Hello,

Unfortunately, this was not a valid fix for this detected issue. You cannot simply "extract" a part of the building to a different layer to avoid the error report. There isn't great Mapillary footage available here, but you can see this is a building passage through the same building https://www.mapillary.com/app/?pKey=108548951312564&focus=photo And if you enable the cadaster layer, you can see it's the same building. These are extremely common in Latvia.

Please take care not to guess at the solution if you aren't sure and instead skip it or mark it as "too hard" in MapRoulette. A local may be able to fix it.

Thanks

As I mentioned in changeset/140536183, this minor ditch is almost certainly not a river, please don't change it to such classification.

You specified that way/1203094652 and further is a tunnel and a pipe - are you sure this is correct? This stream/ditch is visible of LIDAR layer and it doesn't look like it's some sort of underground pipe. If you could explain what you meant, I could fix it and set the right tags.

Hello!

Can I ask where the name "plavūpe" is from for way/1203087391?

This is also almost certainly not a river. It could be considered a stream, but it's most likely just a ditch. I'm not sure how you are measuring the width, but it's definitely not 5 meters wide water.

Also, tidal=yes shouldn't be specified on any water bodies so far from the coast - none of these are affected by the tides.

The geometry of way/1203090258 looks incorrect - it's drawn almost on top of the road but the road here is 2 lanes wide.

I fixed some other issues with waterway relations. Notably, there are no tributaries or springs here.

I'll leave some additional comments on some other of your changes.

Thanks

P. S. I can reply in Latvian or Russian if English isn't the best.

Sveiki!

Vai varētu lūdzu precizēt, kas tieši bija domāts ar šo izmaiņu? Cik zinu, šajā posmā nav divvirzienu kustība un sabiedriskais transports to lieto tikai austrumu-rietumu virzienā (no Tilta uz Tvaika vai Ganību, kā tas arī minēts LSM rakstā).

Paldies

reverted as vandalism

changeset/140358537

Hello again,

Please check the comment I left on a previous changeset. Before continuing with many more changes, please do check your previous edits.

This changeset has another example where a new relation relation/15672280 does not include part of the road
way/129959284

Or when you create a relation like relation/15666998 and there is an obvious section in the middle named differently way/24257928 then you should be checking the cadaster data for the street layout and naming rather than creating a relation for incorrect data.

As mentioned, there have been multiple previous mistakes too. Just one example of a relation
relation/15672275 that didn't include segments like way/153204065 or way/4962589/

Sveiki!

Zīmējot pilnu kontūru parastām ietvēm un celiņiem kā piemēram way/1201188517 , šie "laukumi" nebūs
`highway=pedestrian` + `area=yes` bet gan `area:highway=footway`. Gadījumos, ja tie ir plati un kustība ir visos virzienos (nevis galvenokārt sakrīt ar pašu ceļu) būs `highway=footway` + `area=yes`. Līdzīgi ir visiem `highway` variantiem. Bet tāda lieta kā `highway=pedestrian` + `area=yes` ir speciāli veidoti laukumi, kur parasti ir visādi soliņi utt. - tādi nekad nebūs tur, kur parastas ietves. Vēl pieminēšu, ka te ne visur ir `footway` - piemēram way/889204955 posms ir `cycleway` un tāpēc arī kontūra būs `area:highway=cycleway`. Un vēl kontūras principā nevar pastāvēt pašas par sevi bez iezīmēta ceļa pa vidu, kā piemēram way/1201188518/ - šeit jābūt `highway=footway` līnijai maršrutēšanai un savienojumiem.

area:highway=*#Differentiation_area:highway_vs._area=yes_on_a_highway

Hello,

Can you please provide more descriptive changeset comments in the future when making so many changes to road relations? osm.wiki/Good_changeset_comments

Also, please edit more carefully. There have been multiple imprecise or incorrect relations you have created over time.

For example, this one relation/16227717 has literally one tiny segment in it.

Another example, where this segment was included way/25867387 yet this one way/706072588 wasn't. This leads to a question of how are you selecting the segments for inclusion? It seems you may be creating these relations by searching the road name rather than actually looking at the map or cadaster data.

In relation/16227714 you seems to have "missed" several segments of the road.

There is also an open question of having road routes that get "split" in an intersection like relation/15686391 . For road routing purposes, these obviously are part of the same route even if the intersection segment in between is a different road.

Thanks

I almost always put `surface=compacted` on these. Here, there are only rare loose rocks. Personally, if one can bike there without rattling and sliding on rocks, then it's most likely not gravel.

