I’m thinking that it might be an option to tag the steam train line with https://wiki.openstreetmap.org/wiki/Tag:service=tourism and exclude lines with such tags from the render style.

I have created issues #1 and #2 for this.

It seems like anything with the tags https://wiki.openstreetmap.org/wiki/Tag:route='train' or https://wiki.openstreetmap.org/wiki/Tag:route='railway' or https://wiki.openstreetmap.org/wiki/Tag:line='rail' is rendered as a train line.

Looking at the docs, https://wiki.openstreetmap.org/wiki/Tag:route='railway' should probably not be marked as a train line. It seems that the Openptmap styles have a strong focus on Germany.

Looking at the way you linked, it does seem to be part of one https://wiki.openstreetmap.org/wiki/Tag:route='train' relation (Steamtown Excursions). I would argue that rendering it as a train line is not wrong according to the tags. If you find a tag that would distinguish such occasional services from regular ones, I can think of trying to adjust the style.

Hey Hartmut, MyOSMatic looks like a quite useful tool, and I would love to integrate it into FacilMap one day to make maps exportable there.

Could you point me to where the Openptmap style is defined in the MyOSMatic source code? Does MyOSMatic use the original osmfilter file, or did you replicate the original Openptmap style in another format?

@Severak: The frontend connects to the server using socket.io, and the protocol is quite simple. Unfortunately it is not documented so far, but maybe you can get an idea by reading the code of the server-side part. If you need any more detailed information, let me know, I could also document the protocol if you need it.

@Marián: Yes, unfortunately the Nominatim reverse geocoding is neither very fast nor very clever. I hope to find a different provider…

@CloCkWeRX: Good idea, will add it soon.

Okay, I have activated some detailed logging, and it seems that there is in fact a dead lock somewhere.

Running more tasks at a time will not bring a speed improvement but only consume more memory as the limitation of speed is the OSM API.

The fairest policy would be to not have a queue at all but to process everything at the same time. The problem is that for an analysation to run, lots of OSM objects need to be kept in memory. That’s the reason why I implemented the queue, because this way the memory consumption is kept as low as possible.