There are benefits of surveying entrance=yes objects. Especially in areas where entrance=staircase/entrance=yes is likely to have ref and signed apartment range (for example in Poland), while say entrance=servicewill not have such data.
It also has some value in routing to distinguish actual building entrance from possibly multiple service or shop entrances.
But some entrances of residential buildings cannot be conclusively surveyed.
entrance 37 and 39 are clearly staircases, but 37a is just an entrance. (And left and right of that entrance, you actually see two more entrances but without a housenumber). For these entrances, it is not possible to get more detailed than yes.
I propose a new entrance=unclassifiable value (unspecifiable?) value that could be used in such cases - where entrance=yes cannot be replaced with more detailed value despite survey. Or maybe entrance=yes entrance_not_classifiable=yes for full backward compatibility?
This post and attempt to invent a new tagging is motivated by attempt to make it possible to implement related quest in StreetComplete and in general to allow automatic detection of places where survey will be useful. This would help to collect address data in some parts of the world.
Opinions whether it is ok to start using such value are welcome, suggestions for better tag value/key are also welcome.
Note that I am not planning to make a full-scale proposal with a vote so please comment also if you dislike this idea.
I have used level=0 and level=1 to distinguish in similar cases entrances from the ground floor from those on the first floor. In analogy, your example could maybe get level=0.5 tags. Also, drawing the stairs with step_count etc. could make sense and will likely be evaluated by common routing engines out of the box. wheelchair=no also comes to mind.
I don’t think I fully understand. You want to have a tag to say “this is an entrance, I just don’t know if it’s a special one”. Isn’t hat exactly what entrance=yes is used for?
In this context I understand “yes” to mean that there is some type of entry way but not why it exists. If you include the fact that it has double swing doors, I still only know about the doors present. It would still be “yes” since there is still nothing describing the purpose of the entry.
In my opinion, if a mapper is not able to use a detailed value for a tag, either because he/she doesn’t know what to use, or a value for that specific case doesn’t exist, the mapper should use the more general value (in this case “yes”).
In specific cases where a help is needed, a note would be useful to highlight that.
In such case it is impossible to distinguish places where only remote mapping from aerial images was done from places where on the ground survey was attempted.
So you cannot reliably find places where on the ground survey would be useful.
I did aerial mapping and there is some sort of entrance, please go there and check.
I did on-the-ground checking and there is an entrance, but I have no clue what kind of.
Ultimately, this comes down to “there is an entrance”, which is exactly what entrance=yes is used for. If you want to, you could add a source:entrance=survey, or check_date:entrance:=<date>, but I think even that is unnecessary…
… unless the whole point of trying to distinguish these 2 cases is to write a SC quest where you want to be able to stop showing an entrance, because it already has been surveyed.
yes, I want option to mark “I went there, I surveyed location, I was unable to change entrance=yes into something more specific”. It could be an extra tag or new entrance value.
And yes, I want this to make possible surveying this with StreetComplete (though could be also used by other tools listing places worth surveying)
Then creating a new value of entrance=* instead of using a prefix/suffix like check_date sounds like a bad idea, because you’d want to have a general solution, while a new value would only solve this one use case.
On the other hand, just because you were unable to classify an entrance into something more specific doesn’t mean no one else would be able to. Would you stop showing a SC question to other people, because a single person wasn’t able to answer the question?
If the goal is to stop asking you the same question over and over again, that should be implemented in the SC application, by remembering that you (as a user) are unable to answer the question and don’t want to see it again, but not by means of tagging this in OSM.
@Nadjita: now please have a little more understanding for the users of StreetComplete, who are asked the same question over and over again.
(I can’t decide right now whether I should be amused that people come up with such ideas at all, or annoyed because with the help of SC the database is increasingly filled with things that are actually default, but SC can’t distinguish whether it has already been checked on site and found to be default or a more specific entry is simply missing).
Just to avoid misunderstandings: I use StreetComplete myself. But sometimes … argh.
It’s a general thing in OSM: you can’t tell the difference between unsurveyed, surveyed but not an approved type, and surveyed but I don’t know these things. I don’t think classifying it as unclassified would solve this any time soon, and I don’t think we should invent tags for individual cases to tell the difference. It’s not an attribite of the entrance that the mapper doesn’t know the type, and it’s not a mandatory tag so you can’t force mapping it.
Tools should find their own way to deal with that. If a question keeps bothering the users, find a way to allow them to stop getting this question.
And while talking about it, I find entrance=staircase badly defined. It says just “A staircase entrance, almost always a door” which does not help much. Ok, most of the entrances to buildings are doors, so that does not seem specific. But neither is “A staircase entrance” really any more descriptive than entrance=staircase.
Only picture on entrance=* wiki helps a little (surprisingly, picture at supposedly more detailed entrance=staircase does not help at all!) - it seems that after a door is opened there are stairs. But do they lead to single private home (eg. building=home)? Or to a many different households (e.g. building=apartments)? Or it doesn’t matter and it can lead to both private homes and condos?
And do stairs have to be after the door, or can they be before door (or both before and after?). Do the stairs even matter, or can it be just a step or two (e.g. if it is single-floor apartment building)?
And if stairs must be there, must they begin immediately after the door as in that picture, or can there be a short hall (e.g. containing multiple postboxes) before stairs? (as is most common case in Croatia for multi-floor apartments buildings / condos, for example)
I agree, it isn’t clear from the pictures either because of the photo on the entrance key page (which seems to show an internal staircase of a single dwelling), and it misses indications when to use it (because almost any multistory building will have its entrances somehow leading to staircases so there is significant overlap with main and yes).
I think that for residential buildings we should map only the entrances as such anyway (entrance=yes). @dieterdreist is right, with multi-storey residential buildings there is always a staircase somewhere behind it, unless it concerns individual flats (or shops) on the ground floor or in the basement.
I think the distinction was once meant for public (publicly accessible) buildings like office buildings and shopping centres. There it is also important to distinguish between main entrances, entrances to stairwells, staff entrances, supplier entrances and escape routes.
In residential buildings, main entrance and entrance=staircase are synonymous for me. At most, one can still differentiate between entrance=home for direct flat entrances and entrance=shop for direct entrances to shops. I personally would not use service entrances (mostly direct entrances to the basement) in residential buildings.
If I compare your green marked entrances to: Key:entrance - OpenStreetMap Wiki all three would be entrance=home. I agree with @Matija_Nalis that entrance=staircase isn’t a suitable value-definition, as it gives no information regarding the purpose, like all the others are doing. But even though, it’s not reflecting your doors, which have the stairs outside.
The door between 37 and 37a looks like entrance=shop where the door between 37a and 39 looks to me like entrance=maintenance.
In summary: I rather would add more purposes for entrances than a second “yes”.
I agree that the wiki description is not clear. My understanding is that the type of entrance is based on what it lead to or way from. So a main entrance usually leads to some type of a lobby or central area. A service door would lead to an area designed for building staff to gather before going start their shift.
Using the same logic a entrance=staircase should be a door that leads directly to a landing with stairs or elevator bank. There should be nothing else directly accessible from this door. If there are connected rooms or other hallways than the entrance should be considered secondary.
agree that the wiki description is not clear. My understanding is that the type of entrance is based on what it lead to or way from. So a main entrance usually leads to some type of a lobby or central area. A service door would lead to an area designed for building staff to gather before going start their shift.
IMHO the “entrance” value is referring to the type of entrance, only loosely connected to the kind of area it leads to. You could have a service entrance leading to the main hall (as a secondary entrance) but likely not a main entrance that only leads to a service area.
