For the last 10 days I’ve been travelling in Europe to go to OSM events. I have some spare time before my flights to get home to Vancouver, so I’m posting a brief overview of the events and what got done.

My first event was the Karlsruhe hack weekend. This event is held twice a year and attracts lots of German and Swiss OSM developers. I started out by discovering that the jet lag is worse as I get older and I missed the pre-event social.

At the event itself I focused on the Shortbread Vector Tiles specification and osm2pgsql. In total three people from the Shortbread steering committee were at the event. The main Shortbread coding that’s been completed was a cleanup of the tables in the 1.0 spec. I think this is the last change we’ll see to 1.0 and all my Shortbread focus is now on 1.1.

I haven’t yet published the changes but support for multiple languages in the OSMF shortbread vector tiles is basically done. I just need to clean up some changes and add them to the WIP commit. Initially the language list is only en and de, but it’s ready to be added to.

I met with the two other maintainers of osm2pgsql and we discussed what’s necessary to tag releases of osm2pgsql-themepark. I think we’ve got a route forward there.

I spent the Monday after the hack weekend around town. The next morning I started for Nottingham via the Eurostar train. While I was in Nottingham I visited relatives and didn’t do anything OSM-related.

On the trip north to Dundee I got a cheap seat upgrade on the train to Edinburgh. This gave me three hours of focused time to work on my SOTM EU presentations.

The first day of the conference was good. The talks I attended were - Human elements of trust and data misuse in open maps - Equal Access at the National Trust. The National Trust is doing good collaborative work with OSM and trails. I need to follow up with them about trail tagging efforts. - What’s New with our Website. The answer? Lots. Thousands of PRs were merged in the last year and development is moving at a good pace. - Servers on Fire: Keeping OpenStreetMap Online. This talk was by Grant, a fellow sysadmin, and was an interactive talk with “war stories” from running all the osm.org services. It was part of an effort to get more people interested in ops work, so if it sounds like something you’re interested in, please get in touch. I wasn’t the target audience for the talk because I had heard most of the stories. - Future of the OpenStreetMap Foundation discussion group. This focused on the funding shortfall. A full review of the session is more than I can do before my flight. - From Coast to Coast: OSM Coastal Landforms Tagging & Mapping Strategies across the Atlantic. This was a talk by someone from the Canadian Red Cross about mapping stuff near the ocean. I need to get in touch with the presenter because they’re a fellow Canadian - Shortbread Steering Committee meeting. Three of the steering committee members were at the conference. I wasn’t sure if this session would be just us three or more. It was way more, with probably 20-40 people. There was a lot of longer-term discussion but less on the issues that were holding back a 1.1 release. One misunderstanding during the session is that a lot of people thought that we didn’t want to put out a minor release with backwards compatible changes. This is wrong, because that’s what 1.1 is! There was also the common issue of software developers treating all technical computer projects as software development. Shortbread isn’t a software project, it’s a vector tile specification. We were able to iron out most of the 1.1 holdups in a couple minutes at the end.

After the talks we got locked into the building. When we found our way out we went to a local Weatherspoons pub.

I had two talks the next day, so I got some more work done on talk prep. The talks I went to were

• “Survey Me” - Using external data comparisons to flag issues in OSM. Lots of interesting tools are being run in the UK
• Using New Vector Tiles, one of my talks
• Make your own base map with MapLibre and Planetiler. I wasn’t the target audience for this workshop so I worked on my slides while following along.
• Minutely vector tiles deep dive, another one of my talks
• Everything Everywhere All at Once. This was about OSM Spyglass, a tool that is likely to replace the data layer on osm.org. You can see it at http://test.osm2pgsql.org/
• Overpass Turbo goes PostGIS. This was about postpass, a SQL-based server similar to Overpass
• Overpass API. This talk focused on the current issues the Overpass API is facing with limited capacity and the steps that are being taken.

My first talk was on how to use the vector tiles the OSMF makes. I’ll have a blog post version of this soon, but the example I walked through was building versatiles colorful to use the vector tiles, building a pmtiles file with wall data for the whole planet, and using Maputnik to add the wall data to the style.

The second was a deep technical dive into how the tiles are served. The focus was on the move from storing tiles to storing tile layers to improve performance. I felt less prepared for this talk, but people afterwards told me is was the better of the two talks I gave.

We went to a museum on the old jute industry in Dundee for the social. I stuck it out at the social until the hours of standing on a hard stone floor made me limp back to the hotel. I’m not as young as I used to be.

For the makeathon today (Sunday) we had all the ops people there. We had some internal dicussions about usage policy changes and preventing abuse of the OSMF-run services. Nominatim in particular is overloaded, but AI scrapers running through mobile device botnets are a problem everywhere.

We had a second meeting about how to get more people involved in ops. This is a bigger subject than I can write about before my flight, but if you’re interested in helping, get in touch. We have some tasks well suited to onboarding new people. We also had some discussion around longer-term planning.

All of us had been hoping to get more coding done during the makeathon. Spending most of the time in discussion meetings isn’t as fun but I’m glad we got the chance to sort things out face to face.

After the makeathon we went to a pub and had a fun time trying to identify the date of a map of Dundee that was on the wall. The map said 1982 but based on the lack of the road bridge and tram lines we think it was 1932.

I caught the direct train from Dundee to Glasgow and stayed at a hotel near the airport.