make tracks more prominent on z13 and z14

[zdila]

Please make tracks more prominent on zoom 14 and 13 by adding the glow effect as it is on zoom 15+.
On zoom 14 the tracks are very hard to see eg. in forests mostly because of the pattern used (#965).

Zoom 15+ is OK:

Zoom 14 is the least readable:

Zoom 13 is better than 14 because of the missing forest texture, but still should be improved by adding the glow effect as it is on the zoom 15+:

[imagico]

This would mean more or less reverting #747 w.r.t. tracks - which i am not too sure is a good idea. Showing all tracks at z=13 is questionable in many areas though - especially of course at low latitudes. At least differentiating grades is fairly pointless.

http://www.openstreetmap.org/#map=13/48.0813/7.6565
http://www.openstreetmap.org/#map=13/3.9537/11.5951

I would not put too much emphasis of improving readability in forest since forest color is likely to change - see #1242.

[matthijsmelissen]

I'd propose to put this on hold because of @matkoniecz's google summer of code project, which is expected to cause major road rendering changes overall.

[zdila]

Showing all tracks at z=13 is questionable in many areas though

some tracks are paved roads (tracktype=grade1) which I think is good to have also on z13. hard to tell the upper tracktype grade...

[matkoniecz]

I am considering rendering paved tracktype=grade1 tracks like service roads.

Function of these roads for pedestrian and cyclists is exactly the same. It is also very similar in case of drivers - both may be used for shortcuts and as first/last part of route but it is not typical, frequently use is restricted.

In all cases paved grade1 track is closer to highway=service than to other tracks, though it is the strongest for cyclists (and I am aware about my bias here).

The negative side is increased complexity (wait, I though that I mapped highway=track - and I see service road). The positive would be reduced mapping for renderer - I am frequently fixing paved tracks mapped as highway=service.

[zdila]

@matkoniecz, I mostly agree. tracktype=grade1 and highway=service imply paved and often narrow (single lane) road.

There may be still some differences - in some countries it is illegal to drive a motor vehicle on highway=track.

[dieterdreist]

2015-07-03 18:24 GMT+02:00 Martin Ždila notifications@github.com:

I mostly agree. tracktype=grade1 and highway=service imply paved and often
narrow (single lane) road.

service does not imply a paved road. tracks might be single laned or more
(typically will be single lane).

There may be still some differences - in some countries it is illegal to
drive a motor vehicle on highway=track.

My suggestion would be to add (explicit) access tags accordingly, or you
will break a lot of applications: basically it is nearly impossible or at
least a huge load of work to check for all areas which legal road classes
are mapped how to osm tags, and which are the access "defaults" for these
particular kind of road class. For instance in Germany there are only 1 or
2 "Länder" which do generally forbid the use of tracks by some kind of
motorvehicles (e.g. Baden-Württemberg, IIRR mofas (<=25ccm) are excepted),
so this is not even uniform within country-granularity but would have to be
researched seperately also for smaller admin entities.

[kocio-pl]

The idea of common rendering tracktype=grade1 and highway=service sounds interesting.

[matthijsmelissen]

As far as I'm concerned, this is no (longer) an issue - other opinions?

[SomeoneElseOSM]

All the paths, including the tracks, at http://a.tile.openstreetmap.org/14/8135/5332.png are still virtually invisible, although, of course, the decision to remove that information from the map was a conscious one taken for genuine reasons in #747 . I can't see how if that change stays in the issues here can have been resolved? Are there some as-yet-uncommitted changes that would revert some of the changes in 747?

[imagico]

I think the tracks at z=13/14 are still a problem but more in terms of clutter and this is not going to be solvable by adding casing.

Since i was at it anyway here how it looks like without tracktype differentiation and thinner solid lines:

A tracktype cutoff would be possible as well.

[mboeringa]

That is a major improvement, but it will require people accepting that it is probably unrealistic to render tracktype at Z13/14... lets see...

[jojo4u]

I would rather accept not distinguishing tracktype at certain zooms than unifiying tracktype=grade1 with service. The latter are different classes of road (agricultural/forest use vs. general access) while the first is (only) about surface.

[kocio-pl]

Interesting analysis by @imagico about track rendering:

http://blog.imagico.de/drawing-the-lines/

[Adamant36]

@imagico, do you think any of the suggestions on your site about this would be implementable?

[imagico]

The road rendering changes of the ac-style depend on not having purple boundaries and are kind of incompatible with #3467 (well - not strictly technically but design wise they go in a very different direction).

[Adamant36]

The road rendering changes of the ac-style depend on not having purple boundaries

That's what I was thinking.