People walking there does not make this a path. This is a railroad and not an abandoned one (though I remember it having a lot more frequent traffic ~20-25 years ago)

https://i2.wp.com/viimsiteataja.ee/wp-content/uploads/2021/11/IMG_1317-copy-scaled.jpg?fit=2560%2C1920&ssl=1

http://milstrand.ee/en/

I think you can connect the path where it ends directly to the railroad to hint that people continue walking on it, but you shouldn't misguide users by drawing paths which don't exist.

Hello, you've added a path next to the railroad, it's even so close that it goes on top of railroad ties. Are you sure it exists? I don't see it on aerials.

Please note that if someone is using a railroad to walk it doesn't make it a path.

Hello,
those roads you've added on T-junctions don't really exist. You shoudn't add a separate highway just because the junction doesn't have perfectly square corners. The convention is to only add a separate way if there is some kind of physical separation or atleast road markings which separate carrigeways.

* to the north of the bridge

I don't think there is a sidewalk to the bridge.

Hello, I think you've made some mistakes.

1)I see you've marked many ways as bicycle=yes. Wiki says bicycle=yes "bicycle=yes is applied where cycling is legal, not obvious from other tags and where it is not explicitly designated for cyclists (that would be bicycle=designated)."
Estonian law https://www.riigiteataja.ee/akt/103012017018?leiaKehtiv says it's illegal to cycle on a footway, unless you are a small child.

bicycle=yes is rarely used in Estonia in cities, since cycleways and combined pedestrian-bicycle paths are tagged with bicycle=designated and actual footways, which are not marked by signs 431, 433, 434, 435 (http://www.liiklustestid.ee/liiklusmargid/kohustusmargid/), are implicitly bicycle=no.

2) way/288898677 This way is definitely not a cycleway, it's a forest path.
This way should be changed to a highway=path for the reasons listed in the first point. Usually forest paths are not tagged with a bicycle=yes or any other bicycle=* tags, since it may require some skill to ride it and only the rider can judge for himself.

Also tag combination like highway=cycleway and bicycle=yes is very unusual. When you edit the map in ID Editor, you'll see that when selecting a cycleway, it already shows bicycle=designated as a default greyed-out value. And it's logical, since cycleways are ways where cycling is designated, there is no need to repeat that. Having a highway=cycleway and bicycle=yes raises some questions why is it a cycleway, but cycling is NOT designated?

I've also noticed that here
changeset/117336964
there is no reason to turn the tunnel into a cycleway (it has stairs, it's not legally a cycleway and it can be dangerous in a narrow tunnel with limited visibility)

Please be careful next time, consult OSM wiki and local law. And it would be nice if you fixed this.

Why did you add a service=driveway to Pilvetee? It's a small street which gives access to other smaller roads. The driveway value is for smaller roads, which are usually not streets and usually connecting to one building. I'm not even sure highway=service should be used.

If you meant to tag a street with a living street sign, then highway=living_street should have been used.

1) operator=Selver
Cadastre map shows that this parking is outside of Selver's territory. Don't add values just by assumption, it's ok to leave fields empty.
2) parking=surface - street_side may fit better parking=street_side
3) Street-side imagery suggests that parking starts earlier than mapped.

Priisle selver is under construction. You can add "construction:" prefix to fix it.
construction=*

>I think the general conclusion from that and the way things are is that both scrub/grass scheme and simple man-made=clearcut tag are insufficient for given purpose.

There is no "general conclusion". On the same man_made=clearcut there is lot of arguments for it. I think that it's the best approach given the tags we have today, while obviously not ideal.

*obviosly meant "not implying" in the first sentence.

I'm implying that opinions have no value or that you have no point. I'm saying that opinions are DIFFERENT. And as long as we don't have a stricter tagging scheme in OSM, they should coexist and allow a bit of leeway.

Clear-cut in a landuse=forest filled with "scrub" can be interpreted as a (very) young forest area. You may not like it, you may have points against it, but today that's how it is.

The solution may be that you stop arguing with one person and use your energy to help define a better standard.

This is going nowhere. Don't expect me to reply anymore. I understand your position, you aren't attempting to understand mine.

I'm sorry if I made you feel bad.

>I'd also use common sense, but I don't get why you'd consider forest of about 1 m tall trees against common sense.

I'd consider removing scrub against common sense.

Everything else you wrote is again opinions, which can be different. Estonian goverment does not govern OSM and does not define which tags should be used.

Sadly, it would be hard for us to create a big enough discussion in Estonian OSM community to accept some local standard.

>I'd say a clear-cut rather falls in the temporary category as it's expected to develop into something else quite shortly, in a few years time.

This is your peronal opinion which has nothing to do with best practices you mentioned.

>Small forest surrounded and managed grasslands exist, but I don't think anyone on the other hand would really consider a clear-cut filled with stumps a grassland. Most of given clear-cuts here also are no longer dominated by grasses. Due to sparse canopy and bare crowns, it may not show well on an orthophoto, but these "grasslands" are filled with (small) trees.
>Note that man_made=clearcut tag exists. It's not rendered on osm.org and it could be used to sort of work around grassland-scrub.

I'll add man_made=clearcut you mentioned. But this tag is not exclusive to the vegetation. And the areas with small trees which are mistagged can be mapped as scrubs like it's mentioned on the man_made=clearcut wiki page.

>how does a mapper know at all that it's no longer something less than a forest and that area can be removed?

There is no clear standard on this. I'd wait until the trees are at least 3m in height. Someone else may have a different opinion and I won't be against it as long as he uses common sence and doesn't consider 1m trees grown. As with many things on OSM, common sence should be used and different approaches to tagging may exist until a stricter and more exact standard is implemented. Which highway tag value should be used for a combined foot and cycleway - footway, path or a cycleway? Can a small group of trees be tagged as natural=wood or landcover=trees? How sparse can they grow until they stop becoming one? Can you precisely answer these questions without bringing your PERSONAL opinion? Probably not.

Overall, I get your point and I'm not against having just a landuse=forest with man_made=clearcut, but I also don't agree that it's the only way and think that clearcuts can be descibed in more detail.

>I wouldn't say that any one these areas is a grassland only because grasses were dominant for a couple of years.

But this is what a grassland is...

>As for scrub, I realize that OSM mappers interpret it very differently, but it'd probably be more intuitive to consider it as a habitat where shrubs are more or less persistent (not very common in our climate, but juniper shrubbery might be a good example), rather than just young forest.
>based on what criteria would you consider a young forest simply a forest again then?

I agree that definition of a scrub is a bit fuzzy, but it is sometimes used for any small trees. I think it's ok to use it here - trees there are very distinct from the surrounding forest.
natural=scrub
The only reference to height you can find here is "less than 5 meters tall" here. I'm not really strict about (as with many things in OSM, it's hard to objectively tag) BTW, notice that the next sentence on linked page mentions young trees.

> Then again such clear-cut areas are expected to be temporary and it generally isn't considered a good practice to map temporary features.

Buildings are temporary too. Trees don't grow overnight, Considering that it's a place not far from the capital, it's unlikely to be abandoned by mappers and get too outdated. What this good practice means is "Don't map cars, festivals or weekly markets".

> Also, a young forest by most definitions in common use is also a forest.

I'm not mapping landuse with grassland and scrub, I'm mapping what's there in addition to landuse.

I agree that landuse should be fixed, thanks for pointing that out. I'll still map these areas separately, but they will be under landuse=forest

Some time ago you asked opinion about your imports. Well, I've noticed that you've put e-cigarette shop INSIDE McDonalds and a pharmacy in an empty field. :(

Clearly could use some basic manual validation.

Hello, I recommend you to use aerial imagery from Maa-amet. It has a lot higher resolution and accuracy than Bing, especially cadastre map. These parking lots could have been a lot more accurate. (While this changeset is old, I see that you are continuing to use Bing)

Good job!

BTW,

bicycle=designated
foot=designated
highway=cycleway

is not in any way different from

bicycle=designated
foot=designated
highway=path

but may be incorrectly interpreted by some data consumers, such as cyclosm.org

we can also mark it as

bicycle=designated
highway=footway

I find that path is a good middle ground for combined pedestrian/bicycle roads.
See also osm.wiki/Path_controversy

Thank you for participation.

The crossing you've added is not a bicycle crossing, but a usual one, so there is no sense in tagging it as a bicycle crossing.

Also, the nodes were not connected to paths. While it may look ok on a map, a routing software would not be able to find a route if ways are not connected.

I've fixed these problems.