Talk:Colombia/Project-Ciclovías

From OpenStreetMap Wiki
Jump to navigation Jump to search

Help received from others

This is a list of the help that we have received to define this project:

Cycle route as relation not lanes

Under Colombia/Project-Ciclovías#Mapping it should be To map bicycle routes, the sections of the highways must be added to a relation<. Please, do not use relations for lanes. --Skyper (talk) 13:48, 10 February 2022 (UTC)

Corrected, this a translation error. Lanes are not grouped, but the highways.
--AngocA (talk) 17:33, 10 February 2022 (UTC)

Conditional oneway lanes

I think oneway:lanes=* is a poor choice and personally would use lanes:backward=*, lanes:forward=* and lanes:both_ways=* as default plus with postfix :conditional. This works as long as the both_ways lanes are right of backward and left of forward lanes (right-hand traffic). Additionally, it is useful to add oneway:bicycle=no in case of oneway:bicycle:conditional=yes @ (…). --Skyper (talk) 13:48, 10 February 2022 (UTC)

Just to be sure, what you mean is to change this oneway:lanes:conditional=yes|no @ (Su,PH 07:00-14:00) for this oneway:lanes:forward:conditional=yes|no @ (Su,PH 07:00-14:00)? As in case 4.
Following the same tagging structure, what about oneway:lanes:conditional=yes|yes|-1 @ (Su,PH 07:00-14:00) in case 5 or 6? Should it be change to oneway:lanes:forward:conditional=yes|yes|-1 @ (Su,PH 07:00-14:00)
Finally, I thought the following could be the same and written interchangeably, but now I am not sure about that, and I cannot identify what is difference or implications: oneway:conditional=no @ (Su,PH 07:00-14:00) and oneway:bicycle:conditional=no @ (Su,PH 07:00-14:00). I put that in cases 2, 3, 6, 7 and 8.
--AngocA (talk) 01:14, 11 February 2022 (UTC)
Yes, oneway:bicycle:conditional=no @ (Su,PH 07:00-14:00) is not needed together with oneway:conditional=no @ (Su,PH 07:00-14:00). Similar is true for oneway:bicycle:lanes:conditional=no|no|no @ (Su,PH 07:00-14:00) which can be shorten to oneway:bicycle:conditional=no @ (Su,PH 07:00-14:00) or in even oneway:conditional=no @ (Su,PH 07:00-14:00). --Skyper (talk) 13:32, 2 March 2022 (UTC)
I tried to come up with examples and thought about this over and over again, but I found not solution, so far. The problem is the definition of lanes:*=* and the post-fix :lanes. It always considers the driving direction which means without :backward, :forward and :both_ways is only a special case for one-way roads with the only exception change:lanes=no|no. That is why oneway:lanes=* does not work as expected.
Plus we have a strict definition where these lanes per direction are placed depending of side to drive. For right-hand traffic the fixed order from left to right is, first :backward, then :both_ways and last :forward. The problem that it is not all the time a one-way road might be solved but the fixed order is my major concern.
For 4., I came up with following:
highway=secondary
lanes=3
lanes:forward=3
lanes:forward:conditional=1 @ (Su,PH 07:00-14:00)
lanes:both_ways=0
lanes:both_ways:conditional=2 @ (Su,PH 07:00-14:00)
oneway=yes
oneway:bicycle=yes
oneway:bicycle:conditional=no @ (Su,PH 07:00-14:00)
vehicle:both_ways:conditional=no @ (Su,PH 07:00-14:00)
bicycle:both_ways:conditional=designated @ (Su,PH 07:00-14:00)
foot:both_ways:conditional=designated @ (Su,PH 07:00-14:00)
This does not work with 5., though, unless we have a tag that changes the traffic_mode to left-hand traffic for the segment of the road.
I guess we need to get more opinions, maybe, you can reach out for imagic but I am not sure how active he is, atm. Another concern are all the QA tools which will probably go crazy with this detailed tagging and will produce many warnings.
I am sorry, that I have to disappoint you in that way, that it is even more complex as you might have thought. --Skyper (talk) 18:13, 14 February 2022 (UTC)

