Proposal talk:Runway=displaced threshold

Turning pad

With reference to Talk:Proposed_features/aeroway=stopway, should turning pads be considered together or separately? Although as of now, only 2 instances of runway=turn_pad.

\---- Kovposch (talk) 13:05, 10 November 2020 (UTC)

Yes, I think turn pads very much deserve their own tags. I am struggling a bit with a simple enough mapping definition though as they come in various shapes and forms. Sometimes they are just a circle extension to the runway pavement (Skiathos: https://www.openstreetmap.org/#map=19/39.17027/23.50104 ), then there is some that are fully marked with an ellipsoid driving path (Kristiansund: https://www.openstreetmap.org/#map=19/63.11584/7.85069 ) and then there's these 90 degree drops (Myambat: https://www.openstreetmap.org/#map=19/-32.37378/150.61342 ) that look like https://caasref.wordpress.com/runway-turn-pad-making/ - I haven't seen enough of these to come up with a tagging proposal, but my guess is already that these might work on both nodes and ways. I won't include them in this proposal, but think it deserves its own.--Claudius (talk) 08:49, 11 November 2020 (UTC)

Indeed, about this question, someone else realized that there are 16 aeroway=turning_circle instances to crudely tag this difference by copying highway=turning_circle's value (though some of them are used on taxiways).

---- Kovposch (talk) 13:36, 11 November 2020 (UTC)

Generally approve

As with Proposed features/aeroway=stopway, good job overall.

I do see a similar issue: I don't think the direction of the way should matter, since a displaced threshold can only ever be at the end of a runway. And if it does matter, it's sufficient to say it points "towards the runway it belongs to", rather than "towards the closest threshold of the runway it belongs to."

Regarding rendering, you suggest "Way rendered with width and opacity between runways and taxiways". Grammatically, I think this is missing a word or two: "Way rendered with width and opacity between that of runways and taxiways". However, I'm also not sure I agree. The displaced threshold is part of the runway; we've already established that. Why is it beneficial to render it narrower than the rest of the runway? In the common map styles, runways are just rendered as a large opaque rectangle (without any regard for the runway's width=*, annoyingly), which makes sense because those styles are mainly for people navigating on the ground by foot/bicycle/car; runways are a nice landmark, but the details of a runway are irrelevant to most people. Then again, we render taxiways in a different width and that seems reasonable, so maybe given a lack of other visual distinctions there is some merit to graphically depicting a displaced threshold distinctly, too. --BigPeteB (talk) 20:57, 1 December 2020 (UTC)

As mentioned in the stopway proposal: Agreed to simplify the orientation/direction description.

In regard to rendering: You are totally right. Should render same style in width and opacity as runways (with a potential bonus rendering to mimic the arrow markings).

I've updated the proposal accordingly. Thanks again for the time to review and share your feedback. --Claudius (talk) 09:37, 4 December 2020 (UTC)

These stretches of runway inherently have a directionality; very clearly marked by the customary arrow markings. To give renderers a way to use that information, I would suggest keeping the way's direction pointing towards the runway proper and documenting it as such (kind of like how we map a oneway street). The example below gives an idea of what renderers can do with that information: I use the way's direction to have the arrows point in the right direction. It's easy to include in the tag's documentation now and not much of a burden on mappers; much harder to retroactively fit it in later. --JeroenHoek (talk) 13:52, 14 June 2022 (UTC)

Rendering experiment

I am experimenting with rendering this feature along with the runway proper based on its width=* over here. This is just a personal experiment, so I don't expect Carto will render it like that at the moment, but it might be interesting to have a look at. --JeroenHoek (talk) 19:38, 12 March 2021 (UTC)

Disagree

Excuse me for rowing against the stream ;) I have long been thinking about a consistent and coherent ruleset for the mapping of runways, and the more I think, the more confused I get. Still, a few points are clear to me:

• we should not be too eager to map details that are only relevant to a specialised minority; displaced thresholds are not relevant to the public at large. Some hold that anything may be mapped, others (especially in Germany) would rather limit mapping to generally useful information. The whole discussion is not super-relevant, IMHO;
• we should map in the first place what comes most readily to the eye, as visible on the ground. In the case of a runway, that is its total area;
• subdivisions of the runway can (and, for me, should) be handled as optional extra information; the more so that this information is rather volatile - it only requires a brush and a pot of white paint to move them around (not really, of course, but I hope my point is clear).

