[Tagging] Removal of "amenity" from OSM tagging

[Daniel Koć]

W dniu 15.05.2015 15:11, Martin Koppenhoefer napisał(a):

> either one of the available tags fit for your purpose or you will have
> to invent a new tag, that's how OSM works. The situation would be

Sure, I know! =}

However when you have only few fixed categories, it's much harder to 
invent a proper one - coherent with the rest (also language-wise) and 
not overlapping with other definitions. That can be less of a problem if 
you have just a dozen of tags, but when there is more of them, lack of a 
clear system will hit you more and more.

> different for you if we had a common tag for any kind of higher
> education, but it seems that other mappers have thought this approach
> would be less useful for them (or we would already have this tag).

I don't think it's simply "if we don't have it, we don't need it, 
because if we need it, we would have it already". =} You probably 
underestimate the power of inertia and "good enough" system. In theory 
we encourage "any tags you like" approach, while in practice we tend to 
treat Wiki as the highest truth to deal with scattered cases and we 
render only a small subset even of already "approved" tags.

Most casual users are afraid and don't know the system, so they will 
just choose existing tags no matter what, and if there are no clear 
guidelines/categories to create new tags AND they need to put a lot of 
work (like writing the proposal, subscribing to the list, discussing, 
then voting) AND there is no chance to see it work in the short time, 
they will abstain or tag for existing tags or renderer.

There should be easier way of "creating" (defining and populating in our 
ecosystem) new tag schemes within more useful category tree. It's more 
reachable goal than trying to change casual users - exactly because they 
are not involved too deep and most of them never will.

> IMHO in general we have to deal with what is there, yes, we can try to
> make some changes to make the system more consistent or complete, but
> if you come and want to change everything there will naturally be some
> reluctance. This is ever more true with established usage of a tag. If

Of course. But I think we well need to change it anyway and it's better 
now than when we grow much more and the problem will get even bigger.

However real crisis may be exactly the point where any big change is 
really possible - I can only make the community aware of it and start 
discussion early enough to not make it too big.

> you still want to try, please use a namespace/ keys that are not yet
> in use, so to avoid conflicts with other mappers and to allow for
> parallel tagging during "transition" (also look at the tags path and
> public_transport which are examples for previous attempts to redesign
> parts of the tagging scheme)

I see the area=* namespace as the most interesting and realistic 
candidate, because it's really basic word/object, and while it may look 
like a highly conflicting one (almost 700k uses already! - 
http://taginfo.openstreetmap.org/keys/area), in reality it's underused, 
because 99,39% of values are just yes/no and there's no need to touch 
them. Other popular namespaces area:highway=* and area:size:ha=* are 
just a different notation (area=highway + highway=*, the second one is 
harder to translate, but not that important) and as I said: which 
notation will prevail is a secondary problem.

So all the needed elements are there. What we need is just a mental 
shift first (area=* may be a primary tag, not just an optional feature 
of some objects), then documenting it, tagging and changes in tools 
later. For me the biggest question is: are we ready for this and only 
inertia is at play or maybe there are some valid objections or issues we 
should think about before the change?

> cheers

+1 =}

-- 
"The train is always on time / The trick is to be ready to put your bags 
down" [A. Cohen]