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

RedWizzard · Oct 28, 2014

Snuble said:
Do you have a defined upload (url + what you expect in the json) where I can dump my data? The alternative is for me to take one of the defined single system web forms and just loop POSTs to it.

I don't have a url yet - hopefully later today. The format can either by the full system format from systems.json or just simple distance data ala distances.json.

- - - - - Additional Content Posted / Auto Merge - - - - -

Snuble said:
Just want a dump of gathered coords, to make it easier for me to find empty places in the larger pill. Of course if tgc is up in a couple of days that is not as relevant anymore.

I'll try to keep systems.json updated until we have TGC.

Biteketkergetek · Oct 28, 2014

Sorry to ask it here, but I am thinking of a problem for some time to help visualize all the stars in B2/B3. Is there a simple way to calculate the general axis direction of the pill, so I have a 3D line that has the smallest distance from every system? Perhaps I am not phrasing my queries right, but could not find anything so far.

Chromatix · Oct 29, 2014

Mathematically, you could use a correlation function.

Visually, you could rely on my maps. Those don't have B3 stuff on them yet, though.

RedWizzard · Oct 29, 2014

Biteketkergetek said:
Sorry to ask it here, but I am thinking of a problem for some time to help visualize all the stars in B2/B3. Is there a simple way to calculate the general axis direction of the pill, so I have a 3D line that has the smallest distance from every system? Perhaps I am not phrasing my queries right, but could not find anything so far.

I came up the following empirically in Excel for B2. I haven't tried to calculate a "best fit" axis (and a best fit would not necessarily be correct - we need the surrounding out-of-pill stars to get an accurate axis).
z = x * 0.45 + 7
y = x * 0.025 + 35
A 33 Ly radius cylinder using the above axis contains all pill stars (more or less). I was kind of expecting the radius to be 30 Ly but I don't think it is that small.

JesusFreke · Oct 29, 2014

Biteketkergetek said:
Perhaps you could make it python 3 ready with just

Code:

from __future__ import print_function, true_division

at the top. At least that is what I usually do. Then replace / with // for integer division and print xyz with print(xyz) as needed.

Thanks for the tip! I tried it out but am getting different results when using python3. I'll have to look into why later. For now, I'm going to dig into beta 3

Harbinger · Oct 29, 2014

Only just got Beta 3 up and running in the last hour or so so I'm not going to do any mapping tonight but one thing I noticed when trying to head out there using TradeDangerous route planning is we seem to have lost some of the beta 2 stars. These 2 I've noticed are not within the playable area anymore:

Code:

'G 205-38',-51.65625,19.125,11.40625,'Beta2-Inferred','2014-10-21 17:16:31'
'Wregoe WV-E d11-104',-45.4375,31.3125,-25.125,'Beta2-Inferred','2014-10-21 17:16:31'

Wregoe WV-E d11-104 actually points to YIN SECTOR YE-A D104 (well outside the playable area) now.

JesusFreke · Oct 29, 2014

Harbinger said:
Only just got Beta 3 up and running in the last hour or so so I'm not going to do any mapping tonight but one thing I noticed when trying to head out there using TradeDangerous route planning is we seem to have lost some of the beta 2 stars. These 2 I've noticed are not within the playable area anymore:

Code:

'G 205-38',-51.65625,19.125,11.40625,'Beta2-Inferred','2014-10-21 17:16:31' 'Wregoe WV-E d11-104',-45.4375,31.3125,-25.125,'Beta2-Inferred','2014-10-21 17:16:31'

Wregoe WV-E d11-104 actually points to YIN SECTOR YE-A D104 (well outside the playable area) now.

From the patch notes, it sounds like they changed the generation logic a bit, for these generated stars.

Rejig of galaxy mass, so that the arms are more populated, and more systems are generated in the thick disc
The gaps between arms now have some sparse stars

RedWizzard · Oct 29, 2014

Harbinger said:
Only just got Beta 3 up and running in the last hour or so so I'm not going to do any mapping tonight but one thing I noticed when trying to head out there using TradeDangerous route planning is we seem to have lost some of the beta 2 stars. These 2 I've noticed are not within the playable area anymore:

Code:

'G 205-38',-51.65625,19.125,11.40625,'Beta2-Inferred','2014-10-21 17:16:31' 'Wregoe WV-E d11-104',-45.4375,31.3125,-25.125,'Beta2-Inferred','2014-10-21 17:16:31'

Wregoe WV-E d11-104 actually points to YIN SECTOR YE-A D104 (well outside the playable area) now.

Oh no! They've changed the procedural generation and the naming. I wasn't expecting that. We're going to have to recheck all the beta 2 stars. At least the Wredguia and Wregoe systems are completely gone though the search takes you to other stars (stale index I bet).

Can we even rely on the beta 2 reference list?

Harbinger · Oct 29, 2014

Looks like a lot has changed in Beta 3 with regards to the systems we once knew.

3 of our Beta 2 reference stars no longer exist (Clotti, San Guaralaru & Haras), 2 of them are massively out of position (Fu Haiting & Arabha) and there is a slight variance (under 0.005 LY) for each of the distances reported for the remaining 4 reference stars when you recalculate the coordinates for a star.

