Yeah, the tiny little piece had it set, so that's where I took the 30 from. I just assumed it hadn't been mapped and I was too lazy to check Mapillary.

If the pedestrian street counts as an intersection here, then the sign would indeed apply just to the tiny section before it. But then it should be mapped as maxspeed:forward/backward only in that direction until the intersection, since the signs aren't "symmetric".

I see there are still many roads marked as tracks causing bad routing, so you haven't fixed them yet.

I tried going to that path from West, but I didn't see a bridge and I couldn't even find where the path is. The stream is some 5 meters wide, so it should have been a bridge big enough to see.

If there was a bridge, I think it is gone and the path is no longer used. I see Strava foot heatmap around it, but the GPS is kinda bad there and I am not sure if and where people actually crossed (or possibly just turned around).

Hi! For path way/941358486 , do you remember if there was a bridge over Vēršupīte?

Sveiki! Vai tourism=attraction tiešām ir nosaukums "OGRE"? Dabā šeit nekādu uzrakstu nav. Tūrisma info arī nekur tāds nav norādīts (piem. https://www.visitogre.lv/lv/kulturvesture-un-arhitektura/547/mols-pie-ogres-ietekas-daugava)

Yeah, an error, I am not sure how that got added, I am 80% sure I was just selecting the concrete plate surface value in iD. Fixing in changeset/105113958

Izlaboju

changeset/104261840

I don't use JOSM much, so I don't know how it converts tracks into paths. I record tracks with OsmAnd too, but I've never converted them directly. I'm sure there must be some JOSM option when you are importing tracks?

If you didn't intend to add these elevation tags, I can probably remove them for you (probably in JOSM) since they don't seem accurate anyway. But I have no idea what you have to change in JOSM so it doesn't add them automatically.

Hi! The elevation data on points seems very wrong, as the terrain here is almost flat, but the values range widely from 14 meters to 30 meters. If this is from raw GPX data, then it is very inaccurate and I am not sure there is any point in keeping these values on points.

Is there a reason you added individual point elevation in this particular area?

Same for changeset/83663012 or changeset/83783494

Thanks, I thought it had something to do with the Soviet trampoline installation, just not sure what since I didn't recall anything at all being there.

Hi! What's the man-made feature node/6924839392 ?

I have now mapped all the intersections and various details plus some of the bigger features between the blocks. And I removed the original tags from the main road. There's still plenty to map, but the main routable ways should be up to date barring any mistakes.

Hey!

I realize the main street has the basic tags, but I intended to micromap the fuller details along the road. Almost none of the intersections have crossings explicitly marked for pedestrian routing or accessibility. A lot of them have tactile paving and flush curbs now. Cycle lanes have separate marked crossings. There are no stop/traffic light locations mapped. Public transport platforms sort of overlap with cycle lanes now and there's a give way warning. Roadside parking where it's still possible is now a bit different and adjacent to the cycle lanes. (There are obviously all the service roads and driveways, none of which are mapped yet, but which all cross the cycle lane. Not to mention all the shop and building entrances, which are a long way from being added.) Sidewalks also have different surfaces now, mostly concrete mini-plates and not just asphalt throughout.

As for original tags, I haven't removed them because I haven't transferred and confirmed all the values (yet). For example, the above-mentioned asphalt value is outdated.

Yeah, during the Soviet era, these were legally classified as allotments (as in дачний кооператив and similar). They were locally named/organized as "dārzkopības sabiedrība" and similar. But since the dissolution of USSR, they were all reclassified as part of the closest village, as its own "vasarnīcu ciems" or just individual dwellings and all given a proper address. None of them are legally allotments anymore, since there is no such concept in the law in Latvia. Each plot is private land and there are no requirements to cultivate the land (like, for example, German allotments). Colloquially, people still call them дача/огород or whatever, but that's mostly just jargon and no one actually has to keep gardens or produce food. No one can evict them for failing to cultivate the land. I find that most OSM uses of allotments are just editor misunderstanding about what allotments means, especially due to historical use, and because, as you say, for lack of "a better tag".

All that said, some of these plots are still arguably used that way and the produce sold at the farmer's markets, so we could map it that way. But the problem is that being allotments implies a large area of this use, where everyone is participating, not just individual plots. Personally, I have biked through dozens of these "allotments" and they are all just private land with old and new houses and gardens in various states of use. Yes, there are fruit trees and greenhouses and vegetable beds, but I could not definitely state that any of them are used for actual community-focused food production. It all looks personal use.

