCD compulsion...however I've discovered so many new places through my mapping to visit in "real life". It's been very satisfying.

To learn more, I recommend perusing the OSM wiki. Just keep in mind it's not curated by one person or formal committee, so you'll see some inconstancies stemming from this being an international project of volunteers with different viewpoints on how things are done in their part of the world.

I'm happy to chat more too. I'm active on the OSM Slack instance.

Happy mapping.

Hello. Please refer to my email on the other changeset. The same feedback applies to this changeset.

Hello. Welcome to OpenStreetMap.

Please explain why you removed "park" from a couple of parcels and added "natural=wood".

Generally speaking, land cover tagging should not go on the same ways that are used to define a parcel boundary.

These parks do need some tagging clean up, but your changes IMHO are not correct. Please let me know your thoughts based on this feedback. Thanks.

Congrats!
---

Published using OSMCha: https://osmcha.org/changesets/116475694

Welcome Mass Audubon! I've edited many of your properties throughout the state. Please let me know if you have any questions. Cheers.

Hello. Thanks for the contribution. In the future, please consider creating multiple changesets over a much smaller geographic area. It makes it easier to review and provide feedback if needed. Happy mapping...

Hello. Thanks for the contribution. Please consider in the future creating multiple changesets over a smaller geographic area. It makes it easier to review and provide feedback if needed. Happy mapping...

It's customary, and polite, to reach out and wait a day or two before reverting a changeset that isn't obviously vandalism. txemt seems to be engaged and very responsive. Why risk making a motivated mapper feel disenfranchised?

"Broken data can be fixed easily, but a broken community is not so easy to restore."

osm.wiki/JOSM/Plugins/Reverter

Hello. This is not the proper way to map in OpenStreetMap. The "name" is not for instructions like "dogs must be leashed in this area". And these are not proper "nature_reserve" as they are already within the park feature. I urge you to take a look at other similar parks to get a feel for how things are mapped in OSM. Please let me know if you have any specific questions. Thanks.

Leftover parcel boundaries before combination for overall property boundary. I just removed it. Thanks.

An oversight. Fixed. Thanks!

That level of micro-mapping is not usually done as there are trees everywhere around here. I'm not one to prevent someone from mapping what they want but perhaps there are other higher priority things that need attention? Cheers.

Hello. I see you "review requested" this changeset. Though not technically "wrong", what's the motivation for mapping individual trees? There are a LOT of them in this cemetery. It's inconsistent with mapping practice nearby. Thanks.

Hello again. My main point on the other changeset discussion is that any landuse/landcover tagging, in this case natural=wood, should NOT go on the way(s)/mulitpolgon(s) that defines the property boundary. You just switched tags on the same multipolygons. If you wish to do landcover mapping, you should create a NEW multipolgon relation for that. This new multipolgon relation can recycle the same ways from the parcel multipolygon relation. In the examples I gave earlier, notice that there are two separate multipolygon relations for those properties: 1 for cadastre, and 1 for land cover.

Makes sense?

BTW, strictly speaking, these parcels are not even all "woods". Per MassGIS data, there's a good amount of wetland areas presently not mapped.

https://massgis.maps.arcgis.com/apps/mapviewer/index.html?layers=4c940b24f05f4de9bdb9a467e555aaf8

IMHO, a blanket natural=wood for all those parcels would not be completely correct. It's this sort of tedium that has led me to personally deprioritize doing natural=wood land cover mapping.

Thanks.

This is a bit outdated but still a good read:
osm.wiki/Massachusetts/Conservation

And:
osm.wiki/United_States/Public_lands

Generally speaking, you should NOT place a landuse/landcover tag on the same closed way or multipolygon relation that defines the cadastre or property boundary. These boundaries are virtual as they exist in a GIS tax map. (of course, sometimes the property boundary is a physical feature like a stone wall or river edge, but...) Whereas, landcover/landuse is directly related to what's on the ground. Ideally, OSM would have a different layer each for cadastre, land cover, etc but alas it's all just one jumble.

Having said that, I used to do detailed woods mapping that was bounded to the "public land" property boundary. See here for an example:
osm.org/relation/11873101

I have stopped doing this because I think it's just "tagging for the renderer" because I wanted the area to show up darker green on the standard OSM carto map. Furthermore, for me, woods mapping in New England is not the highest priority as almost everything is covered with "woods" nominally.

If you felt compelled to land cover or land use map, please create a new multipolgon. This will allow you to "cut out" ponds, parking lots, swamps, etc that are not covered with "woods"...and have this "cutting out" not affect the definition of the property boundary.

Another example of this, which I did, is the Middlesex Fells:
osm.org/relation/11875172

I'm happy to chat more about this or other mapping topics. I'm active on the OSM Slack channel if you're so inclined. https://slack.openstreetmap.us/

Happy mapping...

Hello. Per the wiki, and local mapping convention, landuse=forest is for parcels that are "Managed woodland or tree plantation maintained by human to obtain forest products". I'm fairly certain this land is not regularly cut for timber and thus this tagging does not apply. Please let me know your thoughts on this feedback. Thanks.

Hello. Can you explain what you're getting at with this changeset? This is a LOT of added nodes. Thanks.

Looks to be a mistake. Fixed. Thanks.

Hello. This parcel looks completely wooded. What's your reasoning for removing nature_reserve and switching to a recreation_ground? Thanks.

osm.wiki/Tag:landuse%3Drecreation_ground
---
#REVIEWED_BAD #OSMCHA
Published using OSMCha: https://osmcha.org/changesets/115280174

Sorry about what sounds like a bad experience. I did a bit of investigation. These tracks were part of a TIGER roads import. Looking at a 1960 USGS quad, it looks like this is the old route for the AT. However, it's right close to some other tracks that show in lidar data so I think just bad or redundant data. I deleted out:
osm.org/changeset/115278986