Railways: Draft Manual

[Duke87]

Does the Akita Shinkansen start in Tokyo, or does it start in Morioka?  I would assume Tokyo with a concurrency to Morioka.

I agree with your assumption.

[nezinscot]

Most metro stations in Japan have a station code that is based on the metro line that they service. These codes are listed on route signs and maps so they are known to the public (and from experience, are really useful if you can't decode some of the more esoteric station names).  Could or should a station code be incorporated into a station's waypoint name? 

I'm torn, because I would prefer not to, but English versions of Japanese place names when abbreviated all too often start with the same syllables -- e.g. Tak, Tok, Fuk, Shi -- so station name abbreviations can be very similar. The station code would help disambiguate the station name abbreviations.  (The ideal solution would be using the station name kanji as the waypoint name because they are unique for that line and already short. Will multi-byte character text ever be supported?)

[michih]

The station code would help disambiguate the station name abbreviations.

Our users can simply use the .list tool editor in showroute.php. Nonetheless, I have no strong feeling if we should add station codes or not.

[Jim]

Will multi-byte character text ever be supported?

I have no objection but also have no time available to make it happen.  I think with the .list tool, we wouldn't need to worry about people having to type in the non-English characters if they did not want to/know how to do so.

[Duke87]

I mean, SOP already accomodates this situation by allowing additional letters to be added to avoid duplicate labels, though I agree with a lot of would be duplicates this gets messy - which one is Osa and which one is Osak anyway?

One potential solution would be to interpret "first three letters" to mean the first three letters before transliteration into the Latin alphabet. So if you have a station named Fukushima, it would be labeled Fukushi. This reduces collisions while keeping label lengths reasonable.

I think with the .list tool, we wouldn't need to worry about people having to type in the non-English characters if they did not want to/know how to do so.

Ehhhh anyone editing their list files manually is still going to find it an annoyance to have to handle that though. I would prefer to continue sticking to the basic Latin alphabet for anything that goes in a list file.

Also note that other fields like system name and city name do already support non-ASCII characters (e.g. "Iceland Þjóðvegir").

[si404]

I'm using the signed station codes in Singapore, which Tokyo seems to have copied the idea from. The important bit is that the code is signed.

Here's some Tokyo signs - pretty clear they are signed as part of the station name.





Codes are perhaps an unhelpful term - they are station numbers and the analogue is junction numbers - which in highway systems trump any name-based label for waypoints.

[neroute2]

Yeah, those look analogous to exit numbers.

[nezinscot]

Thanks to everyone for their comments.  I'll use the station code when available (mostly metro and commuter rail), and try using the English version of the first 3 or 4 kana of the station name for stations lacking a code.  Three kana will be 6 to 9 characters.  For example かごしま (Kagoshima) would abbreviate to Kagoshi.  Or I could use 4 kana and the entire name could be used.  A lot of station names are four kana in length.

My next question is about named shinkansen routes.  The Tohoku Shinkansen has four named services.
 - Nasuno, a local from Tokyo to Koriyama
 - Hayate, a local from Morioka to Shin Aomori
 - Yamabiko, a rapid from Tokyo to Morioka
 - Hayabusa, an express from Tokyo to Shin Aomori

All the services run entirely on the Tohoku Shinkansen line.  Hayabusa travels the entire length of the line.  Should there be four entries for the Tohoku Shinkansen, one for each named service, or only one entry?

[Duke87]

All the services run entirely on the Tohoku Shinkansen line.  Hayabusa travels the entire length of the line.  Should there be four entries for the Tohoku Shinkansen, one for each named service, or only one entry?

One entry that includes every station is sufficient. Mapping distinct local and express services is only necessary if they diverge at some point.

[si404]

All the services run entirely on the Tohoku Shinkansen line.  Hayabusa travels the entire length of the line.  Should there be four entries for the Tohoku Shinkansen, one for each named service, or only one entry?

One entry that includes every station is sufficient. Mapping distinct local and express services is only necessary if they diverge at some point.

But, at the same time, if they have separate designations, they may be mapped individually.

For instance the C in the NYC subway doesn't diverge from the A (which runs express), but we map both.

[Duke87]

Propose adding the following to the "scope" section:

- train services that are subject to access restrictions such that the general public cannot casually just show up and ride them should not be mapped. Examples of access restrictions include trains contained within government facilities that may only be ridden by employees and authorized visitors, and trains located airside (beyond security) at airports such that they may only be ridden with a valid plane ticket to or from that airport. For the purpose of this rule, the need for a pre-reserved train ticket is NOT considered an access restriction, nor is the need for travel documents (passport, etc.) on trains that cross international borders.

Currently the draft manual does not address this issue. There has been some controversy about it. In the interest of having a written rule about it so the project can move forward, this proposed rule aims to simply codify the status quo given the lack of any consensus in favor of altering it.

[si404]

Currently the draft manual does not address this issue. There has been some controversy about it. In the interest of having a written rule about it so the project can move forward, this proposed rule aims to simply codify the status quo given the lack of any consensus in favor of altering it.

It's sort of like the highways and people wanting a 'pure' thing of State/US/I- only. I feel that airside / Capitol Subway probably should be provided at some point, but as a 'bonus' style system, rather than a core thing. Other than the Capitol Subway, the access restrictions here seem less than the roads clinchable on military bases (though, tbf, we have tended to debate those and minimise to those where you'd have an reason for the guard at the gate beyond 'I want to clinch this road'). We also have bits of road mapped that you need a ferry ticket to drive.

As for stuff where you need ID to travel - I think the Spanish AVE trains have it (most don't cross the border, but even those that do never leave Schengen so wouldn't require passports for the actual travel - it's security theatre after the Madrid train bombings), so perhaps a bit of clarity there.

Also, when I went to the Zoo a month ago, was I not an authorised visitor - they let me in (after a bag search) onto their private land, and had the right to turn me away? Wasn't my valid zoo ticket an access condition to ride the train (at extra cost, so I didn't*) in the way a plane ticket is an access condition to ride airside people movers?

I have no problem with the status quo, but would like a bit crisper wording of what we're not/not yet including.

*A bit annoying as I can't make out from sources whether there's a second stop and thus something mappable. But not that annoying as all you get to see is some deer things a bit closer than if you were walking around.
---

Looking back through the thread, the only outstanding issue where there might not be consensus that I can see is the 3 letter labels, but the implementation of longer ones wasn't seen as that useful by the person who implemented them on the systems they did so:

It's crap for metros and S Bahn services which mostly run through one city. My city district proposal really sucks.
I've (mostly) implemented it for HH S Bahn now but not for HH U Bahn (metro).

https://github.com/TravelMapping/RailwayData/pull/15/commits/efabe446010a5470e765f6ddb317785754e931c7
https://tmrail.teresco.org/hb/?u=michih&sys=deuhhs

It's different for long-distance services which usually have one stop per city / town only. I don't have such route in Germany though.

[michih]

I do still sit in the 'let's map the routes by infrastructure, not by service' boat. I found several supporters on that last year: panda80, bubaclex and bartpetat. They agree with me that the service approach is not reasonable for Germany, Ex-Yugoslavia or Romania when talking about regional or national connections.

[Jim]

For the creation of the TM Railways manual, how much overlap is expected between this manual and the TM Highways manual?  This answer will determine if we should have some PHP conditions, like we do for the main web front end code, for the manual, or if an entirely new set of files should be created.

[Duke87]

What's currently proposed is a completely separate document. Not sure where opportunity for overlap would be.