[OSM-talk-ie] Railways

[Colm Moore]

Hi Michael,
Thanks for coming back to me.
1. I haven't been able to access http://product.itoworld.com/map/14 in months. ITO World seems to be down over this weekend.
Similarly, the railway layer on http://tools.geofabrik.de/osmi/ hasn't been update in months.
2. The objective is to show the potential of railway lines, both in the use of multiple tracks and the use of passing places. Number of tracks is a key indicator of railway capacity. Unfortunately, just like non-motorway dual carriageways, map rendering software isn't able to automatically work out how many railway tracks / carriageways there are, so tags are needed.
Note that the amount of double track is modest: http://www.bueker.net/trainspotting/maps_british-isles.php although there are many passing places that aren't on that map, due to scale. There are perhaps less than 20 km of 3- and 4-track sections.
Misguided or not, I had added tracks=number to the entire network, even where I had mapped both / all tracks, as this was supported by ITO World. Since my last email, I have drawn the second track and applied passenger_lines=number on most of Dundalk-Belfast-Bangor and Larne-Newtownabbey, leaving only Belfast Great Victoria Street to Newtownabbey to be checked. 

That other users were doing the same probably suggests what we were doing has some logic and that the 'official logic' might not be intuitive. So, as I go over the network the next time, do I change track=number to passenger_lines=number ? 
4. I will be adding railway:traffic_mode railway=signalrailway=switch

I have been adding railway=signal and railway=switch but I'm not seeing them on OpenRailwayMap. Does a full scheme need to be defined before they are included?
To date, the following have been added: http://stat.latlon.org/ie/latest/tags-r.htmlsignal (172)signal_box (13)switch (1063)
I suspect that is less than 50% of switches and only 5-10% of all signals have been mapped.
5. Railways in Ireland have timetable numbers, e.g. Dublin - Sligo is timetable 7, but not line numbers. The timetable number isn't very visible, but more visible since www.a-b.ie and and Ireland was added to Google Transit.
Stations have both station codes, e.g. CNLLY for Dublin Connolly and ticketing numbers (100 for Dublin Connolly). Certain stations and ports also have (British) National Rail codes for ticketing purposes.
Station codes seem to be in use on the signalling system.
Bridge numbers are in the style of UB_ (underbridge) and OB_ (overbridge) with _ being a line letter, e.g. "UBR 52" is the 52nd bridge on the Rosslare line. Older signs will only show the number, e.g. 52. Confusingly, there is a separate system of numbering for restricted bridges: http://www.irishrail.ie/media/bridgeheights1.pdf I don't know why bridge 52 was renumbered 264.
6.> [2] In Germany
> a railway line has – by definition – only one or two parallel tracks
> (outside stations). If there are more than two parallel tracks outsides
> stations (e.g. Augsburg–Munich), they are defined as two independent
> double-tracked railway lines.
That seems ... strange. :) If there are four tracks, do the northbound tracks have one number and southbound the other. Or are the inner tracks one number and the other tracks the other number? On Köln-Düsseldorf, there are **lots** of tracks.
7. 
> > 1. There are about 4,000 instances of tracks=number that need to 
> > change to passenger_lines=number. Can someone help me? A mass edit 
> > seems attractive, but can that be restricted to Ireland? 
> 
> I don't think that a mass edit is possible. Every track should be
> checked manually if track=* is right or wrong.

I have manually checked every single piece of railway except Belfast Great Victoria Street to Newtownabbey (10km). Does this mean the change can be made (if a mass edit is possible)?
Thank you
Colm
---------------------------------------------------------------------------------------------------------------
Never doubt that a small group of thoughtful, committed citizens can change the world. Indeed, it is the only thing that ever has. Margaret Mead