inexorabletash
Feature Request: "It is not possible to display custom data on the main map itself."
Please consider changing this.
I use travellermap.com live in my game, except when I need to add custom data. Then I'm reduced to paper handouts. It makes me sad.
Alternatively, how hard would it be to port to Nginx/Mono? ;-) Actually, I'd consider contracting someone to do this.
Maybe a browser extension could be contrived?
Looking at the code, there is one reference to a PDF dll, so migrating to Mono might need some work on that. Probably doable, there are plenty of good PDF libraries out there in the wild.
I would also like some form of Docker deployment for Travellermaps, so people could create their own variants easily.
Anyone tried to port this project to Mono yet?
No need to port to mono. Just update it to modern .Net which is cross platform.
.Net isn't the issue with getting TravellerMap to run on Docker. It's the reliance on IIS, which is a bigger hurdle.
I tried a port but made no real progress. I don't know C# all that well. Likewise, all the contracting options I found declined or clearly didn't understand the task.
Haven't worked with IIS in the last 25 years - but is it bundled with the app or just a reverse proxy? I guess bundled, otherwiese it would be quite easy. I see that C# does offer a "self hosted" server, but I haven't done C# in the last 10 years either, so I am unsure where to begin.
Last time I looked, TravellerMap was using ISAPI calls for all the URL handling, which isn't a fun thing, nor easy to port.
Best way forward is to reimplement the web api front end. I also can't remeber if the rendering was using directx graphic calls or not.
Thanks for pointing this out. Argh, I see. ISAPI... Memories 😱 Right, these will not be completely easy to port.
I'm an IT Architect, not a full stack developer. I actually looked at TravellerMap with Josh with an eye to porting it to Azure web apps. ISAPI was the major roadblock, with the tileing engine being a secondary.
I'm pretty sure the ISAPI calls can be eliminated with some work and converted either to full .Net methods or to another webserver like nginx. The tileing/rendering I didn't understand enough of to be able to offer an opinion.
My current opinion is that the code needs to be ported to either kubernetes or podman using multiple containers. One for api requests, one for rendering, one for data requests, one for pdf. Perhaps others. That way it can scale better.
I am a full stack developer with a strong tendency towards architecture and I would generally agree. That said, I am not doing and C# any more (ten years ago, yes), so I am not too confident with the language any more.
I guess my first step is a local windows service and gradually factoring parts over to nginx on Windows. I hate windows, BTW. Certainly not promising any real progress, but I'll try.