because you know them all, or how do you check whether there are still visible remains?
Yes, I know all the railways. /s
Obviously Iām just removing railways that are clearly paved or built over or otherwise completely removed. Youād be surprised at how much rail you find on OSM that is now just farmland, roads, parking lots etc etc with no visible remains.
I disagree. It is an interesting fact that there was a railway once, but that fact does not belong into OSM where we record what can be verified on the ground - and with bits and pieces of a former railway line on the ground, you need external knowledge to say āthis was the so-and-so gauge passenger line that connected X to Yā. It is fascinating to research these things and I like looking at maps of former railway lines, but I donāt think that this āshould be mappableā in OSM.
If there are other areas that you believe are comparable to this, then let us discuss whether other data also needs to be deleted (or whether that other data is deemed important enough to make an exception), rather than using those as a reason to map non-existing railway information.
3 Likes
The problem is that this is that this goes down the rabbit hole because e.g. most valleys are shaped by now non-existing rivers. Itās much better to map younger changes in geography like railways and intentionally dried out rivers than dried out former natural rivers thousand years ago.
Orā¦ how about no?
3 Likes
There is no rabbit hole, we just map what is on the ground today.
(Actual rabbit holes excluded, as the wiki reminds: " You should not map natural nesting sites")
2 Likes
where does it stop? If you see a yield sign you need external knowledge to say it is a yield sign, otherwise you just see a white triangular traffic sign with a red border.
This is true for most things we map, if you map a street cabinet you typically need external knowledge to say what kind of cabinet it is, and also to map a power plant you need external knowledge. Some people are interested in power plants, others in modeling bicycle related legislation, or historic remains, or remains of former railway infrastructure. Just because some railfans have overdone it and added things that are gone without a trace (probably happens in other fields as well), there are now some autodeclared ghostbusters who delete whatever railway piece that is tagged as not functioning, and DWG keeps a hands-off approach
4 Likes
This debate? Never. It will keep going even after every single OSM server has been switched off and every copy of planet.osm been lost to posterity. After the world has burned up in climate change-led heatwaves. After the AI apocalypse. After the regional wars of the 2020s ignited into WW3. After nothing is left but three primitive micro-organisms subsisting in a freak temperate micro-climate somewhere in the high Himalayas. One of those organisms will be painstakingly etching long lines across the cave walls with Cyrillic graffiti. The other two will be arguing about whether dismantled railways should have been allowed in OSM.
16 Likes
You have an interesting but very theoretical point. If I see a street, and there are two street name signs on either end, then I make the informed guess that the bit in between is probably the same street with the same name. A railway nerd might see two bits of track with a certain curve radius and a new settlement in between, and conclude that there clearly was a railway there, and I will tell them to stop mapping guesswork razed railways under new-built houses. But how small must the gap between the two bits of track be to allow a continuous route to be drawn? This is an interesting question.
I think philosophically we need to be aware that weāre discussing two different concepts here. One is āthingsā that physically exist, like a street cabinet or a sign. If you donāt know what they are, you can at the very least record that thereās something there with certain characteristics because you can use your senses. The other is āideasā, like a former railway line. A former railway line isnāt a thing (any more); it was a thing. Mapping something that was is fundamentally different from something that is. Natural language allows you to say āthere is a former railway line hereā but it would be more correct to say āthere was a railway line here, and now there is something elseā.
I am in favour of mapping things that are, not things that have been.
4 Likes
A railway is not just tracks (as in iron and wood/concrete) but many opponents argue if you cannot see rails it must be removed. I have occasionally encountered a former railway while walking in the forest, and there werenāt tracks, but there were clearly remains of a railway, you could see where the tracks had been, there was still the ballast on an embankment, a bridge and the remains of a signal.
Regarding your question, I do not advocate mapping a course where there are now houses, and I think for the peace it would be better if former railways were not continuous but only where actual traces can be seen, but this doesnāt mean there have to be rails, it is sufficient if you can see where the railway was.
this could easily refer to two different situations: the first when the former railway is perceivable, and the second when it is completely gone. And it depends whether this something else is completely unrelated (like new houses) or builds on the remains (like a cycleway).
4 Likes
Very interesting thesis. If we follow this reasoning and see a cycle path at the end of which there are information boards about a former railway line, the cycle path itself bears the name of the former railway or is called Railtrail XY, then I can just as well assume that the route in between runs on a former railway line. I donāt even have to make any major assumptions about the course of the railway line. But it is precisely such railway lines that are deleted without you taking action against them - contrary to your own thesis.
Others, on the other hand, claim that only what you can actually see should be mapped. If you follow this, we only map street names as poi with name_sign=, just like all the traffic signs. The highway itself then no longer has maxspeed= or name=* on it (unless it is painted on the ground) because the fact that the signs are assigned to the road in question is only an assumption, a guess by the mapper.
translated by deepl
1 Like
At the risk of inserting myself into a loooong discussion with some very strongly held opinionsā¦
A railway isnāt just tracks, but is a railway a railway without, yāknowā¦ rails?
