Please, DO NOT create no_u_turn restrictions with identical ways in from and to roles. This type of turn restriction relation DOES NOT add any value to the map. Please learn carefully and understand what turn restrictions are for in OSM before mapping these. Turn restriction relations in OSM map the FLOW OF TRAFFIC, not the legal situation on the road. What you are trying to map here is the legal situation on the road. If you want to map the legal situation on the road use the traffic_sign=US:R3-4 tag.

👍 Danke!

Kannst du bitte mal prüfen ob die Buslinie A614 nun weiterhin korrekt durch den Kreisverkehr in Eiserfeld verläuft? Es ist schwer aus dem allgemeinen Verlauf einzuschätzen wie der Kreisverkehr durchfahren wird. Danke!

Or, look at “SW Moody & Gibbs” (https://www.mapillary.com/map/im/9G0D4nrLEYzlb2yQM2_LAw) which is served by bus routes 35 and 36, and the Portland Streetcar NS line.

>> The phone is a public phone, not a private phone.
>
> Not sure why you think so because it's not. In fact, I didn't see a phone inside
> at all. It must be hidden somewhere behind the counter.
Perhaps it is.
>> The same applies to “7 days open” on the sign. As an observer you are allowed
> to make assumptions, regardless whether you are on site or sourcing data from
> imagery.
>
> Making reasonable assumptions, like "the phone number on the sign is probably
> correct" is fine, although you should even be wary of doing this in areas that
> you are not familiar with (anyone native to Oregon would know that a 7-digit
> number is not valid).
I am and I was aware that the phone number is probably not valid. Nevertheless, it is what the sign says. Besides, exceptions may still exist even after renumbering has been completed. As a mapper for OSM, my responsibility for verifying the validity of most data basically ends with reading what a sign says or what you can seen in an image. If I can verify data more thoroughly or to further ends then fine, I can map more accurately. So, it is what it is.

Assuming that “open 7 days" means 24/7 is a poor assumption.

That’s your opinion but I think it is still a reasonable assumption if you happen to know that most convenience stores in the US (and around the world) are open 24/7. Technically, if read through the eyes of a lawyer then sure, “open 7 days” does not necessarily include 24/7.

>> Bus stops are fully accessible when people on wheelchairs can get to the bus
>> stop without obstacles, that is just roll through. The same applies to boarding
>> and disembarking a bus. The bus stop does not seem to have been outfitted with a
>> ramp or any other device to enable wheelchair users to roll on board.
>
> Where did you get this idea from? The wiki says "Use this tag to mark a feature
> in the United States as compliant with the Americans with Disabilities Act
> (ADA)." To the best of my knowledge, the ADA doesn't require the stop to have a
> ramp to the bus.
Right, because ADA is not concerned with any particular pieces of infrastructure (at least not directly) but with organizations, businesses, and state authorities. So, you have to make a distinction between a transportation entity (or agency, network, etc), that is the bus fleet operator, who serves any particular bus stop and a bus stop as a physical piece of infrastructure. ADA does not require putting compliance stickers neither on bus stops (or buildings) nor on businesses or organizations. Simply put, an entity is ADA complaint as long as nobody complains or an oversight authority does not step in. But, this is not exactly how OSM works. OSM does put compliance stickers on physical objects in its database, by mapping with three different `wheelchair` values. Labeling with any particular value depends on the properties of the object. For example on buildings, `wheelchair=limited` usually means nothing more than that the entrance has a step which potentially can be overcome by most wheelchair users on their own. You can be more specific, map the entrance separately and tag it with `wheelchair=limited` and map the building with `wheelchair=yes` where building amenities might be accessible but not its entrance. On bus stops however, `wheelchair=limited` means that you can get to the bus stop but once you are there, you cannot expect to find a ramp or a raised curb there and you are at the mercy of the bus (and driver) to pick you up (or that there is some sort of obstacle to access the bus stop but otherwise getting on and off the bus is smooth). Basically, it means that you are gambling unless you are sure that the route is served by an accessible bus.

> People in wheelchairs can get to the bus stop because the curb
> is lowered on both sides. If there is no ramp, the bus is not wheelchair
> accessible, but the stop still is.

This is only half of the story to a bus stop. As you said yourself, the bus stop itself may be wheelchair accessible but the bus may not (or vice versa). And, as a wheelchair user when you look at a bus stop you are not only interested in getting to the bus stop but also in getting on and off the bus. This is why in OSM terms `wheelchair=*` is more about practical accessibility, not so much about a legal or technical standard. Again, OSM sources its data from observable reality like signs or missing ramps, not from legal standards.

> I've never even heard of having such a ramp at the bus stop rather than on the
> bus. I can't imagine any practical way to make that work, since the bus would
> have to stop in a very precise place for the ramp to extend inside. If you know
> of anywhere that this exists, I'd like to hear about it. It certainly doesn't
> exist in Portland.
Just because you have never heard, seen, or cannot imagine something does not mean that it does not exist. Generally speaking, this a bad premise to operate on.
Look at “South Lowell & Bond”: https://www.mapillary.com/map/im/QqYwsTwqdVVZd9IWiefGMw
Although this stop is primarily served by street cars (a `tram` in OSM terms), it is fitted with a raised curb. So, it can also be served by buses without a ramp. This stop has a modern raised curb but there are also older style raised curbs in Portland, like at “West Burnside & Burnside Bridge”.
See here: https://www.mapillary.com/map/im/E0iJuaxW1-NCTh-1-Eyl_w
There are many more similar examples in Portland (and around the world).

> First, private phones inside of buildings should not be mapped, and certainly not with a node randomly dropped inside of the building.

I did not add the phone, so I did not drop it randomly. All I did was augment it with additional tags. The phone is a public phone, not a private phone. OpenStreetMap accepts and appreciates indoor mapping.

Since OpenStreetMap generally operates on the “posted” or “signed” rule of thumb and the market did not have a `phone` tag it was safe to assume that the phone number as it appears on the sign is the market’s phone number, regardless of any fact that it may or may not be in use anymore. It is completely admissible to collect data to the best of your knowledge. Should anybody survey the area and collect any updated data then they are free to update the market’s phone number, including a `note` that the market’s sign is misleading or outdated.
The same applies to “7 days open” on the sign. As an observer you are allowed to make assumptions, regardless whether you are on site or sourcing data from imagery. What OpenStreetMap just “expects” of you to do in terms of quality of data, is to collect data to the best of your knowledge and ability, and to annotate the source of your data.

> Please don't map based on assumptions.
I will most certainly continue doing that because this has been OpenStreetMap’s mode of operation ever since. OpenStreetMap is not a court of law where facts have to be proven beyond reasonable doubt. OpenStreetMap collects DATA based on an observer’s best effort interpretation of artifacts (like signs) and natural phenomena, while data sources are cited to be able to verify (not prove) an interpretation later. Moreover, OpenSteetMap is a consensus driven project where certain habits and rules of best practice have crystallized. And so far, I fail to see where I might have deviated from the consensus or rules of best practice.

Bus stops are fully accessible when people on wheelchairs can get to the bus stop without obstacles, that is just roll through. The same applies to boarding and disembarking a bus. The bus stop does not seem to have been outfitted with a ramp or any other device to enable wheelchair users to roll on board. Whether a bus stop is exclusively served by bus vehicles which may or may not be boarded by “just rolling on board” is a property of the bus, or a fleet of buses, not the bus stop.

> A bus stop is the area around a sign on the sidewalk.
Well, according to the new public transit mapping scheme “the area around a sign” is actually a `platform`. In the legacy scheme the boundary of a bus stop is rather blurred and not clearly defined. In most cases this includes the sign (if any), the platform, any shelter, bench, ticket vending machine etc. So, you could also map all these things separately and interpret the bus stop sign (or the platform) as `highway=bus_stop` only. There is nothing wrong with adding `shelter` or `covered` tags in the legacy scheme. `shelter` and `covered` are here often used and interpreted interchangeably (or synonymously).

If you have surveyed the area recently then you may update the OpenStreetMap database anytime you want.

To be more specific, I have added the Mapillary image keys.

@SomeoneElse Please excuse the fact that this conversation has gone off topic. However, I do not know of any more adequate place. If you know a better place to continue the conversation about this topic then please share a link or reference of some kind. Thank you.

> content (collectively, 'Contents') to the geo-database of the OpenStreetMap project (the 'Project')

Please note that granting copyright applies explicitly only to “the geo-database of the OpenStreetMap project”. Thus, please explain what covers changeset comments.

> With regards to the legality of a third party company using OSM data, please re-read the agreements that you signed when you started contributing to OSM

As far as I can tell changeset comments are not covered by the ODbL, simply because it is not part of the map database. Please enlighten us what agreement covers changeset comments because as far as I can tell neither OSMF's terms of service nor the privacy policy mention that commentators relinquish their copyright or agree to transfer rights to a third party.

> (there's a changeset comment fedd for USA+Canada on the OSM US' Slack channel)

Does this mean that changeset comments are automatically and without notice forwarded to a private third party company? If that is so then this practice is surely illegal and may constitute a criminal offense.

> In that case, the 'on the ground' rule doesn't make sense because the signs are not being replaced on the whole street at once.

It was not me who came up with this rule. So, I do not think pointing this out in some edits could be considered as nit picking. I also do believe that this rule has its merits and drawbacks. So far it has worked well for OSM.

