[OSM-talk] Key:smoothness

Matt White mattwhite at iinet.net.au
Sun Feb 1 12:18:11 GMT 2009

Shaun McDonald wrote:

> "Good" means different things to different users. A racing cyclist,  
> touring cyclist, and trail/mtb cyclist will all have difference views.  
> These will be even more different to a wheelchair user, hiker, 4x4  
> vehicle driver, smart car driver, and a tractor driver. More  
> importantly take any one of the above, and they will have arguments as  
> to how suitable the way is for passing over. It is also a very  
> relative thing as someone in one part of the world will say a road has  
> one quality, while someone else will give something completely  
> different as that is what they are used to.
> Every single one of these types of people will have a different view,  
> thus it is a very, very subjective tag. You need to go back to first  
> principles and try to workout what makes a good road/path. Then each  
> data user will need to work out what they think is a good road on that  
> particular way. This makes it objective. The data is far more useful  
> to a wider range of people. It will be somewhat more work to process,  
> however it will give much more customisation for the end users of the  
> data. This will mean that the OSM data will have a much greater value.
> The surface tag is far better, maybe some more values are needed  
> there? Maybe some sub values are required?
The smoothness proposal sucked - I got semi involved in the discussion, 
but actually never voted either way  - came in a bit late once the 
barney was already in full swing (mostly trying to locate a nice way of 
tagging 4WD only roads). And what fun it was....

There are a couple of proposals kicking around under fairly average 
names that attempt the address the subjective issues raised in 
smoothness, and to improve the surface type and surface condition 
descriptions (as well as attempting to deal with this issue across 
multiple transport types).


Parts of a proposal I made on the Smoothness talk page are incorprated 
into the surface unification feature, some of which I think has 
potential provided sufficient loving is given to the proposal.


More information about the talk mailing list