1 Like
This discussion isnāt about railway=rail (where situation is hopefully crystal clear to everyone that is very welcome always to map in OSM!), but instead about lifecycle-prefix related tagging (and its alternatives) of things that are nowadays less then fully functional railway with rails and everything.
E.g. much like pile of bricks in roughly rectangular shape definitely isnāt a building=house but it very well might be ruins:building=yes, so a railway remains without actual steel rails might still be abandoned:railway=yes (or railway=abandoned, whatever); e.g. this:
or this (with just wooden sleepers remaining, and all metal long scavenged):
Or even this (hopefully you have a sharp eye!):
Sunny, yes. But I fail to see the connection between ālovelyā and āsunnyā.
Sun in the Solar maximum cycle, highest #ClimateWarming heatwave in all recorded history approaching its most deadly to all living creatures (except apparently mosquitos, which are waiting in ambush in every little shade in record numbers), official repeated national warnings of avoiding leaving house between 10am and 5pm etc; isnāt really something Iād really call ālovelyā even in current state of dehydration threatening loss of higher brain functions.
ā¦ so, razed:railway=* discussion it is 
3 Likes
Youāre going to have to help me; I donāt see any railway in these photos.
Railroads also have ballast, signs, electrification wires, sleepers, protected road intersections, dedicated bridges etc etc. If a few of these are still left with the railway still being verifiable, but the metal rails themselves are gone, I would have little problem with it being mapped with a lifecycle prefix. As I said before the railway=razed tag is redundant since we have the lifecycle prefix.
But cycle paths, information boards and specific street names are not characteristics of rail infrastructure that actually still exists, so thatās all disqualified.
I though I did, with link to that lifecycle prefix wiki? Please read it. Weāre not talking about (working or complete) railway here, but about *:railway tag ā where lifecycle prefix (indicated by *:) drastically changes the meaning.
I.e. in case of those particular images and railway=abandoned / abandoned:railway=yes tagging, the meaning is āpartially dismantled tracksā which means that a part of tracks (like metal rails in those pictures) is removed, but part (like sleepers) still remains visible on the ground.
To draw a parable, looking at e.g. a picture of street light tagged with highway=street_lamp and saying āI donāt see any highway thereā is problematic, because one has to consider whole tag (whole key=value pair) to get its meaning, not just some arbitrary part of it.
So if I understand correctly, you have no problem with razed:railway=yes, but do object to railway=razed tag? Lifecycle prefix wiki discusses it and explains that railway=* predates lifecycle prefix but means the sameā¦ So to me, both are synonymous (and thus both are fine by me), although one might do well to use one with better support & higher usage.
Youāre going to have to run this by me again, because I donāt see āpartially dismantledā anything in those photos. Well maybe the second one, at a stretchā¦ But this one:
Whā Whereās the ārailwayā? I see patches of grass and a little old bridge with someā¦ concrete and stones? I donāt see the means by which a train/streetcar/minecart/whatever can travel across this space.
5 Likes
It still depends on the situation of course, I wouldnāt map railways that are completely gone or of which only ballast remains hidden under new vegetation or something, regardless of tagging. But in cases where it makes sense to map the remains of a railway I prefer consistency with the other lifecycle tags.
(Iām going to assume youāre asking in good faith once more)
Nowhere. Why do you keep talking about it, even after it was explained several times now? There is abandoned:railway=yes, however. Have you read the wiki links Iāve shown you? Did you manage to grasp the difference between railway=yes and abandoned:railway=yes? They are not the same tag, and they donāt have the same meaning. Understanding that is crucial prerequisite to even attempting to read the rest of this thread.
old bridge with someā¦ concrete
Yes, that concrete part (on the upper part of the bridge) is part of the original railway infrastructure; it used to support metal rails (which had since been removed). Thus āpartially dismantledā part, i.e. abandoned: lifecycle prefix.
I understand it may not be instantly recognizable to every layman, but it is pretty obvious to anyone with interest in the stuff, even without local knowledge. (which is also why I think people who do not understand what some specific tag means should not go around willy-nilly removing them, but instead ask original author in changeset discussion about it)
Youāre not supposed to see that (as you keep using wrong tense). If you currently see a means by which ātrain can travel across this spaceā, that is to be tagged with railway=rail (or maybe disused:railway=rail depending on train company schedule/plans) instead!
abandoned:railway=yes however means that train could have moved over it only in the past, while that railway infrastructure was not yet āpartially dismantledā, but that it can no longer do so.
Much like your car likely wouldnāt be able to move far after it was partially dismantled (i.e. had tires removed). (or see my previous example or rubble of stones not being a building, but still being tagged with appropriate lifecycle prefix).
I hope that clears the confusion?
10 Likes