Looks like we may have to remap everything we previously knew.

ShadowGar · Oct 29, 2014

Not only have they changed locations of stars but also changed a majority of the stations in each star. Some stars lost stations, some gained stations and many station names have changed. A lot of work down the drain. I wish FD would just give us a list dump of all stars locations and station names per star.

We know they have a method of storing this data. Can we just push for the dump of the data? We will need all of us behind it. Because if we don't, whatever work we do now, will possibly be lost again on next update.

Chromatix · Oct 29, 2014

On the plus side, at least the galaxy map now has a working route planner in it.

RedWizzard · Oct 29, 2014

ShadowGar said:
Not only have they changed locations of stars but also changed a majority of the stations in each star. Some stars lost stations, some gained stations and many station names have changed. A lot of work down the drain. I wish FD would just give us a list dump of all stars locations and station names per star.

We know they have a method of storing this data. Can we just push for the dump of the data? We will need all of us behind it. Because if we don't, whatever work we do now, will possibly be lost again on next update.

They obviously don't have a list of all 400 billion systems. Rather they generate them on the fly as needed.

Knowing how it worked in FE2 and FFE I can understand how it could require some effort. They would probably have to generate all systems in sectors that overlap the beta 3 sausage and then check each one to see if it actually is in the sausage. They may not even have a way to running the generation outside the game (though they probably do - they gave pretty precise counts in the newsletter). Anyway I can understand why they don't want to spend any time on it.

By the way, anyone else having problems with the new forum? I've had at least 3 or 4 posts eaten after getting "token out of date" or "you've logged in since" messages. Very annoying.

Biteketkergetek · Oct 29, 2014

JesusFreke said:
Thanks for the tip! I tried it out but am getting different results when using python3. I'll have to look into why later. For now, I'm going to dig into beta 3

The difference may come from the fact python 3 (and true_division) changes the logic of the division operation. In python2 3/2 is 1 but in python3 3/2 is 1.5 (ie. division always produces a real value). If I'm not mistaken you need integer division in some places, there you should use the // (double slash) integer division operator instead of / (single slash) which is for "true division".

TornSoul · Oct 29, 2014

OMG.. can't believe anyone else haven't mentioned it yet....

The Galaxy Map now only shows distances with TWO decimals.... not three....

That really sucks....

I'm gonna put the TGC on hold a bit - Until we figure out how to make that work
(I'm gonna be running alot of tests...)

wolverine2710 · Oct 29, 2014

TornSoul said:
OMG.. can't believe anyone else haven't mentioned it yet....

The Galaxy Map now only shows distances with TWO decimals.... not three....

That really sucks....

I'm gonna put the TGC on hold a bit - Until we figure out how to make that work
(I'm gonna be running alot of tests...)

Depening on the TZ some of us just woke up ;-) But you are right, that is really totally suboptimal. Can any of the tool authors check what the consequences of this are with the already crowd sourced data when go from 3 to 2 decimals. Have this feeling it won't be pretty....

TornSoul · Oct 29, 2014

That's what I'll be testing for a bit...

TornSoul · Oct 29, 2014

A quick run through shows that my "draconian" test for coordinate validity is no longer useful.
It simply rejects too many (about 50% - eyeballing it)

EDIT: Yup - Rejecting 66% now - as opposed to 3% before.
Meaning that in only 33% of cases will trilateration report the same coordinate for the various combinations of p1-p5 reference systems.

RedWizzard · Oct 29, 2014

wolverine2710 said:
Depening on the TZ some of us just woke up ;-) But you are right, that is really totally suboptimal. Can any of the tool authors check what the consequences of this are with the already crowd sourced data when go from 3 to 2 decimals. Have this feeling it won't be pretty....

Yeah, that's a big blow. The trilateration will be harder with 2dp but with the right references it should still be doable. I expect there will be a lot more cases where multiple grid locations can satisfy a particular set of distances. I'll probably have to switch to something like JesusFreke's technique of identifying and searching the whole candidate region.

All our existing data is potentially wrong now, including the reference systems, as systems have been moved (and removed entirely). It will all need to be checked.

I don't think I'm going to do anything further until we get a beta 3 reference list from Michael. And even then... the route planning problem is pretty well solved in game. That was my main interest.

TornSoul · Oct 29, 2014

RedWizzard said:
the route planning problem is pretty well solved in game. That was my main interest.

Only if you know where you want to go.

Trade route planners still need the data - As they need to incorporate it in the calculations of where you *want* to go.

wolverine2710 · Oct 29, 2014

I just send Michael Brookes a PM about the partial list of coordinates.
Also mentioned the 2dp vs former 3dp precision in the GM - and the consequences of it for the project.

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

RedWizzard

Biteketkergetek

Chromatix

RedWizzard

JesusFreke

Harbinger

JesusFreke

RedWizzard

Harbinger

ShadowGar

Chromatix

RedWizzard

Biteketkergetek

TornSoul

wolverine2710

Tutorial & Guide Writer

TornSoul

TornSoul

RedWizzard

TornSoul

wolverine2710

Tutorial & Guide Writer