Good job Gerd,

Two remarks:


mtb:scale=0 can also be used for paved trails. I would suggest not to assume this means unpaved if mkgmap:unpaved=0

So better use this rule >0 and 0+


I guess bridleways can be paved too, so also apply this for mkgmap:unpaved!=0


Van: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> namens Gerd Petermann <GPetermann_muenchen@hotmail.com>
Verzonden: donderdag 9 februari 2017 08:58:58
Aan: Development list for mkgmap
Onderwerp: Re: [mkgmap-dev] [Patch] unpaved roads
 
reg. smoothness: the patch contains this line , forgot to comment it
Rule 7:
highway=* & mkgmap:unpaved!=1 & smoothness ~ '.*(bad|horrible|impassable)'  { add mkgmap:road-speed = '-2' }

reg. double values: I agree that most of them come from wrongly merged roads.
The patch treats them all as unpaved, the unpatched version treats them like paved.
I think the patch makes it more likely that someone will fix the wrong value(s), and it is producing fewer
errors.
I see no reason to believe that the first value is better than the 2nd.

Gerd
________________________________________
Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Alexandre Folle de Menezes <afmlistas@terra.com.br>
Gesendet: Donnerstag, 9. Februar 2017 17:45:46
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] [Patch] unpaved roads

Great, but now you are not handling smoothness anymore.

Regarding the double values, i would get the first and ignore all the rest.  I believe these come from wrongly merging segments with different surfaces, there's no way to know which one is correct.

Best regards,

    Alexandre

Em 09/02/2017 13:43, Gerd Petermann escreveu:

Okay, attached is v6 (more or less a mixture of v2 and v3)

Some comments:
Rule 1) surface values meaning paved:
highway=* & (surface=asphalt | surface=paved | surface=sett |
    surface=concrete | surface=concrete:lanes | surface=concrete:plates |
    surface=paving_stones  | surface=cobblestone  |
        surface=cobblestone:flattened  | surface=metal  | surface=wood)
{ set mkgmap:unpaved=0 }

Rule 2) tracktype=grade1 without explicit surface implies paved:
highway=* & tracktype=grade1 & surface!=* { set mkgmap:unpaved=0 }

Rule 3) highway which was not set to paved before and has a surface or valid tracktype tag is unpaved:
highway=* & mkgmap:unpaved!=0 & (  surface=* |  tracktype ~ 'grade[2-6]')  { add mkgmap:unpaved=1 }

Rule 4) highway with mtb:scale or climbing attributes is unpaved, even if surface says paved:
highway=* & (
    mtb:scale=* |
    sac_scale ~ '.*(mountain|alpine)_hiking' |
    sport=via_ferrata)
{ set mkgmap:unpaved=1 }
Rule 5) default for hw= path or track is unpaved:
(highway=path | highway=track) & mkgmap:unpaved!=0 { add mkgmap:unpaved=1 }
Rule 6) set special highway types to unpaved even if surface says paved:
(highway=bridleway | highway=unsurfaced | highway=via_ferrata) { set mkgmap:unpaved=1 }

I am not 100% sure about hw=bridleway in rule 6. I think horses don't like paved ways, so a paved
bridleway is probably a tagging error.

I am quite happy with this.

Does anybody know a good way to handle tags with lists like these?
1) surface=asphalt;paving_stones (all meaning paved)
2) surface=grass;earth (all meaning unpaved)
3) surface=asphalt;sand (mixed)

Taginfo shows that most values containing a semicolon are of type 2 or 3.
Maybe rule 1 should also contain asphalt;paving_stones , asphalt;paved and paved;asphalt

Please suggest improvements, else I'll commit this patch on Monday.

Gerd
________________________________________
Von: mkgmap-dev <mkgmap-dev-bounces@lists.mkgmap.org.uk><mailto:mkgmap-dev-bounces@lists.mkgmap.org.uk> im Auftrag von Carlos Dávila <cdavilam@orangecorreo.es><mailto:cdavilam@orangecorreo.es>
Gesendet: Donnerstag, 9. Februar 2017 14:02:05
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] [Patch] unpaved roads

I would interpret that the other way round. If surface=paved, concrete,
... treat it as paved, no matter what tracktype is. I think tracktype is
very often misinterpreted by mappers, and wiki doesn't clarify it much
either. Additionally, tracktype>1 may be an old tag placed before track
was paved, in the same way that you can find a lot of
highway=residential, tertiary, etc + tracktype=*

El 09/02/17 a las 11:25, lig fietser escribió:



My OFM will mark them as unpaved. The combination of track and grade3
counts more than the surface. I think the current default rules assign
them also as unpaved. Your patch v3 too, is that correct?



------------------------------------------------------------------------
Gerd wrote:

And what about combinations of surface meaning paved with tracktype?
The wiki says "Usually ... " or "Almost always an unpaved track" for most
tracktype values.

I see quite a lot of ways with
highway=track & surface=concrete:lanes & tracktype=grade3
in my Niedersachsen:
http://overpass-turbo.eu/s/lNI

I think those are not unpaved, no matter how nice it is to cycle on
them;-)

Gerd



_______________________________________________
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk<mailto:mkgmap-dev@lists.mkgmap.org.uk>
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev




_______________________________________________
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk<mailto:mkgmap-dev@lists.mkgmap.org.uk>
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

_______________________________________________
mkgmap-dev mailing list
mkgmap-dev@lists.mkgmap.org.uk
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev