Open Train Times site launched

[Poggs]

http://www.opentraintimes.com/ launched at 7am today.

Real-time data and maps will be coming in the next couple of weeks

 

[All Line Rover]

"historical data coming soon"

How will the website get that information? The train companies would want to keep it secret, I'd assume. Seems a bit dodgy to me!

 

[MidnightFlyer]

"historical data coming soon"

How will the website get that information? The train companies would want to keep it secret, I'd assume. Seems a bit dodgy to me!

That would be from TRUST surely? I'd like to know how they are allowed to get their hands on that (legally!).

 

[tsr]

Looks quite good to me.

For comparison, I might point you in the direction of http://www.traintimes.im/

I hope to be working on some similar projects in the next few years. Naturally, I am keeping any ideas secret for now.

 

[WestCoast]

For comparison, I might point you in the direction of http://www.traintimes.im/

I hope to be working on some similar projects in the next few years. Naturally, I am keeping any ideas secret for now.

May I ask why that website has an Isle of Man domain?

 

[Tom]

Best available that fit my needs

 

[LNW-GW Joint]

Seems pretty good to me.
Gives WTT info (headcode, routeing, recovery time etc) which I am not aware is available elsewhere?
Can't seem to set time of day (defaults to "now", then you have to shift an hour at a time each way).
The Highland Sleeper seems only to go to Inverness (maybe the Aberdeen/FW sections are considered separate trains).

 

[brompton rail]

Doesn't seem to be able to give real time running for trains en-route. E.g. 18.40 DON to LDS is shown as on time, yet other mobile systems say it is 6 late.

 

[Eagle]

(maybe the Aberdeen/FW sections are considered separate trains).

Split trains have separate headcodes... but it should be possible to get it to say "splits to form service 1B99 to wherever" at the splitting point.

 

[swt_passenger]

Neither site seems to fully deal with joins on route either - based on looking ahead at the Southampton to Victoria service. Tom's site explains it joins at Horsham in the detailed view, but both sites basic lists just show a Southampton to Horsham service initially...

 

[amcluesent]

Excellent idea if you can add history of late arrivals to confirm whether a delay\repay claim is viable on a journey

 

[Tom]

Neither site seems to fully deal with joins on route either - based on looking ahead at the Southampton to Victoria service. Tom's site explains it joins at Horsham in the detailed view, but both sites basic lists just show a Southampton to Horsham service initially...

Mine does deal with join/divides just fine in the 'direct journey search' - try doing Southampton to Clapham and it does show the through service. http://traintimes.im/services-between/SOU/CLJ/2012/01/18/

The reason why I don't put the false destination in on the location page is primarily down to how the data is set up. Searching for the association to create the false destination would take a LOT longer as I'd have to search for each service etc...

 

[bluenoxid]

How is this running? Has the data been released in a public format like promised?

 

[Tom]

The data for both sites is pulled from Network Rail for the raw timetable data which requires a server system for receiving full extracts/updates daily, etc.

Passenger trains only are available from http://data.atoc.org - has been available for about a month and a half now.

 

[All Line Rover]

The data for both sites is pulled from Network Rail for the raw timetable data which requires a server system for receiving full extracts/updates daily, etc.

Passenger trains only are available from http://data.atoc.org - has been available for about a month and a half now.

But historical information? Neither ATOC or Network Rail would release that, surely?

 

[Tom]

Historical TRUST data (poggs has access to TRUST data via TD.net right now, and it is being released publicly in April supposedly) can just be created by building up a present day profile...

So yes, they are releasing that

 

[Eagle]

I notice on Tom's site you can get all the intermediate TIPLOCs (whether station or not) that a train doesn't call at, and also they're presented in a human-readable format "Gibbet Hill Jn" rather than "COVNGHJ"; which is useful for me as there's a TIPLOC right outside my window :P

 

[willas00]

This site looks very similar to Toms including it's look o.O

 

[Eagle]

This site looks very similar to Toms including it's look o.O

Both created on Google's API developer, no doubt.

 

[Tom]

Both use Twitter's bootstrap css - http://twitter.github.com/bootstrap/

 

[Jan]

I notice on Tom's site you can get all the intermediate TIPLOCs (whether station or not) that a train doesn't call at, [...]

At Open Train Times, you need to turn on the advanced view via the settings at the top.

@Tom: In your detail view, the 'path' column should rather be shown to the left of the 'pfm' one, as it indicates the line/route taken approaching the location.

 

[Eagle]

Couple of things I've noticed in advanced view: the WTT timings seem to use a mixture of xx.xx½ and xx.xxH to denote xx.xx.30 timings, rather than xx.xx½ consistently. Also this ECS movement seems to have first class available... it's a 153 :P

EDIT: the NMT also has first class... well, it is a HST.

 

[The Planner]

That is just ancillary information on the schedule, we don't bother removing it for ECS and stuff like the NMT. I've no doubt that if you check a freight schedule it will offer reservations...

The ½ and H suggest that the data is coming from different sources, ½ sounds like the CIF from TPS and H direct from TRUST.

 

[Poggs]

The "H" is just where I haven't passed a time through the code which converts an H in to ½

Thanks to those of you who send in emails about bugs and new features, I've got a list as long as my arm of things to work on.

Watch out for Network Rail's public release of real-time train data in April, as mentioned in the Chancellor's Autumn Statement late last year.

Incidentally, none of the technology I'm using or have access to is in any way secret - Network Rail's TD.net platform already feeds ATOC's Darwin system, and I've had real-time TRUST, TD, VSTP and TSR data from that for the past nine months or so. There's nothing to hide here.

 

[umontu]

I quite like this site.
But why is the time always on 10:26 anyone know?

 

[the sniper]

This'll sound like a really noobish question, but I thought there was some kind of issue around freight train timings, yet this site has them if you put in the reporting number? For example: http://www.opentraintimes.com/schedule/H50007/2012/01/18

Are freight train timings now freely available too, or has nothing changed?

Thanks anyway Poggs. Your site will be handy to me for looking up freight times (before anyone says, I know these greatly vary from the timetabled times ).

 

[voyagerdude220]

I think it's a very good website.

It may not be perfect, but it's got far more info (like passing times/ECS movements etc) compared to any other website i've seen.

 

[BRX]

Will be useful if and when full freight working info is included.

 

[tsr]

I quite like this site.
But why is the time always on 10:26 anyone know?

I don't know, since I am not the developer, but I would hazard a guess that it's set at 10.26 because I seem to recall that's also the time that demonstration iPhone units used to be set to when screenshots of them are taken and put on Apple's website, so it may be a reference to geek culture. They have now changed this to different times of day.

If not, I have no idea!

 

[Poggs]

I don't know, since I am not the developer, but I would hazard a guess that it's set at 10.26 because I seem to recall that's also the time that demonstration iPhone units used to be set to when screenshots of them are taken and put on Apple's website, so it may be a reference to geek culture. They have now changed this to different times of day.

If not, I have no idea!

The answer is simple - or rather, it is now. To speed up rendering the front page, the web server caches it in memory. The problem is that the time field *should* be updated with the time you requested the page - and I made the change live at 10.26am.

I'm working on fixing that