Release EDSM - Elite Dangerous Star Map

iain666

iain666
I've just submitted 40 distances to Yaimeia KN-S e4-2 and EDSM is not trilaterating.

ED Discovery has been happy with
Trilateration successful (25.2430ms), exact coordinates found.
x=26237.375, y=-168.59375, z=17798.59375 Sol: 31705.2 Sag A* 27435.9
Click to expand...
for the last 30 systems, not sure what's throwing you out...

EDD is suggesting the distance is 0.01ly out for Fe 1 Sector OS-R b5-4, G 123-49 and Ts'in Yanka if that helps.


 

GregMalcolm

GregMalcolm
iain666 said:
I've just submitted 40 distances to Yaimeia KN-S e4-2 and EDSM is not trilaterating.

ED Discovery has been happy with

for the last 30 systems, not sure what's throwing you out...

EDD is suggesting the distance is 0.01ly out for Fe 1 Sector OS-R b5-4, G 123-49 and Ts'in Yanka if that helps.
Click to expand...

I've created a number of points recently and seem to find it's like a dice roll. Sometimes it works out under 10, sometimes I have to give up. I seem to have slightly better luck if I sink down to at least -300ly where that's possible.
 

Inhumierer

Inhumierer
iain666 said:
I've just submitted 40 distances to Yaimeia KN-S e4-2 and EDSM is not trilaterating.

ED Discovery has been happy with

for the last 30 systems, not sure what's throwing you out...

EDD is suggesting the distance is 0.01ly out for Fe 1 Sector OS-R b5-4, G 123-49 and Ts'in Yanka if that helps.
Click to expand...


Web page says: Found alternative positions without errors, we need more reference distances!
So there is more than one position in the 1/32 Ly grid where all distances are ok. EDD should not find exact coordinates, or we use a different distance calculation. This may happen if the calculated position is close to a 1/32 Ly grid border (I believe), it's hard to predict and hard to get rid of. Probably RedWizzard's page will ask for better reference distances, it should be worth a try at least. A GIT clone can be found at http://inhu.edsm.net/rw/ed-systems/entry.html
 
Last edited:

AnthorNet

AnthorNet
GregMalcolm said:
I've created a number of points recently and seem to find it's like a dice roll. Sometimes it works out under 10, sometimes I have to give up. I seem to have slightly better luck if I sink down to at least -300ly where that's possible.
Click to expand...


You could try to fill the last field reference on EDSM.
It is a test with a system under or above 750LY.

Will see if it helps.
 

AnthorNet

AnthorNet
iain666 said:
I've just submitted 40 distances to Yaimeia KN-S e4-2 and EDSM is not trilaterating.

ED Discovery has been happy with
Code: 
x=26237.375, y=-168.59375, z=17798.59375

for the last 30 systems, not sure what's throwing you out...

EDD is suggesting the distance is 0.01ly out for Fe 1 Sector OS-R b5-4, G 123-49 and Ts'in Yanka if that helps.
Click to expand...

From what I understand about EDDiscovery code, it is happy if there is at least 5 good distances, and 2 more than previous results.
I've found by testing, that it could work but in some cases, we could find other results that are good for each distances.

On EDSM for your particular example, we have found the exact same coords, but also found that the variation
Code: 
Checking variation: 26237.375 / -168.5625 / 17798.59375

also works for all distances, so how can we know for sure which coordinates are the real one?
Might be other coordinates that works, but we stop after we found a good variation to avoid too much time on calculation.
 
Last edited:

iain666

iain666
That's fair enough then - although if it's not gotten there by 40 I have hit my 'cannot be ringed anymore' threshold :)
 

AnthorNet

AnthorNet
AnthorNet said:
You could try to fill the last field reference on EDSM.
It is a test with a system under or above 750LY.

Will see if it helps.
Click to expand...

Ok, now you can try the outlying references.
We will test if those references helps getting trilateration when far outside the bubble.

Reference 1/2 are more than 750 on X axis
Reference 3/4 are more than 750 on Y axis
Reference 5/6 are more than 1500 on Y axis
Reference 7/8 are more than 750 on Z axis

We might adjust the values to find something easier for explorer trilaterators ;)
 

Inhumierer

Inhumierer
Sometimes the server has some memory issues, I'm still investigating. And I wonder why those issues occur only from 5 to 6 AM EST... Anyone got an idea? ;)
Sometimes the database server gets shut down because the main server is out of memory, and I can't figure out why this happens. It's probably some kind of web server issue, but I don't know where's the reason for this. I'll setup some RAM monitoring, anyway. And I'l let you know if I make any progress.
 

GregMalcolm

GregMalcolm
Inhumierer said:
Sometimes the server has some memory issues, I'm still investigating. And I wonder why those issues occur only from 5 to 6 AM EST... Anyone got an idea? ;)
Sometimes the database server gets shut down because the main server is out of memory, and I can't figure out why this happens. It's probably some kind of web server issue, but I don't know where's the reason for this. I'll setup some RAM monitoring, anyway. And I'l let you know if I make any progress.
Click to expand...

Ah, I assumed it might have occurred while the server was running batch jobs. Mysterious... :)
 

AnthorNet

AnthorNet
iain666 said:
The EDSM site has three 'Top 100' lists - what is the 'First discovered' list actually measuring?
Click to expand...

It is measuring the total of systems the user have first discover. With the data we got, it corresponds to the first log entry for a system.
It is also marked on each system header where we have a flight log entry.

We are planning more of incentive thing like that in the future, now that we have more and more data coming from various software.
 

AnthorNet

AnthorNet
Hi guys just a heads up.

We have been under heavy load yesterday, mainly coming from a russian bot crawling our pages very fast and a faulty index in our database.
All of this is now corrected.

While working on resolving that situation we have put a lot more memcache in our backend to ensure distances listing are faster than before.

We also move the heatmap charts and the statistic chart on asynchronous loading, to make the homepage a lot more responsive.
We also put the trilateration on systems asynchronous to load the page faster and don't block you in the frontend.

Here is an example on how heatmap can look:

user-statistics.png
 
Last edited:

GregMalcolm

GregMalcolm
FYI: "JW 392" co-ords seem to be way off. It's come up a couple of times tonight and it's "expected" coords are off by quite a long way off. I've noticed other people have registered points in recent times that are in the yellow or reds. Maybe the star got moved by FD at some point?

For example from Fegoe DV-S c6-4 it should be 19813.05Ly but the expected value is 19861.71
 
Last edited:

AnthorNet

AnthorNet
GregMalcolm said:
FYI: "JW 392" co-ords seem to be way off. It's come up a couple of times tonight and it's "expected" coords are off by quite a long way off. I've noticed other people have registered points in recent times that are in the yellow or reds. Maybe the star got moved by FD at some point?

For example from Fegoe DV-S c6-4 it should be 19813.05Ly but the expected value is 19861.71
Click to expand...

Coordinates are fixed. Thanks for the report.
 
You must log in or register to reply here.
Back
Top Bottom