I’m proposing to add surface=resin_bound as a subtype of surface=paved.
Resin-bound paving is commonly used on playgrounds, parks and private areas as a hard, porous surface suitable for pedestrians and light vehicles. It consists of small stones or other particles coated with resin.
No existing surface=* value fits that description. Also started a discussion on the wiki.
Here is a photo, which I made myself and which I could add to the wiki:
+1 from me. I encounter this more and more, and have no real way to tag it. Funnily enough I am currently helping a friend with binding the gravel they have on their garden path!
Resin bound is a good tag, do you think it will be its own tag, or used alongside other mediums, e.g. resin bound polymer / gravel / wood chips?
Thank you for your effort. I also had it in my backlog to create such value, so I can add to this thread that:
I also took photos some time ago. I think the zoomed one might be a good choice for some apps with smaller screens. Uploaded to Wikimedia commons, so feel free to use it.
https://upload.wikimedia.org/wikipedia/commons/2/2a/Surface_resin_bound.jpg
EDIT. 2 images included, but for some reason the post shows only one
surface=resin_bound_gravel has some use, surface=resin_bound has none.
Sounds reasonable to have specific tags like surface=resin_bound_gravel; but one should also note that for it to be useful it would require adding support in various data consumers - from editors (like iD, StreetComplete, JOSM …) to navigation apps (like OsmAnd etc).
So proponents should at least be prepared to go open and issues for dozen most popular data consumers.
Otherwise they’d not understand it, and would either show it as unknown (which is much less useful than paved), or even prompt users to overwrite such unknown/unpopular values with more common values.
(of course, it is often chicken and egg problem – editors might not include such preset until it is in popular use, and it is hard to get popular use until editors offer it)
The aggregate in resin-bound surfaces isn’t always gravel, it can be pebbles, glass, recycled material.
Which was the gist of that comment, no?[1]
“resin-bound crushed gravel” is so different from e.g. “resin-bound smooth glass” (or “resin-bound smooth pebbles”) that it makes no sense to tag them with the same surface=resin_bound value, as those surfaces would behave completely differently[2], so IMHO they should have completely different surface=* values.
update: apparently no, it was not the gist of it; I misunderstood “has some use” as “it has some usefulness” instead of “it is present in OSM database”. My point centered on “usefulness” is thus orthogonal to that @SomeoneElse post. ↩︎
e.g. when wet, the former behaving more like surface=concrete, and latter more like surface=sett ↩︎
I don’t think you can easily tell apart glass, plastic granulate, or stones on a survey. The feel and slipperyness will mostly be determined by the shape of the particles and the type of resin I imagine, you won’t even feel the material below the resin as long as the coating is still intact at least.
And IMHO different behaviour when wet is not a good reason to create multiple top-level sub values. Different types of asphalt or paving stones or earth can also have various levels of slipperyness when wet. And if slipperyness when wet is the main factor to distinguish them, that means we can’t survey that when it’s dry.
If anything I’d propose
surface=resin_bound
resin_bound=gravel
Or something if you want to be super precise, but for reasons mentioned above, I don’t think the lower value will be tagged often enough to warrant documentating.
If material is indeed undeterminable by on-the-ground mappers ([citation needed]); then perhaps some basic differentiation like surface=rough_resin_bound and surface=smooth_resin_bound (if “rough resin bound stone” behaves the same as “rough resin bound glass”, etc.) would encompass minimal common set of those big differences?
Uh, I disagree. Different behaviour of the surface in different typical conditions is IMHO actually the main (if not only) reason why you’d invent a new surface=* value at all in the first place. I at least enter data in OSM so it would be useful for data consumers.
What would be the purpose of inventing new surface=* value, if the new surface looks and feels practically identical to the old popular value? (I.e. duck tagging principle).
Or are you saying that how the surface behaves between wet/dry weather is irrelevant (or of extremely low priority) when tagging surface? If so, I would disagree; e.g. my main use of surface=* tag is to decide (in router apps, or when route planning) how usable/comfortable taking road A vs. road B would would be to me in certain situation.
If some surface=* tag doesn’t give me information to help solve that problem, then such surface=* is worthless to me. Might as well be just generic surface=paved.
But we do differentiate in topsurface=* between:
asphalt vs. chipseal vs. tarmacpaving_stones vs. sett vs. unhewn_cobblestonecompacted vs. fine_gravel vs. graveldirt vs. sand vs. lateritemetal vs. metal_gridexactly because their behaviour differs in certain conditions (like rain).
I fail to see why resin_bound should be the exception?
Which is why I’m proposing to tag what the surface actually is, instead of how it behaves in specific circumstance (which could be inferred from the former).
IOW, if user sees “resin-bound pebbles”, that surface=resin_bound_pebbles is what they should tag (as it does not differ during dry or wet seasons, or during summer and winter etc. so should be deterministic in all situations), and it is up to data consumer to decide what to do with that definition.
But it needs to be tagged with enough detail in the main value if surface behaviour differs significantly. Otherwise one would end up with just another “generic” value like paved, with vastly reduced utility.