My approach would thus be:

• to map the complete runway in its full length, including stopways and displaced thresholds, and what not, not from threshold to threshold as currently recommended in the wiki;
• to optionally map displaced tresholds, stopways, and perhaps more such, as markings on the runway, much as a holding point is marked on a taxiway.

The worst idea of all is the current recommendation in the wiki with aeroway=runway+runway=displaced_threshold! By adding a tag "aeroway=runway" it suggests the presence of a second runway, apart from the main one. That is obviously not the case. The previous suggestion "aeroway=displaced_treshold" is at least coherent, though I still find it needlessly complicated.

Jan olieslagers (talk) 10:25, 19 June 2022 (UTC)

A highway=residential cut into two or more ways because the tags differ (extremely common) is still conceptually one street. The same holds true for aeroway=runway. Also: you can't map the displaced runway as markings without cutting up the runway way (unlike the holding points, which can be mapped as nodes or separate ways perpendicular to the taxiway). These markings apply to a section of the runway way, not to a node. This proposal is entirely consistent with how we map other linear features.

That argument turns up again and again, but it is not really applicable: linear features, such as roads or railways or canals, can be split up in sections, which can then be joined (in OSM relations) to create routes. A driver, or her/his route planner, will plan a route by joining route segments, similar for a yacht captain. A pilot OTOH will plan a route as a list of waypoints - which will by the way normally not include runways. I agree that the similarity to highway & railway has a certain degree of poetry, but it does not go any further.Jan olieslagers (talk) 11:45, 19 June 2022 (UTC)

Displaced runways are not volatile at all. They change, but just about as often as the whole runway. That's not really an argument here. Besides, displaced runways are just about the most visible and accurately mappable feature there is due to their high visibility on imagery (i.e., “readily to the eye, […] visible on the ground”).

The mapping of aerodromes mostly consists of mapping details that are only relevant to a specialized minority. We might as well leave out most of aeroway=* and stop naming taxiways if avoiding that is a criterium. --JeroenHoek (talk) 11:20, 19 June 2022 (UTC)

I would quite support such a proposal, indeed :) Jan olieslagers (talk) 11:45, 19 June 2022 (UTC)

As with all matters of detailed mapping: you are under no obligation to map such details yourself, but others do have the freedom to do so if done in a reasonable and non-destructive manner. This proposal, together with aeroway=stopway, help more clearly define the concept of a runway, and formalize how to tag a runway's common components. I wouldn't even classify this as micromapping; more along the lines of lanes=* or sidewalk=* for highway=*. --JeroenHoek (talk) 11:58, 19 June 2022 (UTC)

I say again: there is no parallel between aeroway:xxx and highway:xxx. Highways are routeable - aeroways are not. Jan olieslagers (talk) 14:15, 19 June 2022 (UTC)

You also wrote “we should map in the first place what comes most readily to the eye, as visible on the ground. In the case of a runway, that is its total area”. For the total area, use area:aeroway=runway. This is about the linear way representing the conceptual runway (from threshold to threshold, as it should be), not the total surface area which may include stopways or blast pads. The Wikipedia article Grille_Chompa links to helps explain this. --JeroenHoek (talk) 11:28, 19 June 2022 (UTC)

The tag area:aeroway=runway annoys me above all else: in 99% of cases, it only serves to add one runway twice to our database, with identical information (especially after a recommendation in maproulette). This is in violent contradiction to our basic principle "one feature, one OSM entry". I will admit though that there are cases where it can add useful information. Jan olieslagers (talk) 11:45, 19 June 2022 (UTC)

Disagree with the proposed name

Displaced thresholds are not sections of runways - they are thresholds beyond which certain activities aren’t allowed (in this case touch-downs aren’t allowed before crossing such a threshold). So they are actually lines across the runway. Displaced threshold is like a runway end/threshold (a line across that marks the edge of a usable runway surface), but located not at the end of the runway but further down the runway surface.

