Station as part of the (rail)way?
- So the node of the station should not be part of the ways? Why not? --Bahnpirat 14:16, 8 December 2008 (UTC)
- If a station has more than one railway, on which railway should the station tag be assigned? A station is not a physical part of a railway. If you define train routes, then the station will be part of the route and that is the only needed information for routing and displaying. ck3d 17:37, 8 December 2008 (UTC)
- I agree. What most people probably want to find their way to, by car or by foot, is the main station entrance. If we say that the station is somewhere out in the tracks, we risk that turn by turn navigators will lead people to some road near the tracks, but not where the platform is. Perhaps it is even on the wrong side of the tracks. That is dangerous! I think it makes a lot more sense to put railway=station near the entrance. Or perhaps somewhere in the middle of the station building where the biggest information board is, if the station is big. Thats where the station is for me. --Henriko 21:18, 10 May 2009 (UTC)
Did you ever settle on anything regarding this? Not being a developer I'm not sure how navigational software would see this. If i were developing a city navigation system, I'd want it to work something like the tfl (Transport for London) route finder. Including walking directions and train directions.? Martin Renvoize 14:40, 9 December 2009 (UTC)
- All good points but a bit one-sided... There is of course huge value in placing the node as part of the way, which is why most mappers have (I believe) ended up mapping it this way. Using the data to render, or otherwise understand a rail network map. If there's no node on the railway, this becomes quite tricky.
- Some related discussions that I can remember: Talk:Tag:railway=subway entrance#Several railway lines going into a large railway station and Talk:Railways#Tube stations and tube lines. But overall it seems like a general problem with our strategy (or lack of strategy) for tagging and handling railway networks. There's several difficult problems in around this. Not the kind of thing we do very well at solving in any consistent way, but we can always try. Maybe we should discuss this somewhere at Talk:Railways. There is also a talk-transit mailing list.
- -- Harry Wood 13:25, 15 December 2010 (UTC)
UK station three letter references (CRS codes)
I notice that http://www.nationalrail.co.uk/stations_destinations/ uses a three letter code for stations. Is there a list of these we could use? Is it useful to add them to OSM? --LeedsTracker 19:32, 26 January 2009 (UTC)
- Oh look: http://deaves47.users.btopenworld.com/CRS/CRS0.htm Wonder what copyright status is? Would be good to add references to stations in the same way we add refs to roads or amenity=post_box --LeedsTracker 19:37, 26 January 2009 (UTC)
- I think this is very unlikely to be copyright free. Unless this information is actually displayed at stations somewhere (and hence we could actually go and find out anyway) - then it's likely been copied from something like National Rail Enquiries or similar. Post boxes have their reference numbers displayed on the box itself. Road numbers have information on signs Richard B 12:38, 27 January 2009 (UTC)
- I've added the CRS codes (ref:crs) from https://dataportal.orr.gov.uk/ (OGLv3) --DaveF63 (talk) 10:38, 30 January 2020 (UTC)
- If other railway facilities such as junctions, crossovers and freight-only stations had CRS codes and they were used for management of the infrastructure, I would recommend to use railway:ref=* instead. That tag is used accross countries to tag the reference code used by the operator of the infrastructure. OpenRailwayMap supports it (both rendering at lower zoom levels and search). --Nakaner (talk) 15:01, 30 January 2020 (UTC)
Station an area ?
- Good point. I have updated the page to give more details about mapping stations as areas. PeterIto 12:24, 22 April 2011 (BST)
- I'd still like some clarification on this matter. For larger stations/stations where the building is mapped, is it preferred to have a node with a station tag and the building tagged as building=train_station, or should we skip the node and have the building tagged railway=station? --Mike` 20:35, 27 April 2011 (BST)
- You should not map the building as railway=station, because railway=station as an area is seen as the area of the railway station (how obvious ;-) ), which is often from the entrance signal or at least from the first switch/point (AE/BE...) of each direction. --rayquaza (talk) 12:46, 23 July 2013 (UTC)
Former stations are currently being mapped in many different ways, including:
- railway=station; disused=yes
- railway=station; abandoned=yes
One could possibly add the following option:
Any suggestions one which one we should standardise on? To my mind the first is incorrect in that disused should only be used in circumstances where the facility could easily be brought back into service, and in the great majority of places where it is used the station and often the line itself have been removed. The last one, which is my proposal fits with some more general lifecycle modeling of railways and other features where any tag can be prefixed with 'proposed', 'construction' and possibly also 'historic' to indicate a upcoming or previous change to a tag value.