I am still not convinced about oneway:lanes=* and oneway:lanes:conditional=* but I have not found a better replacement. I guess oneway:lanes=* should not be used but rather lanes:backward/forward/both_ways if possible. At least, I would expect to find a oneway:conditional=no @ (Su,PH 07:00-14:00) to overwrite the general oneway=yes and only use an additional oneway:lanes:conditional=* if there are different values per lane. --Skyper (talk) 13:32, 2 March 2022 (UTC)

yes or designated as access values

Sorry, I am not 100% familiar with access restrictions in Columbia, but I think yes and designated need to be rechecked. As long as there are no additional signs, I would say yes is the right choice. With signs I would use designated for the signed transport modes. If bicycle are allowed on the lanes all the time, it is not designated for motor vehicles (only use yes) but if the lanes are designated for bicycles, pedestrians or buses (only at some time), I would use designated as value of the access tags respectively the conditional access tags. --Skyper (talk) 13:48, 10 February 2022 (UTC)

Could you please expand on this specific issue? Perhaps by adding a few examples to help illustrate your point on this would help too.
--JAAS (talk) 19:27, 10 February 2022 (UTC)
In Germany, we have a strict separation between the two values. As roads only for motor vehicles have their own tags, either highway=motorway or motorroad=yes, you will not find many motor_vehicle=designated, even with :lanes postfix. Even pedestrians are allowed on the road but should use the sidewalk if possible. If the sidewalk is blocked for some reason or you are pulling a hand-carriage which does not fit on the sidewalk, you are always allowed to use the road. Similar is true for bicycles with a cycleway=track or cycleway=lane.
designated is used e.g. bus or bicycle lanes marked with the blue signs, DE:Tag:traffic_sign=DE:245 or DE:Tag:traffic_sign=DE:237. These highways or lanes are exclusive unless other traffic modes are allowed with DE:Verkehrszeichen_in_Deutschland#Einspurige_Fahrzeuge_frei_DE:1022 or DE:Verkehrszeichen_in_Deutschland#Mehrspurige_Fahrzeuge_frei_DE:1024 or lanes marking allowing to use a short section of the lane for e.g. turns. As yes is the default, I have to use no for forbidden traffic modes on the exclusive lanes but can drop yes in all other cases unless a higher level of access tag like vehicle=no is tagged.
Ok, I try to get to the examples. First question would be, if the lanes are in general mode exclusively for (motor) vehicles and pedestrian are not allowed under any circumstances? The same for bicycles, are they restricted to any lanes? In my opinion, I do not see any vehicle=designated or vehicle:lanes|designated|designated|designated but only designated lanes for bicycles and pedestrian.
E.g. "2. Restricción sin desvío para automóviles"
* bicycle:conditional=designated @ (Su,PH 07:00-14:00)
* foot:conditional=designated @ (Su,PH 07:00-14:00)
and "4. Restricción con un carril permitido" should be:
* vehicle:lanes=yes|yes|yes (or even dropped)
* vehicle:lanes:conditional=no|no|yes @ (Su,PH 07:00-14:00) or vehicle:lanes:conditional=no|no|designated @ (Su,PH 07:00-14:00)
* bicycle:lanes:conditional=designated|designated|yes @ (Su,PH 07:00-14:00)
* foot:lanes:conditional=designated|designated|yes @ (Su,PH 07:00-14:00)
but I might be wrong. --Skyper (talk) 16:38, 14 February 2022 (UTC)


no route relations as members of route relations

Please, do not use route relations as members of route relations. In order to split long route relations, we have type=superroute and in case of collections like Ciclovías de Bogotá we have type=network if network=* and cycle_network=* on the route relations is not enough, thanks. --Skyper (talk) 13:48, 10 February 2022 (UTC)

Corrected to type=network. I didn't know it, and it is the best fit.
--AngocA (talk) 17:46, 10 February 2022 (UTC)
Nice, route=* should be removed, too, as network=lcn already includes the information and we might get warnings from QA tools about incorrect value of type=*. --Skyper (talk) 18:42, 10 February 2022 (UTC)
Removed, I left name, network, operator and type in the parent relation.
--AngocA (talk) 01:18, 11 February 2022 (UTC)