FAA AIM corroborates this: “2. Displaced Threshold. A displaced threshold is a threshold located at a point on the runway other than the designated beginning of the runway. Displacement of a threshold reduces the length of runway available for landings. The portion of runway behind a displaced threshold is available for takeoffs in either direction and landings from the opposite direction.”

So the proposed practice of referring to a portion of the runway as the “displaced threshold” would be misleading and confusing. BodhidharmaI (talk) 06:51, 3 September 2022 (UTC)

Yes, the concept of the "displaced threshold" is indeed represented in real world just with the marked perpendicular runway threshold bar. The area behind that threshold doesn't seem to be named anywhere in FAA or IATA documents, but just described as "The portion of runway behind a displaced threshold". The wikipedia article https://en.wikipedia.org/wiki/Displaced_threshold refers to it in what seems to be colloquial pilot speak as "displaced threshold area". I used in this proposal what was already an in use tagging since I couldn't really come up with a better name.

Any suggestions what would be more descriptive? Added "area" to the value as runway=displaced_threshold_area didn't feel to add much. --Claudius (talk) 13:03, 26 September 2022 (UTC)

I wouldn't rename it at this point just because of a minor semantic disagreement. I think the current tag-name makes sense; Wikipedia does use displaced threshold in the sense this tag is using it. Personally, I would solve any ambiguity with clear documentation here on the wiki. --JeroenHoek (talk) 13:49, 26 September 2022 (UTC)

Quite opposite to JeroenHoek, I would rather rename it. In fact, I think the proposal above by Claudius to name it runway=displaced_threshold_area seems quite sensible. In fact, this also leaves space for marking actual thresholds (as points on runways) as runway=displaced_threshold without further confusion as to in what context "displaced_threshold" value means "area" and in what context it means "the threshold itself". BodhidharmaI (talk) 12:20, 6 March 2023 (UTC)

You should probably draft a new (mini-)proposal for that rename, because this tag is decidedly in-use, and several tools make use of it. This proposal was never brought to vote, but the tag was adopted under Any tags you like. Renaming it would affect the tags used, so it would need approval from the community. --JeroenHoek (talk) 13:03, 6 March 2023 (UTC)

I should think that, since the use of this tag was never approved by the community, neither should any changes to it be? Yes, the tag is in use, but that is not "official" use, given the lack of approval. People who use unapproved tags must not complain about subsequent changes. Jan olieslagers (talk) 13:08, 6 March 2023 (UTC)

Sorry, that's not how it works. Anyone can introduce and use a new tag as long as it is reasonable. This was done with this tag. Renaming it would require approval because unlike introducing a new tag it has direct consequences on data consumers, renderers (like OpenAirportMap), preset, map paint styles, etc. Bear in mind that some of the most popular and common tags are unapproved. --JeroenHoek (talk) 13:24, 6 March 2023 (UTC)

In regard to BodhidharmaI's comment regarding the name: There's already aeroway=threshold in use which is some times being used to tag the position within the runway of the runway thresholds - displaced or at runway end. So I don't see a conflict or possible misunderstanding between this threshold tagging and tagging the runway segment. --Claudius (talk) 10:18, 14 August 2023 (UTC)

Can we come to consensus?

After several months, no new arguments have been added, and my points have not been answered. I think it is now time to come to a consensus, and, for lack of counter-arguments, I think it should be as follows:

• we should map as "aeroway=runway" the full length of the runway, including displaced thresholds, but perhaps excluding blast pads and overrun areas, as suggested in the article pointed to by @Jeroen ("The runway is the surface from threshold to threshold (including displaced thresholds)";

• a displaced threshold is a mark on the runway, no less, no more; (some seem to interpret it as an area, a part of the runway, probably due to the arrow markings, but it is not: the arrows indicate the part of the runway between its beginning and the displaced threshold);

• the comparison with "highway=" does not work, and is not applicable, because roads (and also railways and waterways) are intended to link up, creating routes, and do so by nature; runways, at the contrary, and equally by their nature, do not;

On a side discussion, I still hold that aeroway=runway can, and ideally should, be mapped as an area; only this agrees with our own opening definition in this same wiki "A runway is an area..." and also with the citation above. The additional "area" tagging is normally not needed, is contrary to our own basic principles, and is a potential source of confusion and duality.