-- PeterIto 17:13, 1 April 2012 (BST)
- As I'm partly responsible for having mapped disused stations.
- This was my take on it ...
- railway=station; disused=yes - Buildings extant on active infrastructure, could be returned to use with minor works. (eg. West Watford)
- railway=station; abandoned=yes - Station extant on active Infrastructure, but unlikely to be reinstated (eg. York Road, Primrose Hill)
- railway:historic=station_site - There was a station, but you'll have a job to find evidence of it or the site's been redeveloped. (example Croxley Green LNWR )
- railway=station_site. (Example: Blake Hall )
- I've got no objection to historic:railway=station provided that historic:railway=station_site is also provided, and existing tag usage migrated which is used in the situation where the site is known to have been a 'station' (or freight terminal etc) but where the majority of what would conventionally thought of a station has gone, and it requires a degree of 'archeological technique' to find the remaining evidence...
- -- ShakespeareFan00 00:05, 24 April 2012 (BST)
- Thanks. I have reordering the list and clarifying a few things for my own benefit. I hope I have not damaged the meaning. PeterIto
- OK, so disused matches with railway=disused (ie not currently used, but easily could be) and abandoned matches 'railway=abandoned' (ie there is physical evidence but it isn't likely to be brought back into use). railway:historic=station_site is for what would be tagged as 'railway=dismantled' if it was a railway (which is one where it is probably impossible to see that there was a station/line). All makes sense to me. Do we need 'historic:railway=station' and 'railway=station_site' or are these identical to 'railway:historic=station_site'. PeterIto 08:59, 26 April 2012 (BST)
- To come back on this, I agree with you :) The differing tags seem to be hangover from various different tagging schemes. I'd say combine on railway:historic=station_site as appropriate.. ShakespeareFan00 18:06, 29 May 2012 (BST)
- Agreed, and I also like the use of 'railway:historic=rail' and 'railway:historic=tram' for railway lines (which I have seen used somewhere) more than 'railway=abandoned/dismantled' because it provides more detail about the former use which gets lost when using the railway=abandoned approach. As such possibly we should consider promoting the use of railway:historic for all of these purposes. PeterIto 11:33, 31 May 2012 (BST)
- "railway:historic=station_site is for what would be tagged as 'railway=dismantled' if it was a railway (which is one where it is probably impossible to see that there was a station/line)" in cases of railway leaving no traces - it must not be mapped Mateusz Konieczny (talk) 20:33, 3 April 2015 (UTC)
Cargo-only train stations
What about stations used only by cargo trains, without any passenger traffic? Is it proper to use railway=station for places like that? Example: http://www.openstreetmap.org/?mlat=50.0793&mlon=19.9604#map=14/50.0793/19.9604 Mateusz Konieczny (talk) 20:31, 3 April 2015 (UTC)
- railway=yard on a node (not an area) would be right. See the OpenRailwayMap tagging scheme. The area itself can be tagged landuse=railway and connected to the operating site node using an operating site relation. --Nakaner (talk) 08:56, 4 April 2015 (UTC)
The sentence "In addition, you can use the new public transport tag public_transport=station." is plainly wrong.
"The tag public_transport=station can only be added in the exceptional case when it should not be an area." approximates the truth much better.
For the discussed railway=station areas: the public_transport=station area is normally a different one as it gives the passenger view of the outline.
- public_transport schema is independent from railway=* It only duplicates existing tags, adding no extra quality to the OSM database. It shouldn't be referenced on this page. --DaveF63 (talk) 10:52, 30 January 2020 (UTC)
What does this sentence mean?
I have no idea what this sentence means or if it has any relevance to this page. "The area a passenger has in mind might not be mapped directly. Data users can create this area by calculating a convex hull around the stop area relation like OpenBusMap does." --DaveF63 (talk) 00:19, 6 February 2020 (UTC)
- If a railway station is mapped as a relation containing several other features, you can create an oval or circular area around it for rendering purposes with operations like ST_ConvexHull in postgis. This also works for a relation that has several bus_stop points or several railway_platform areas: you can make a blobby area around these. Since this is a technical rendering hint, it really does not need to be on the page, or it should be edited to something like "even if the railway=station is mapped as a node, map renderers and other database users can still create an area which includes surrounding features" - though I don't think that's really necessary for mappers to know? --Jeisenbe (talk) 02:19, 6 February 2020 (UTC)