This substance category should also be a candidate for things like picnic table and bench material; it’s a popular YouTube Short genre to take fancy-looking garbage and make a resin bound table out of it.
You mention playgrounds.
Many playgrounds with resin-bound surfaces are resin-bound rubber, which behaves very differently from the resin-bound stone when a child falls on it, eg from a climbing-frame.
I think we need to distinguish resin-bound-rubber from hard resin-bound surfaces at the very least. Not doing so would be worse than lumping tarmac and resin-bound-stone together, and defeat any value of surface=resin_bound.
Are you thinking of rubber mulch? That’s recycled shredded rubber bound with polyurethane, laid directly on grass as a soft safety surface. It’s a different product from resin-bound stone and wouldn’t realistically be tagged surface=resin_bound.
I had not realised that could be laid directly on grass (which would then die?).
More pertinently, I hadn’t realised the distinction between resin and polyurethane and would not be able to identify the difference on a field visit.
The site you linked to is advancedresinsolutions.co.uk
I think this is likely to be tagged as resin.
Unless you are actually standing on it it is easy to mix the two up. I suppose that I am saying that we need something like surface=rubber whilst surface=resin_bound is only nice to have.
What tag do you suggest for rubber-type playground surfaces such as this ?
[ I had been afraid that someone might object that this sort of “rubber” does not come from a tree so is called something else. I am glad you did not do that. ]
Resin-bound rubber (rubber mulch, poured rubber) and resin-bound aggregate (stone, glass, or other hard aggregate) are different surfaces: same binder, but one is springy and soft underfoot, the other is hard. That’s also a reliable field test, no chemistry knowledge needed.
For rubber playground safety surfaces, surface=rubber is already documented on the wiki for exactly that use case.
I agree. Since same binder applies to quite different surfaces, I think it would be bad idea to call the new value exclusively for the binder itself (e.g. surface=resin_bound
Given that hard aggregates also belong to two categories which behave quite differently (see above [RFC] Proposal: surface=resin_bound - #7 by Matija_Nalis), we should have 3 different categories:
As you note, for case (3) we already have surface=rubber, so there is no need to try to re-invent hot water there.
But it is hard to find good name for the first two: we might do surface=hard_rough_resin_bound_aggregate and surface=hard_smooth_resin_bound_aggregate, but those (while having a great advantage of being clear and unambiguous!) are also admittedly somewhat unwieldy.
However, shortening them by throwing out most important distinguishing part (i.e. going for just surface_bound
So perhaps some ideas how to shorten those while still having their names being clear and unambiguous would be nice? Any suggestions?
as some (many?) people would see resin_bound and use it to tag surfaces which are indeed “resin bound” – but without reading all the wikis in detail beforehand ↩︎
I might be wrong, but my understanding is that resin-bound surfaces are generally smoother because the aggregate is mixed with the resin before being laid. In contrast, resin-bonded surfaces are made by applying resin to the base first and then scattering aggregate on top, which results in a rougher finish.
If that’s the case, then resin-bonded surfaces would be out of scope, and something like surface=resin_bound_aggregate might be sufficient?
Also, from what I’ve seen, resin-bound aggregate is usually advertised with pebbles or decorative stone mixes rather than crushed stone, which seems to fit the smoother finish.
It’s the second time today I read a post where slipperiness is mentioned as something worth mapping. Maybe we could do that by extending smoothness beyond excellent and introduce smoothness=slippery (so smooth it’s dangerous)?
Uh, that makes it even worse; I’d wager most users would be quite confused by such differences between “resin-bound” and “resin-bonded”, due to similar root of the word. 
Perhaps it is better to avoid both b* words, then.
But that slipperiness is orthogonal to smoothness[1].
E.g. surface=unhewn_cobblestone is often extremely slippery when wet (and often quite slippery when dry), yet I’d never seen anyone claim that it is “more smooth than smoothness=excellent)” ! (try driving fast in your rollerblades over it)
IOW, you can have smoothness=bad + slipperiness=extremely_slippery; as well as smoothness=excellent + slipperiness=not_slippery_at_all
P.S. I though I’d made it up, but it turns out the tag slipperiness=* already exists (in some tiny use) 
at least on macroscopic level ↩︎
I had in mind to introduce it only for surfaces that are so smooth they are dangerously slippery. For surfaces that are slippery despite not being very smooth, you could tag hazard=slippery (even when there’s no sign; that restriction mentioned in the wiki could be removed). However it would probably be simpler to just use hazard=slippery for all slippery surfaces independent of smoothness.
or to just choose one of them (why have different tag values if most mappers don’t know the difference?) and then describe slipperiness and smoothness with hazard=slippery and smoothness=*
“Resin bound” is the established trade name, used consistently by Wikipedia, manufacturer sites, and installer specs, and resin-bonded appears to be a different product anyway (aggregate scattered on top, rougher texture); a wiki disambiguation note handles that edge case better than coming up with a name nobody in the industry uses.