To elaborate, there are very few roads that have actual gravel on them, because cars very quickly compact, break and disperse the rocks. Such gravel roads last no more than a season before becoming like this example. In general, I wouldn't put `gravel` until most of the road is actually covered in rocks and they move as you step/bike/drive on them. Basically - rocks are touching each other and forming a layer.

Of course, this becomes difficult because many roads do indeed start with a gravel layer, but soon turn into what you see here. This is why a lot of mappers do put gravel everywhere if they see any rocks. As I mentioned - not completely incorrect, but also not that useful for data consumers.

To give you a more complex example, https://www.mapillary.com/app/?focus=photo&pKey=177516025291859 This road was repaired by adding a layer of rocks in potholes and the traffic here isn't high enough to turn it into uniform road. Is this gravel? Who knows :) I would probably agree that `gravel` here is okay despite only being in patches and I think most mappers would indeed use `gravel` here. But this will eventually turn into `compacted` unless regularly maintained.

I think I have only seen "full" gravel that lasted more than a summer may be a few times around Saulkrasti and it was someone just placing rocks in front of their property to avoid dust or water buildup or for temporary access like construction vehicles. For example, https://www.mapillary.com/app/?focus=photo&pKey=577720461150720 At the intersection, you can already see that rocks are getting crushed and turning into a more uniform mass.

Hi!

Just wanted to let you know that `surface=gravel` in OSM sense usually means "dense larger rocks". Roads like that in Latvia are actually very rare. Typical unpaved rural and remote roads (so-called "grunts" ceļi/"grantenes") are almost always `surface=compacted` - surface=compacted . It's not completely incorrect to tag it this way, but it does make surface values poorly usable for cases where it matters, like bicycles/wheelchairs/etc.

"Real" gravel roads look like this: https://www.mapillary.com/app/?pKey=795381835143910&focus=photo

Thanks

By the way, there is some recent Mapillary here from May https://www.mapillary.com/app/?pKey=1935502296801471&focus=photo . It doesn't let you see much beyond the immediate fences along the road, but I can confirm there's a new fence. Of course, per richlv, one is wire mesh and the other is wooden boards, so types would get quickly mixed if someone ever maps those. Wire mesh one is also offset deeper into the territory. And I don't know if the fence between the two properties is wooden or wire mesh - the wood one was built first after all.

Sveiki!

Ar šo izmaiņu ir izdzēsti vairāki adrešu punkti, piemēram node/2073584381 . Ja šie dati nav mainījušies kadastrā, tad adreses nevajadzētu dzēst. Arī nākamajā reizē, kad adrešu bots mainīs adreses, šie punkti tiks atjaunoti.

Sveiki!

Piebraucamos ceļus īpašumiem vajadzētu norādīt kā `highway=service` + `service=driveway` nevis `highway=track`, kas ir tikai piekļuvei laukiem/mežiem/zemēm.

service=driveway

osm.wiki/Lv:Latvian_tagging_guidelines#Piebraukšanas_ceļi_lauku_sētām

Ja ceļš no sākuma ir kopīgs uz vairākiem īpašumiem, tad var arī norādīt `highway=unclassified` tam posmam, piemēram way/1197003410 - šeit tas arī sakrīt ar kadastra līnijām.

Paldies

Hello!

Thanks for adding the crossing.

Just wanted to let you know that you connected it to the sidewalk that is above on the bridge. This created incorrect routing showing that you could get onto the bridge here.

I have fixed the issue and disconnected the ways.

Thanks.

Sveiki!

Ceļiem un takām nevajadzētu norādīt visus iespējamos tegus bicycle=yes,
foot=yes, horse=yes, motor_vehicle=yes utt. Šie tegi ir domāti galvenokārt norādīt, vai ar konkrētu transporlīdzekli/veidu ir atļauts pārvietoties, nevis vai tas ir iespējams. Citiem vārdiem - vai tas ir legāli. Ja nav ceļa zīmes, tad šīs vērtības jau tā pat is saprotamas ar =yes vērtību pēc noklusējuma. Un piemēram vietās kā way/629273053 nemaz nav atļauts publiski braukt ar transportlīdzekļiem. Piemēram, šajā vietā iebraukšanai ir vārti un "ķieģelis" https://www.mapillary.com/app/?pKey=849256453280094&focus=photo.

Paldies

Hi!

Like mentioned before, please use `highway=service` + `service=driveway` instead of `highway=unclassified` for roads that go to individual properties. These are not minor public roads, but (likely private) access roads. Only the parts of a road that are used publicly or access multiple properties should be tagged with a higher classification.

Thanks

Reverted in changeset/139444467.

Reverted in changeset/139444467.

Reverted in changeset/139444467.