Anyhow, since @tguen has explicitly surveyed the area now and the City of Portland is going to replace existing street name signs though slowly but steadily, I think it is best to accept the current state of things in OSM because it reflects the city's intention and eventually reality is going to catch up with OSM's state. @tguen thank you for surveying the area and providing the link to PBOT's announcement. I just wish the OSM changesets had been attributed with sources, so that we would not have been confused and did not need this conversation.

I am not going to vandalize the map and if I ever do so then not on purpose (which in that case is not vandalism anymore). You see, the problem we have here is that you could have spared us this conversation if you had attributed your sources. Since readdressing is a big deal, next time somebody, including the DWG, may come along and revert your changes just because you have provided no sources. You could have or rather should have cited city records or the fact that you may have surveyed the area. Even the DWG needs some sort of evidence or attribution to find out what is going on. So, please cite you sources sufficiently. `source=survey;https://www.portlandoregon.gov/transportation/75814` would have been good enough. Anyhow, if this is what is on the ground I am fine with it.

As far as I can tell the notice only says that city records have been updated and property owners have been notified, including other agencies like the IRS and USPS. If things have visibly changed on the ground then why did you not add a `source=survey` tag?

Again, you have referred to an external source. Where is your on the ground evidence?

Actually, you should have updated the street name to `Southwest Kelly Avenue` here instead because this is what the street name sign says. For example, see street name signs on Southwest Gibbs Street and Southwest Kelly Street. Again, please be careful and double check with what is on the ground. Do not simply copy from other sources and do not assume from one piece of data in OSM on other data.

Anyhow, fortunately because of this mistake we can actually fix the street names now, since Southwest Kelly Street is not the only one which is wrong/incomplete.

Please let me elaborate more on the address issue, or mapping addresses from official address registries. While official address entries may exist in official records they may not “exist” in the real world. OSM's mission is to map, or more specifically model, the existing perceivable world. If a building has multiple addresses but only one or few are perceivable (mostly visible) on the ground then only these addresses should be mapped. I am not aware of any tag which could denote that an address is a synonymous or secondary address. OSM does not map or collect addresses for the post office. Post offices are governed by separate rules and regulations. For example, some entities (usually with large volumes of mail) are addressable exclusively or inclusively by a specific ZIP code. This kind of addresses should not be mapped, unless this is what a building or piece of land is signed with.

> This from is RLIS data on oregonmetro.gov.

It is okay to source or import data from other databases or registries as long as the data is adapted or transformed to the OSM-way of doing things. Please note also that OSM collects data based on the facts on the ground, not according to any legal or other third party registries. Official sources are not authoritative for OSM but only the factual situation at the location. So, please be careful not to simply copying from other sources even if they are official. Despite the fact that OSM does not model primary or secondary addresses—address precedence in that sense—there is nothing which would suggest that the building (or actually compound of buildings) has a primary address. Anyhow, I think that in this case it is okay to have the building outline tagged with an address.

> 971 has been a Portland area code for many years.

I did not know that, so I left it as is. So, according to Wikipedia sources 971 has only been introduced since July 1st, 1999. Well, I guess I always learn something new.
NANP is a national standard. Indeed, few other countries have also adopted NANP as a national standard. Yet, this and the fact that NANP complies in the number plan aspect with ITU does not make NANP an international standard. AFAIK, it is not an ITU standard, nor a subset of any ITU standard either. OSM does not follow national standards for good reason. So naturally, OSM prefers international standards, including the ITU standard for phone number formatting. The main reason why OSM tolerates NANP formatted phone numbers is because everybody can commit to OSM, even unaware newcomers who just do know any better and adhere in good faith to their national standard, and because of the shear count of phone numbers already collected in OSM. Nevertheless, for newly added phone numbers the ITU formatting is preferred.

Oh, do not bother. I have reviewed it more thoroughly. The building has multiple addresses, so I have added one which is visible in the Mapillary imagery.

I not sure about the phone number because the rule of thumb is that land line phone numbers should correlate with the geographical location. So, I just left it as is.

I am a bit confused why the `addr:housenumber` changed. Additionally, the phone number you have added does not feature Portland's area code prefix of 503. Lastly, please use ITU's phone number formatting (see the Wiki) when adding phone numbers to OSM. Could you please explain?

Can you please give a more verbose description of what you have done in this changeset or what was the purpose of it? After reviewing your other recent changesets we have to conclude that your descriptions are actually non-descriptive. “Portland”, “Salisbury”, or “Temple, TX” are not descriptions. For example, you have transitioned buildings in the OSM life cycle in this changeset. How can you be sure the buildings do not exist anymore? If you have done this based on imagery please make sure that perform modifications based on the latest imagery or survey.

Since judging by the count of your commits, you are a new mapper, please refrain from modifying the map for now. Please add new map data instead to learn the basics first.