Shortest route on ESI longer than ingame route

Hi, im confused, maybe missing the obvious. But hopelfully somebody can help.

I want to calculate the route from Jita to Irmalin with the API.

The shortest route ingame, takes 35 jumps.
The shortest route on ESI is taking 47. It returns an array with 48 system_ids.

This is my api call:

https://esi.evetech.net/latest/route/30003935/30000142/?datasource=tranquility&flag=shortest

I tried on the EVE Swagger interface, this also returns me the same array. However, when I let everoute determine the route, they do get the same 35 jumps as ingame.

Anybody know what could be the cause of this?

See https://github.com/esi/esi-issues/issues/1285. Looks like there’s actually been a PR open for a while to fix it :grimacing:. I’ll see about getting that merged in tonight.

1 Like

Would be great if that be merged in tonight. How can you tell/determine if it actually happened?

Just wondering, how do dotlan and everoute do this then. They seem also to use the same API.

Not sure if it’s the case here, but the ingame route can take ansiblex jump gates that are accessible for your character into account (if you tick that box).

I tried fiddling with that, but didnt make a difference. And on dotlan or everoute im not even logged on.

I updated the PR for latest data and approved the PR. However there’s probably some process to get it released. So will need to wait for someone from CCP to do so. I pinged one of the devs, so hopefully it gets sorted next week.

1 Like

Thanks for this. Really appreciate it.

I now work around this with adding a connection. Especially that one between Hykkota <=> Ahbazon.

But is there some kind of overview you know of with all the connections? I only really know about this one.

How can we tell the new code is release btw, just checking if it works or is that visible somewhere?

Are you wanting a list of all the connections?

https://www.fuzzwork.co.uk/dump/latest/mapSolarSystemJumps.csv

I suspect the sites mentioned aren’t using the api. They’re probably using their own code. This kind of path finding isn’t particularly difficult to implement. A* or Dijkstra is normally the way to go. like using networkx in python.

1 Like

Thanks. This is helpful. I’ll read on the Dijkstra thing. I’ve seen it before but thought using the API would make more sense since that is directly communicating to the source.

The route api, iirc, doesn’t actually talk to eve. It’s standalone.

Never used networkx and Im not really a pro in Python, so I guess something new for me to work with.

About that fuzzworks dump, that is everything. All connections. Is there a dump with the connections that were new? Meaning, not in the esi? For now, to avoid a complete overhaul of the code, I better stick to the esi first. And in such case, I only need those “new” added connections.

Not something I maintain, I’m afraid.

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.