Discussion What is the most efficient way to crowdsource the 3D system coordinates

The Pill is gone. Seems the whole galaxy is open.

We've got multiple commodity markets per system. In fact it looks like pretty much all stations and outposts have markets.

Holy crap, that is a big galaxy. I know we've been able to see the whole thing in beta... but there's just something about actually being able to go out into that vastness now.... :D:D:D
 
They are named permits but it probably doesn't matter unless a particular permit allows access to multiple systems or a particular system can require one of several permits. Don't yet know if that will be the case but I suspect it won't. I'd probably just go with a boolean for now and if the permit system turns out to be more complex we can always add another field with more detailed info.

The permit list on the status panel has "Region Permits". I think that confirms that they plan to have single permits that are required by multiple systems. I'd still stick with a boolean until we have more information.
 
It's EDSC operational? I get 'Success' message but it returns an empty array, so I guess the db has been wiped out?

I find systems available in the reference cells but every time I try to enter data for a system I get "whoops - returned an error, it's not supposed to do that." Not sure how to go forward with entering the data. Although I was placed - a ways away from the former 'core'
 
It's EDSC operational? I get 'Success' message but it returns an empty array, so I guess the db has been wiped out?
Remember to supply a date filter - or you'll only get changes made the last 24h.
Have you tried the fiddle linked from the API doc page? (works fine for me as of this writing)

I find systems available in the reference cells but every time I try to enter data for a system I get "whoops - returned an error, it's not supposed to do that." Not sure how to go forward with entering the data. Although I was placed - a ways away from the former 'core'

Works fine for me as of this writing - Could you detail exactly what data you are trying to enter?


============

All that said - Expect the DB to be wiped within a few days - Once we hear back from Michael Brooks with some seed data.
Ie. it's not really worth entering any data atm.
 
That's what "we nominate" means.
Totally didn't see/read the "we nominate" bit for some reason... :eek:

If we still had 3 dp that would be fine, but with 2 dp the B3 list is not that suitable for the wider galaxy. It struggles with systems beyond the ends of the cylinder in particular.
Hmm seems we have a difference in approach.

If I understand you right, you want 20-30 systems pretty much covering the whole galaxy, so they can be used for trilateration for any system?

I don't think that's the best approach, as that would mean those 20-30 systems would be *far* apart (distance wise).
Trilateration is a bit sensitive to that... (range) - As you then get *huge* (albeit slim) overlapping areas, increasing the amount of possible candidates.

In my experience it's better to have a bunch of close(ish) reference stars, as the overlapping areas are smaller then.

(the above is a simplification of course - trying to keep it short here :p)


Also - and for me at least, more importantly is "the spirit" of the whole thing.
In my mind we have a small (close by) nucleus of known systems (like the SB3 set)
We then "work outwards" from there - as people spread out.
I personally find that more in the exploration spirit, than having some known systems out where no one has ever been yet.

It is obviously possibly to actually calculate the coords to something far away (1500-2000 LY I guess) given enough spread out refs (as we did with Alpha Cygni in beta).
But that is actually still rather close, considering the scale of the galaxy.



It probably is pretty manual, unless they've already built a tool to do it.
They have to have some sort of tool - I don't see Michael (or one of the devs) doing that SB3 list one star at a time.

I don't know what sort of (easy) filtering they can do - But perhaps a list of all systems from the star catalogs (as those aren't auto generated) within a 200-500 LY radius of Sol (if the SB3 list isn't feasible for some reason)
 
Last edited:
If I understand you right, you want 20-30 systems pretty much covering the whole galaxy, so they can be used for trilateration for any system?

I don't think that's the best approach, as that would mean those 20-30 systems would be *far* apart (distance wise).
Trilateration is a bit sensitive to that... (range) - As you then get *huge* (albeit slim) overlapping areas, increasing the amount of possible candidates.

I was thinking about the inhabited region rather than the whole galaxy. Not sure how big the inhabited region is but I guess not more than a few hundred Ly radius from Sol.

In my experience it's better to have a bunch of close(ish) reference stars, as the overlapping areas are smaller then.

Trilateration works best if the reference points are well distributed around the target point. If they're all in the same direction (which is what happens as you move away from the beta 3 sausage) then it struggles (because the spherical shells overlap more). Distance wouldn't matter at all except the single precision floating point ED is using starts to lose precision after about 500 Ly which increases the thickness of the spherical shells.

Also - and for me at least, more importantly is "the spirit" of the whole thing.
In my mind we have a small (close by) nucleus of known systems (like the SB3 set)
We then "work outwards" from there - as people spread out.
I personally find that more in the exploration spirit, than having some known systems out where no one has ever been yet.

I agree in terms of the wider galaxy, but I don't see much realism/spirit in not knowing the locations of a couple of dozen well spread populated stars. Don't worry, there'd still be 99.999% of the galaxy to work outwards into.

It is obviously possibly to actually calculate the coords to something far away (1500-2000 LY I guess) given enough spread out refs (as we did with Alpha Cygni in beta).
But that is actually still rather close, considering the scale of the galaxy.

The problem is that the beta 3 references start to struggle only a few dozen Ly off the ends of the sausage. Something as far away as Alpha Cygni is quite problematic. With the 2 dp distances we have now using all 12 distances I have is insufficient: there are still 3 possible candidates that I can't distinguish between and in fact I think none of them are correct because none of them match all 12 distances. I generally can't fix it with my current algorithm with 10 randomly selected beta 3 references (and a brute force search of 1728 grid locations around each trilateration candidate). I can get it located with the right 5 references from beta 3 but just picking references randomly would be an exercise in frustration for the cartographer. Alpha Cygni is only the worst of the distant systems (let's not even talk about Sag A*!), some of the others also sometimes fail with 10 random references.

We don't absolutely need other references, of course. We can use derived stars as references, and eventually we'll have to do just that. The danger with using derived references is that any error will propagate so I'd prefer to avoid it as long as possible.
 
I was thinking about the inhabited region rather than the whole galaxy. Not sure how big the inhabited region is but I guess not more than a few hundred Ly radius from Sol.

Ok - That I could get behind (inhabited region).
And yeah it looks like roughly a few hundred LY - Although not exactly centered around Sol (bit hard to tell though)


In that case id' suggest finding 6*4 systems (for 24 systems)

Define 4 spheres around Sol at 50/100/150/200 LY radius.
Pick 6 systems on each sphere in each of the cardinal directions (as close as that's possible).
Maybe a 5th sphere at 250 LY (for 30 systems).
This obviously could be refined - Like rotating each sphere so not all systems are on the cardinal directions, for a bit of variety (not sure how much is gained if each sphere has the stars on the cardinal directions only)

Just need someone to eyeball and name them - Doesn't have to be super accurate.

--

Again, depends on what's possible for Michael.

I'd much prefer a large dump ala SB3 over this still :)
 
Last edited:
I'm going to go out on a limb here... and say that I actually prefer that we *don't* get a dump. As long as we have a good "kernel" of systems to begin with, we can do it all ourselves. Or heck, even if we didn't have any reference systems at all, we could pick coordinates for 3 stars in order to "orient" the coordinate system, and then work from those.

I dunno, I just kinda like the idea of us doing it all ourselves, rather than getting an out-of-band dump from the devs.
 
All true - And I actually agree with the sentiment about "doing it all our self" (more satisfying in the end)

That unfortunately raises the "chicken&egg" issue with all the apps out there.
They need data for people to use them, and using them is hopefully what will get people to submit more data as well.

I haven't seen very much activity with people actually submitting new systems - So it would be a VERY slow start without some sort of seeding.

Else it'll be the same 3-5 people doing all the initial scanning again, just to get the thing started...
 
We do have one reference star: SOL (0,0,0)

It should be possible to calculate some other stars around Sol with the displayed coords grid and distance of the galaxy map.
 
All true - And I actually agree with the sentiment about "doing it all our self" (more satisfying in the end)

That unfortunately raises the "chicken&egg" issue with all the apps out there.
They need data for people to use them, and using them is hopefully what will get people to submit more data as well.

I haven't seen very much activity with people actually submitting new systems - So it would be a VERY slow start without some sort of seeding.

Else it'll be the same 3-5 people doing all the initial scanning again, just to get the thing started...

Right. To a degree there are conflicting goals between us having fun mapping the galaxy and people who want working trade tools as quickly as possible. I wouldn't mind not getting a dump but to maximise participation I think we want to get a good start as quickly as possible. A small group of well spread reference systems within a few hundred Ly of Sol is not a bad tradeoff I think.

I haven't updated my stuff to talk to TGC yet, though I will in the next few days. Then you should see some more activity (even if it's just generated by me) ;)
 
With the pill gone, its not that interesting to trilinerate where you are not. From a practical standpoint, gathering distances for x number of references is time consuming, and come on top of gathering distances to systems around you.

I find distance gathering a bit limited while still in a low range ship. Once going to 20+LY the number of distances gathere will make the incremental approach to system coordinates a bit more interesting as you start to get distances from multiple directions to every system.

I got started in LHS 1913 for Gamma. Not left the system yet, still have 3 docks to explore... From what I see with access to a Viper, Cobra and Lakon 6, it might be that the first time I leave LHS 1913 when I can't find those A modules...
 
Re multiple markets in one system. One thing that would need to be considered is the distance between each station in a system. The dream trading system would be two different economy stations at the same planet/binary planet system. These could be credits/hour on a scale higher profit 2 system routes could never compete with.

Obviously since everything orbits everything else, you would have positions change over time. If you where lucky the two stations could be aligned on the right side of the planets and make for a very efficient run. On other occations it could be on opposite sides, making it a pretty long drive. Same for stations orbiting more distant planets. Sometimes they could be on the same side of the star, other times on opposite side.
 
Works fine for me as of this writing - Could you detail exactly what data you are trying to enter?

Target: Yakabugai

Arcturus 77.82
Sol 78.59
Eta Cephei 73.23
Altais 101.23
Magec 66.83

Considering where I ended up placed it was somewhat difficult to find good candidates in the db that had good differentiation to allow the calculation of where I am.


============
All that said - Expect the DB to be wiped within a few days - Once we hear back from Michael Brooks with some seed data.
Ie. it's not really worth entering any data atm.

Mmmm - I didn't realize the seed data was primary. I thought it was additive to the already existing fixed/core systems. Ok. The way that they are scattering us it will have an interesting impact on the tools.
 
Target: Yakabugai

Arcturus 77.82
Sol 78.59
Eta Cephei 73.23
Altais 101.23
Magec 66.83

Considering where I ended up placed it was somewhat difficult to find good candidates in the db that had good differentiation to allow the calculation of where I am.
Fixed - Was an embarrassing bug in the case where no commander name was supplied :eek:
At least that's the bug I found - You don't mention if you supplied commander name or not.
If you did supply a commander name, then I don't know what the problem is... Need more data then.
Works fine for me now both with and without commander name.

EDIT: Come to think of it though - The fixed bug with commander name should not have resulted in the "whoops - returned an error, it's not supposed to do that." message....
So uhm...??

Mmmm - I didn't realize the seed data was primary. I thought it was additive to the already existing fixed/core systems. Ok. The way that they are scattering us it will have an interesting impact on the tools.
There is no "already existing fixed/core systems" (well one, Sol...) Hence the need(*) for some seed data to get started.

(*) Strictly speaking we don't even need that - But then we'd end up making our own coordinate system which might or might not fit what FD has in their DB.
For any practical purposes that actually wouldn't matter either way - But it just seems "untidy" :p
 
Last edited:
You should seed the system with a few systems that you can confirm from the previous releases of beta. Getting the distance right to SOL/Eranin/Aulin would be a good start. These are the kind of systems that the calculated coords should match distance to 2 decimals rounded no matter how well alternative coords match commited data.
 
Re multiple markets in one system. One thing that would need to be considered is the distance between each station in a system. The dream trading system would be two different economy stations at the same planet/binary planet system. These could be credits/hour on a scale higher profit 2 system routes could never compete with.

Obviously since everything orbits everything else, you would have positions change over time. If you where lucky the two stations could be aligned on the right side of the planets and make for a very efficient run. On other occations it could be on opposite sides, making it a pretty long drive. Same for stations orbiting more distant planets. Sometimes they could be on the same side of the star, other times on opposite side.

I'm not sure I agree. The problem with 2 stations in one system is that when you enter supercruise after leaving a station, you're near a planet, and it takes a while to accelerate up into the 1c+ speed range. As opposed to jumping to a nearby system - after you exit from the jump, you're already at ~.33c. I think it might actually be faster jumping between 2 systems with starports within.. say a couple hundred LS from the star.
 
Back
Top Bottom