Move loose barrier=gate/lift_gate to z18

[kocio-pl]

Similar to #2756.

Currently we show all the gates from z16 and it may be important to see them so early (I'm not sure about it), but at least we should render loose gates (not included in any road) from z18. This would promote more complete tagging (adding roads), which would be good for navigation, and avoid rendering clutter at the same time.

See the example (Italy - Cles, z16):

[kocio-pl]

By the way - does anybody have a clue why do we need such barriers on z16+?

[SomeoneElseOSM]

By "such barriers" do you mean "any gates at all" on z16+ or "gates on roads, paths or tracks" at z16+? I can see the logic of displaying gates in hedges etc. but not on paths later than gates actually on paths (though naturally being less urban-centric than most I'd say "17 instead of 15" rather than "18 instead of 16" - but that's a different discussion altogether).
How technically would you do "this gate is part of a highway so render it sooner" (apologies if this is offtopic; I'm asking because it'd be good to do that at e.g. https://map.atownsend.org.uk/maps/map/map.html#zoom=15&lat=53.16392&lon=-1.38033 ).

[kocio-pl]

I mean both types from the title, however there can be more of them - especially with the new icons designed lately: #846 (comment).

Technically I've been given a hint to look at the turning circle code - it's a bigger SQL statement, so I don't see the immediate pattern, but maybe somebody can use it:

openstreetmap-carto/project.mml

Lines 639 to 666 in 94e4623

	- id: turning-circle-casing
	geometry: point
	<<: *extents
	Datasource:
	<<: *osm2pgsql
	table: |-
	(SELECT DISTINCT ON (p.way)
	p.way AS way, l.highway AS int_tc_type,
	CASE WHEN l.service IN ('parking_aisle', 'drive-through', 'driveway')
	THEN 'INT-minor'::text
	ELSE 'INT-normal'::text
	END AS int_tc_service
	FROM planet_osm_point p
	JOIN planet_osm_line l ON ST_DWithin(p.way, l.way, 0.1) -- Assumes Mercator
	JOIN (VALUES
	('tertiary', 1),
	('unclassified', 2),
	('residential', 3),
	('living_street', 4),
	('service', 5)
	) AS v (highway, prio)
	ON v.highway=l.highway
	WHERE p.highway = 'turning_circle'
	OR p.highway = 'turning_loop'
	ORDER BY p.way, v.prio
	) AS turning_circle_casing
	properties:
	minzoom: 15

[kocio-pl]

I'm getting more and more convinced to start with moving all such barriers to z17+ (and later move detached ones to z18+). What do you think about it?

[kocio-pl]

See also #323 - quite a lot of discussion to read, but maybe we could start with moving all the gates to z17+.

[kocio-pl]

See also #1161 - could be done as a simple rework of both color and zoom level.

[lakedistrictOSM]

I'm getting more and more convinced to start with moving all such barriers to z17+ (and later move detached ones to z18+). What do you think about it?

z16 is generally too early for detached and attached barriers in both urban and rural areas. z17+ would work better.

[kocio-pl]

I think also about moving bollards to z17+. I also expect them to be on local roads, so lower zoom is not needed for showing them.

[Adamant36]

@kocio-pl, are you still interested in moving gates to z18? If so, I'm totally willing to do a PR for it. Along with bollards.

[kocio-pl]

After #3157 (moving barriers from z16+ to z17+) it does not bug me, because there are more obvious things I focus on. But if you think it's worth more attention, feel free to act.

[Adamant36]

I guess its OK for now. Want to close the issue since its essentially dealt with?

[kocio-pl]

OK, it makes sense.

[jeisenbe]

I would prefer to keep them on z17

…

On Fri, Dec 21, 2018 at 11:57 AM kocio-pl ***@***.***> wrote: After #3157 <#3157> (moving barriers from z16+ to z17+) it does not bug me, because there are more obvious thins I focus on. But if you think it's worth more attention, feel free to act. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#3122 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AoxshJ54iZ2nU-Kfa396wVZ5xMCDwGGIks5u7E4ZgaJpZM4SsAbh> .

[Adamant36]

Yeah, I think they are fine there. That's why I had Kocio-pl close the issue.

[jeisenbe]

Sorry, my comment took a long time to post (30min). Slow internet today.

…

On Fri, Dec 21, 2018 at 12:54 PM Adamant36 ***@***.***> wrote: Yeah, I think they are fine there. That's why I had Kocio-pl close the issue. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#3122 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AoxshOEgu6R7EqYSgtpqsX7ID4mOyim1ks5u7FtngaJpZM4SsAbh> .

[Adamant36]

@jeisenbe, oh. No worries. It sucks your internet is that slow.

[kocio-pl]

I'm impressed how much you do with such a bad connection...