footway line styles difficult to distinguish and in general much weaker than cycleway

[imagico]

This is a followup on #1765 which was closed by #1788 - but there are still several problems remaining with the footway styling that should be addressed.

This issue is about problems with the styling of footways and paths - it is not about the tagging semantics this styling is based on - problems with such should be separated. For specific issues at z13/14 see also #1748.

Specific problems are:

• the three different line styles are very difficult to distinguish, in particular of course in areas with a detailed path network at the lower zooms. To a slightly lesser extent this also applies to cycleways.
• The footway color is fairly weak for such a thin line feature leading to poor contrast. This is particularly visible in comparison to the cycleway color which is much darker and more saturated. The visual effect of this is significantly reduced by path/footway/cycleway - render missing surface separately, more prominent rendering for unpaved ways #1788 by increasing the footway width for the weaker style variants and decreasing the cycleway width leading to different stylings for both. But the basic issue of the two colors being very different remains.
• The problem of a very different impression of footways, in particular unpaved, on dark and bright background due to the relatively wide and semitransparent casing remains only slightly reduced through path/footway/cycleway - render missing surface separately, more prominent rendering for unpaved ways #1788.

Examples, all at z16:

Two suggestions have already been made to address some of these problems:

https://github.com/matkoniecz/openstreetmap-carto/commits/solid
https://github.com/imagico/openstreetmap-carto/commits/path-nosurface

Both have issues that would need resolving.

[matkoniecz]

https://github.com/matkoniecz/openstreetmap-carto/commits/solid seems to work well for footways (though testing was limited). Main problem is with cycleways and is described at #1788 (comment)

https://github.com/imagico/openstreetmap-carto/commits/path-nosurface - is it style that was used to render #1713 (comment) ?

[matkoniecz]

The problem of a very different impression of footways, in particular unpaved, on dark and bright background due to the relatively wide and semitransparent casing

Casing is a part of the issue but generally nearly everything (especially thin linear features) will have wildly different effect on untagged land, landuse=industrial, natural=bare_rock and landuse=forest. It is a consequence of rendering so many different landcovers as different colours.

The visual effect of this is significantly reduced by #1788 by increasing the footway width for the weaker style variants and decreasing the cycleway width leading to different stylings for both.

Small note - widths were different also earlier (both in the old style with separate display of highway=path and highway=footway and after #1713).

[matkoniecz]

For people interesting in testing their ideas:

unpaved footways on natural=bare_rock: http://www.openstreetmap.org/?mlat=47.56673&mlon=12.32377#map=19/47.56673/12.32377

high density of paved footways: http://www.openstreetmap.org/#map=15/50.0757/20.0455 http://www.openstreetmap.org/#map=16/50.0795/19.8863

cycleways vs water features: http://www.openstreetmap.org/#map=16/50.0407/19.8556

footways on @grass: http://www.openstreetmap.org/?mlat=33.32792&mlon=-112.08914#map=19/33.32792/-112.08914

footway on field: http://www.openstreetmap.org/?mlat=47.82989&mlon=12.07764#map=19/47.82989/12.07764

[imagico]

A few more notes:

• considerations of colors should probably be made with regards to New road style - rework road colours, road widths and display of railway=rail #1736.
• for a good choice of a pair of colors for footway/cycleway the boundary purple will - like in case of roads - likely be the major constraint. Some general discussion on the boundary color within the overall color concept of the style might be in order. In many problems regarding color this is currently 'the elephant in the room'.

Regarding casing - i was mainly seeing it in comparison to tracks where the width of the casing is the same and since the track line is significantly thicker the casing is less dominant.

[nebulon42]

I'm copying some interesting details from a discussion on Talk-at, so that they are not lost. In Austria we have a lot of hiking paths so a lot of OSM community members are naturally interested in this topic:

@fkv1 suggested to use trail_visibility (currently not possible) when surface is not present. I think that this would indeed an interesting option.

Another test area (https://www.openstreetmap.org/#map=17/47.48884/14.93895) was mentioned by @BorutAtGit:

highway=path with surface=unpaved vs. surface undefined,
above the tree line with partially undefined or different landcover (to the west and north)

[pos-ei-don]

Even worse, it shows via-ferratas as normal path. I met a couple yesterday without any climbing-gear struggeling.

[matkoniecz]

@pos-ei-don This is covered by #1500, maybe also it was tagged incorrectly.

[fkv1]

@pos-ei-don To prepare for a mountain tour, use a dedicated hiking map. The standard map cannot be used for that purpose, as it does not even contain contour lines.

[pos-ei-don]

@fkv1 It's not me, who used this map. I was the mountain rescuer, who had to help them, and they showed me the map on their smartphone ;-)

[jeisenbe]

We could still use help solving the issues with the current footway rendering, including this issue and also issues #1748 and #1765.