I'm not going to change this, since you surveyed it. But I would just tag them as residential areas and that's it. I have retagged a bunch of these after surveys myself. I think allotments is more incorrect than residential is imprecise. I have looked into this before and there was no better tag that I could find. Since OSM maps the present, there isn't any sort of "historical allotments" tag.

Are any of these actually allotments, i.e. land use primarily for individually cultivating gardens? Latvia has almost no real allotments left in terms of OSM. 99% of these are just residential areas and dacha villages (vasarnīcu ciemi).

(I'll reply in English, since that makes it accessible to more people.)

Yeah, these roads have the "braukt aizliegts" signs (https://likumi.lv/ta/id/274865#p259 302.), so only residents/deliveries. There's a bunch of places in Saulkrasti like this, which seems silly. I guess I should tag them foot=yes?

There are repair works happening there right now, so I might re-survey later in case they change the signs, but I doubt they will.

That must have been a glitch with the iD preview version that I did not notice. I was probably moving the whole building to align. Or something else weird happened. Nothing else was different in the workflow besides the iD being in preview version, which I have used a bunch and (as far as I know) without errors like this. It was before they validated stuff like building overlaps.

What does "incompatible source" mean? Was there a licensing issue? It looks like this reverted these edits: changeset/47219776 and this deleted a bunch of buildings, leaving lots of nodes without ways (since the nodes were re-aligned and not reverted, I guess). Can we restore the edits or do we need to redo them?

I can see from your links that "Hamlet Xx" is the official property name and first part of the official address.

But I am not sure I agree that we should map it this way. On the Ground Rule says "whatever [..] used by the people on the ground", which is what my previous reply was about -- on the ground, you see house plates with numbers, prominent numbers with extra info, and almost never "Hamlet Xxx". I would argue that, intuitively, anyone looking for addresses would look for the numbered part. After all, that's what's actually on the ground and that's what local hamlet maps are like.[1] (In fact, some have an additional local street name and number, which isn't official and doesn't even appear in kadastrs.)

The Addresses guideline about addresses with no street names mentions "in small villages the address may be just the village name and the name or number of the house".[2] It seems a reasonable way to map it for rendering and routing software.

I can say from surveying and looking at the map that having house names for each address is illegible and numbers are much more readable. Many names don't even render, because there's no space for them. Often they overlap the streets and nearby features.

I guess the question is how strictly OSM should follow the official name. I don't mind listing it in addr2: (though I don't think there's a way to mark address as "official" versus "on the ground"). But I'm not sure we should list the property names as the first addr:. After all, so many features on the ground do not match what the official sources claim.

[1] https://imgur.com/a/rhrgL
[2] osm.wiki/Addresses#Addresses_without_street_names

Hello,
I was looking into this as I was trying to determine how to fill out the addresses in nearby hamlets. I went over to look at what they actually are in reality (see below). The best I can tell is that those are not real house names. Something like "Rožkalni"[1] is -- you can see it does not have either street or hamlet number, while surrounding houses do. Otherwise, they would be in "Vietvārdu datubāze"[2]. My experience has been that kadastrs.lv mangles addresses due to their own system shortcomings when they cannot organize them correctly for whatever reason and have to resort to novel naming schemes (my own property is like that where actual address is 10-5, but kadastrs lists it as 10D, though nobody, even utility contracts, list it that way).

In case of Silmala 223, the hamlet/village is "Silmala" and the house number is 223 without a street name. If it had a street number, it would likely be Rīgas iela 11 or something, but you wouldn't say "Rīgas iela 11" is the house name. With that "Rožkalni" example, you can see the nearby houses have both the street number and the hamlet number.

If we go by full modern property street plates, they (not that I checked every one) have a prominent number with additionally both hamlet name and street name listed.[3] Since OSM--in theory--goes by what we see on the ground, then we would expect the plate to say "Silmala 143" as a single name if that was the name. In fact, 223 is just a number plate[4] (though I realize we are only using it as example here). I find that this is the case for the properties in this and other nearby hamlets, like nearby Banga, Selga, etc. In fact, most have just a number[4][5] and some don't have anything visible.

[1] changeset/51827482#map=18/57.25089/24.45043
[2] https://vietvardi.lgia.gov.lv
[3] https://imgur.com/a/lOX5I
[4] https://imgur.com/ajSlhlB
[5] https://imgur.com/ARDZP19
[6] https://imgur.com/vYVZl3W