Can we set a date for concluding this discussion? Jan olieslagers (talk) 12:20, 11 January 2023 (UTC)

For area tagging, area:aeroway=runway can be used. I agree that this is something that can be more clearly expressed in documentation. aeroway=runway is linear by definition, in analogy to highway=*. --JeroenHoek (talk) 13:32, 14 January 2023 (UTC)

For the sake of simplicity I will suspend my case about "area" though I still hold that runways are areas and should be mapped as such. But that's not for now. Jan olieslagers (talk) 07:23, 22 January 2023 (UTC)

I think Grille_Chompa's proposal to map the displaced threshold separately is still valid (like this: https://www.openstreetmap.org/way/675887719). As I've shown above, this also offers renderers a way to treat these differently. The proposal also considered this: “The initial version of this draft suggested using aeroway=displaced_threshold instead. Feedback and current usage suggested that most airline industry refer to displaced threshold as being part of the runway”. I don't think switching this up again has many benefits. Like highway=*, there is no need for the runway to be a single section of way. --JeroenHoek (talk) 13:32, 14 January 2023 (UTC)

What "most of the airline industry refer to" is irrelevant - even if it were sourced or referenced. What matters is official definitions. A displaced threshold is a threshold, a line of paint across the runway, nothing more. Jan olieslagers (talk) 07:23, 22 January 2023 (UTC)

I haven't seen a good argument as to why moving to an area only model for runways would be beneficial over the linear tagging in regard to data quality and practicality of mapping. If you feel strong about moving all runway tagging to an area model I suggest putting together a proposal summarizing the benefits. Your first point raised is already covered by this proposal as the full runway length will still be captured in data as multiple linear ways. It could help to also tag those segments with the same ref as the main runway segment. For your second point that the "displaced threshold" is a point on the runway: Yes, everyone agrees to that fact. This proposal is about tagging the runway segment in front of that point differently as it functionally serves a different purpose and is visually quite different. Since right now only 5 wiki users have been discussing here I will move the proposal in the voting phase to hopefully get input and guidance from a broader group of OSM mappers. --Claudius (talk) 08:58, 16 January 2023 (UTC)

The runway segment between the threshold and the runway end is scarcely defined or even mentioned in any official document, and has no need to be tagged as such. Its visual difference can and should be handled by the rendering process, not in our database. This concurs with the remarks by Bodhidharma and Claudius higher up. Jan olieslagers (talk) 07:23, 22 January 2023 (UTC)

and, since @Jeroen and myself seem to be at quite opposite positions, it might make sense to discuss further, offering facts and arguments, before proceeding to voting? Jan olieslagers (talk) 17:30, 22 January 2023 (UTC)

Here's EASA's "Easy Access rules for aerodromes" document from June 2023. On and around page 233 it contains a runway diagram where the section between the start of the runway and the displaced threshold location is called "displacement": https://www.easa.europa.eu/en/document-library/easy-access-rules/easy-access-rules-aerodromes-regulation-eu-no-1392014 - The FAA document ( https://www.faa.gov/air_traffic/publications/media/aim_basic_w_chg_1_and_2_7-16-20.pdf ) does only describe this section, but doesn't use a unique name for it. So I would say that this segment is described but not named in all cases. EASA's naming could suggest a tag value of runway=displacement which we could consider for finalizing this proposal. Since this value would lack "threshold" in the name I think it would be less descriptive. --Claudius (talk) 10:37, 14 August 2023 (UTC)

After three weeks' patient waiting, I must say I am disappointed at the total lack of response. I am a bit lost now, uncertain about how to proceed. Should we not conclude that we can NOT come to consensus? Which would, as I understand, imply that the way of mapping displaced thresholds is at any mapper's discretion? I would find that disappointing, since I prefer uniformity in the database. Jan olieslagers (talk) 14:25, 5 February 2023 (UTC)

@Claudius: given the continued lack of response, we can only progress by passing to the voting stage. Please do! Jan olieslagers (talk) 14:48, 16 February 2023 (UTC)

Since no vote followed, and no consensus could be reached, the proposal must be considered as "not accepted". Jan olieslagers (talk) 10:56, 3 March 2023 (UTC)