I wonder about editors changing road, tram, subway or rail tracks into “operational”, although they have not been opened for commercial/public use already, just because they look like ready constructed.
Unluckily, there is no word about this in the guidelines/documentation about the key “construction”.
From my point of view, because of trial runs, open certifications etc. for railways, because of misleading navigation for roads, footwalks etc., it is not correct to tag traffic routes as “operational” just because they look like finished in construction.
Is there any guideline, consens or discussion about this issue?
From the professional point of view the construction period starts when a traffic way is blocked for users and it ends when the blocks are removed and the traffic way is opened for use. Before this date the traffic way (whatever it may be) may look “finished” but there may still be minor works, tests or certifications missing and as long as that is the case the construction period is still going on. This is the procedure in the real world and it should be the same in OSM.
As this is not explicitly stated in the wiki we should probably add a few words to make this clear.
Well, Mikke, in the recent case, I commented the edit this way, but the editor does not have open mind for this. I will continue into that country forum, thanks for the hint.
As a city and traffic planer, there is no doubt about this in my eyes, Mikke, and I do not want to have people running into trouble, because their navigation is based on OSM.
I would wager most people absolutely agree that the outward appearance of completion is only part of why a feature under construction should have the construction-related tags removed. To me it must be open and in use to have the construction tags removed.
There’s small, indirect mention of this in the highway=construction wiki page (bold is my emphasis):
If a road tagged highway=constructionhas already opened and is passable
(Oh bother, Mateusz just replied above as I’m typing this…)
I’ve also reversed changes made by others where they’ve removed construction tags on a road that isn’t open yet. Many other mappers are keen to keep things as up-to-date as possible, which is commendable, but if a road, a railway, a building, etc. is still closed because it’s still being tested and commissioned: it’s still “under construction”. That testing and commissioning phase is the last part of this kind of infrastructure construction, and it’s incorrect to mark it ‘complete’.
Sure, this may cause serious problems as we all know that some drivers blindly trust their navigation device.
In case of a road not yet open for traffic the most simple way is to just set it back to highway=construction manually (after contacting the mapper who set it operational).
Otherwise there are several reverting tools, like
Yes, I preferred the manual way until now, because there might be correct changes inside the changeset, too. And if You revert the whole automatically, this will be lost.
But a 160 km HSR route is a lot of work and I am a friend of the “polluter pays principle”. ;-)
This topic just came up on the Discord Oceania thread, about when to remove construction from a new international airport Discord & following. Opinion was yes, as construction is finished, test flights have been made, first scheduled flights are due late July & people are starting to use terminal facilities.
After I pointed this thread out, suggestion was made that as construction implies a construction site, maybe there’s room for a commissioning lifecycle prefix for time between constructed and open to the public?
Would certainly work for the latest extension of our local light rail, where they’ve been ding track testing for 2-3 months, all track work & stations are finished, & they’re currently only working on street resurfacing, landscaping etc.
My approach in these cases, where civil construction is absolutely complete but public access has not yet commenced, is to remove the construction tags, but add access=no.
this makes sense when it is certain that people will remember to mark it as open
I seen cases when someone tried to be super-accurate this way, then years after construction ended and it was open OSM still had access=no on some road segments which broke routing
while there is some QA support for highway=construction and similar
This discussion started because @AnnaCatarina asked me why I removed the contruction tag on the Cerkezkoy Edirne High speed railway. I did it because the contruction is completed and the line is used by works trains and the likes. I acknowledge commercial operation has not started yet. The avalaibily of trainsets is the main reason why commercial operation will not commence at once. We are in this grey zone where infrastructure is completed and visible, possibly used by freight trains but not in commercial passenger service yet.
In doubt, I reverted my change. On the great scheme of things, whether the tag construction is present or absent for this particular line do not change nothing to no one for the moment.
I also pointed out that lifecycle tag management is particularly difficult because the information is not readily verifiable by everyone. A good 30% of the rail network in Turkey is wrong from lifeclycle perspective. Even the open data published by the Turkish railway themselves is wrong on this matter.
Nothing to worry about, “polluter pays principle” is the common phrase in English, which is “Förorenaren betalar” in Swedish or “Verursacherprinzip” in German, not using pollution as a term, just to show, the person who is making a mistake is asked to clear it by himself and not wait for others to clear it.
Well, the key:construction runs together with key:opening_date and key:check_date.
So it is always clear, when it has to be checked again, if it has been opened already or not. Or if just the both dates have to be updated, because opening was postponed.
And commonly, operators are precise with an official opening date, if they know precise to the best of human judgement. Otherwise they write/talk about “end of 2026” etc. Which I generally tag as late as possible like “2026-12” in this case, where TCDD originally wanted to open this HSR end of 2022! Delays like this are in my eyes always a reason to be skeptical regarding any kind of announcements. But, for sure, not to know a precise opening date is no argument to change the key into “operational” far before.