Show threadMoof! 🔜 #JetLagTheGame NLNov 6

*looks out the window* but it says “Valence TGV” on that sign!

One of the problems that was talked about at #OpenTransport2025 was quality and availability of data, especially with regards to #CrossBorderRail.

DB Navigator does have some real time data for French trains, but this Spanish train has no live data that DB can access, whether from Renfe or the various SNCF subsidiaries, despite the fact that 80% of this train’s run is in France.

As such, it doesn’t know I’m late and is running per the published timetable.

SNCF Gares et Connexions is showing us as running +30 mins late.

Show threadMoof! 🔜 #JetLagTheGame NLNov 2

At this point, I think back, and wonder if I'm doing something wrong. I decide to try my luck with a python GTFS munging library, and after an abortive attempt with partridge, I settle on gtfs_kit as something a little more robust.

I have a look through the API and finally work out how I can do things on my interactive prompt that make sense of the data, and I continue to query it based on route ICE 91, and find that I still have more than one actual set of destinations.

This doesn't seem right at all.

So I remember back to the talks I went to at #OpenTransport2025 and recall that Transitous does do cleanup of GTFS feeds, and maybe my data needs a little cleanup.

It's definitely possible to download the post-processed so-called DELFI file, and that claims to have the entirety of the DB timetable in it as well as a lot of busses... so I download it and stick it into gtfs_kit, and ask it to give me the timetable for ICE 91, and... I still get a whole bunch of different routes...

Show threadMoof! 🔜 #JetLagTheGame NLNov 2

At #OpenTransport2025 we had a significant number of full and frank exchanges of opinion about the quality and availability of data from service operators. In this case, I know that all European rail operators are required to publish their timetables on National Access Points.

So I googled National Access Point for Germany, and end up on https://GTFS.de

There I find a link to a GTFS feed for Deustche Bahn long distance, which is basically what I'm after, so I download that.

In my defence, it was 2am, a time I'm rarely awake, and I wasn't really thinking straight, but GTFS does sound like something that is useful for me.

So I unzip and look at the file, and... to my surprise it's not that difficult to make out what's there.

Moof! 🔜 #JetLagTheGame NLNov 2

Last night I got massively nerdsniped by @quixoticgeek and stayed up til 5am trying to work out how to plot a route in which I travel one of each type of currently active ICE (1, 2, T, 3, 3MS, 4, 3Neo) in order of entry into service in the run up to taking one of the first ICE-Ls.

The naïve way to do this is a breadth-first search based on the end station and time working backwards through the routes that have the right kind of train on them and discarding any nodes that don't have any connections to the next type of train. As it was early in the morning at that point, I decided it was a good start in terms of algorithm. Optimisation can happen later.

So, I decided to try to find actual data in order to try and create these routes, and, well, I can now start to understand a lot more of the frustration expressed at #OpenTransport2025.

citybikesOct 30

This month in Citybikes, issue 202510

https://eskerda.com/citybikes-changelog-202510/

Home assistant, scaffolding, duckdb-wasm and #OpenTransport2025

This month in Citybikes #202510 | eskerda.com

your daily cup of tea™

MITFAHR|DE|ZENTRALEOct 26

After @julle presented #DBInfraGo's #OpenStations dataset at #OpenTransport2025, I'm now matching the LiftEquipment against #OpenStreetMap... and learn about fascinating elevator projects like the #SchmidPeoplemover

https://de.wikipedia.org/wiki/Schmid_Peoplemover

(HB)

Moof! 🔜 #JetLagTheGame NLOct 26

@sam from what I gathered at #OpenTransport2025 Transitous is starting to consider NetEx, but if I understood it correctly, they want to translate that to GTFS, which seems like shooting yourself in the foot.

I suspect that it has partly to do with the fact that actually parsing NetEx/SIRI is complicated, I’ve found people significantly less capable of dealing with the complexity and nuance of XML, especially with the fact elements can be modified with attributes.

Programmers seem to want everything to be JSON, and for data to be as flat as possible. Which is why I suspect that the CSV-based GTFS is more palatable to many. “Is there a standard relational database mapping for NetEx?” Was a question actually asked at one of the talks, after looking at the increasingly complex nested structures that DB InfraGO was valiantly producing.

Show threadMoof! 🔜 #JetLagTheGame NLOct 26

@rail @darkphoenix I am quite enjoying viaduct on your recommendation. I still have some issues, but it very definitely helped to meet Michael from Viaduct at #OpenTransport2025 and see that he’s genuinely interested in improving the site.

I am still using trainlog.me for my route maps on fedi, because it makes it easy to share, and includes things that aren’t trains, but I am putting as much as I can in viaduct as well…

Moof! 🔜 #JetLagTheGame NLOct 26
@sam this was somewhat of a theme at #OpenTransport2025. However, it definitely needs highlighting, both because people need to know it doesn’t work, and because activism needs to happen to improve data quality.
Volker KrauseOct 25
My impressions from #OpenTransport2025: https://volkerkrause.eu/2025/10/25/open-transport-community-conference-2025.html
Open Transport Community Conference 2025

On Friday and Saturday last week I attended the first edition of the Open Transport Community Conference in Vienna, Austria.