[Tagging] usage of maxspeed:practical is described as recommended on wiki

[Ilpo Järvinen]

On Sat, 23 Aug 2014, Mateusz Konieczny wrote:

> 2014-08-23 10:48 GMT+02:00 Richard Z. <ricoz.osm at gmail.com>:
>       On Sat, Aug 23, 2014 at 09:08:08AM +0200, Mateusz Konieczny
>       wrote:
>       > See
>       https://wiki.openstreetmap.org/wiki/Talk:Key:surface#maxspeed:practical
>       > for proposed change
> 
> with 12000 ways already tagged maxspeed:practical and lack of
> alternatives
> I would think twice removing any documentation.
> 
> 
> 12 000 ways is really low number in this situation. Surface tag is used on
> nearly 9 million roads, number of highway=* ways crossed
> 76 million.

Since you go into the realms of "statistics", could you also estimate few 
things relevant here:

- How much out of those 76 million and 9 million highways/surfaced is 
directly from import or drawn from imagery, and nobody never really 
surveyed by a mapper on ground?
- How much of the remaining you'd expect to need maxspeed:practical?
- How much of such ways that would be a candidate for maxspeed:practical, 
are within reasonable distance from an active enough mapper? ...In 
particular, I don't think we have that good mapper coverage for rural 
areas which is likely very very relevant here.
- Also many mappers are likely to not know about maxspeed:practical
(I suspect this could be seen as supporting your argument too but might 
also be relevant to explain "low number").

Based on the above realism check to your "statistics", I come to exactly 
opposite conclusion than you. 12k is a big number. Out of 4.5M ways with 
maxspeed, a quick calculation yields to 0.25%, IMHO too few zeros were 
necessary to represent it for it to be a "really low number".


I agree with you though that coming up a value for this tag that mappers 
would agree is unlikely in practice so I don't like/use this tag myself
but I can also understand why somebody would want to have such tag.


-- 
 i.