Extremely long times to calcualte routes

[sfaxo]

I've been waiting about 30min now for ED to calcualte a route to a star. I'm in the core region (just by Sagittarius A*) and its just ridiculous. I think we need an option for a quick route, nothing that looks for the most efficient way to get to a place, just a valid way. The algorithm could be go towards destination, any star within jump range select it, repeat until goal is reached. I'm just trying to explore random stars near the galactic core. It takes way too log to calculate a route.

 

[Katejina]

From what I've been reading in multiple threads, others are having this same issue around Sag A*. I'd chalk it up to server/communication issues. There is a larger percentage of clustered stars near Sag A*, right? Could be the algorithms are, for whatever reason, having difficulties in establishing effective routes with so many choices available.

 

[pippo]

Not sure if it actually helps but someone said in another post that switching the map in Realistic helped.
You'll have to test for yourself though.

 

[Surly_Badger]

I've been waiting about 30min now for ED to calcualte a route to a star. I'm in the core region (just by Sagittarius A*) and its just ridiculous.

It has a lotta lotta stars to consider as possible routes. When I was close in what I did was 200ly jumps. It was still slow but it has far far fewer stars to look at that way.

Other explorers go up or down, to where there are fewer stars. You may want to consider exiting up or down because there are big neutron star fields there anyway.

 

[Dommarraa]

Just to make you even more paranoid. I was 40ly from Sag A (had visited), as set a route plotting. After a few mins, the girating icon stopped girating, so I left it for a min or 2, then found that it wouldnt respond to any keypress to exit. I was pondering whether to pull the plug and have to wait all over again when I found myself staring in disbelief at the insurance screen.

No, I wasn't pointing at a star. Ticket raised, no response.

 

[Kuroshio]

My route planner started to behave erratically about 5 kLYs out of Sag Asterisk. For a while, 500 LY routes remained plottable, but it went down to 300, then 250, and then 150. (Realistic map view, neither eco nor fast selected.)

But it was a bit of hit and miss. I got into the habit of cancelling the calculation after 5 minutes and selecting another target close by until something came up in less than 5, and lo and behold, it often worked. It felt - but that's very subjective - that more or less exact multiples of my jump range worked best.

I think the problem is that there are a lot of nearly optimal routes - fast routes surely resulted in a lot of jumps at exactly my maximum range. Considering that the bubble of possible jumps is less of a child's drawing of a star and more like a massive white sphere of light close to the core, it's not that surprising. Even liberal arts me knows that the travelling salesman does not have an was job.

 

[ElectricZ]

We're running a race to Sag A* and of course time is of the essence for pilots trying to get the fastest time. The bad news is there's a certain point where you can't do anything at all to help. It's going to be slow. It has nothing to do with network latency and everything to do with computing power. The route planner works by plotting every course possible using your ship's maxium jump length or most efficient fuel usage in a sphere around your ship. In most cases in colonized space, this takes seconds because systems are spread out. However, in the galactic core, there are 1000x or more systems to map. Check out your nav panel. It lists the 20 closest systems. In colonized space, you might see the 20th system on the list showing 20-30 light years. Near the core, the 20th system might be 3 light years away. Your computer tries to plot the route of every star within whatever radius you tell it, and it grows almost exponentially the closer you get to the core.

The Buckyballers have come up with some remedies to help as you get close, but again they only work to a certain point. After that, your best bet is to plot one jump at a time.

1. Use realistic mode on the galmap.
2. Exit and restart the game. This seems to clear cached routes, but only works up to about 1,000 LY out. After that, you grind.
3. Allow the route plot to reach 99%, then exit the route planner (which may also take a while to do because the CPU is busy.) The computer will remember the route up to 99%, and if you go back to the galmap and select a slightly closer end point to the route, it will complete quickly
4. Plot your jumps in a multiple of your jump range. This was discovered by the current race leader, CMDR Alot. If your jump range is 30 LY, look for stars at 30, 60, 90, 120, 300, 600 light years out, etc and the computer can plot them faster. If your range is 25 LY, look for 25, 50, 75, 100, 150, 200, etc. It requires some math, but Alot reported the effect was reproducible.

Remember, if you plot a route of 1000 LY, your computer has to look at every star within that range to see which ones make the closest route. In the galactic core there are so many stars in that sphere that your computer can't keep up. Anything you can do to reduce the required processing power the better off you will be. But you might also just resign yourself to the fact that getting in and out of the core will take a lot of patience and discipline.

Good luck CMDRs!

 

[Alex Brentnall]

We're running a race to Sag A* and of course time is of the essence for pilots trying to get the fastest time. The bad news is there's a certain point where you can't do anything at all to help. It's going to be slow. It has nothing to do with network latency and everything to do with computing power. The route planner works by plotting every course possible using your ship's maxium jump length or most efficient fuel usage in a sphere around your ship. In most cases in colonized space, this takes seconds because systems are spread out. However, in the galactic core, there are 1000x or more systems to map. Check out your nav panel. It lists the 20 closest systems. In colonized space, you might see the 20th system on the list showing 20-30 light years. Near the core, the 20th system might be 3 light years away. Your computer tries to plot the route of every star within whatever radius you tell it, and it grows almost exponentially the closer you get to the core.

The Buckyballers have come up with some remedies to help as you get close, but again they only work to a certain point. After that, your best bet is to plot one jump at a time.

1. Use realistic mode on the galmap.
2. Exit and restart the game. This seems to clear cached routes, but only works up to about 1,000 LY out. After that, you grind.
3. Allow the route plot to reach 99%, then exit the route planner (which may also take a while to do because the CPU is busy.) The computer will remember the route up to 99%, and if you go back to the galmap and select a slightly closer end point to the route, it will complete quickly
4. Plot your jumps in a multiple of your jump range. This was discovered by the current race leader, CMDR Alot. If your jump range is 30 LY, look for stars at 30, 60, 90, 120, 300, 600 light years out, etc and the computer can plot them faster. If your range is 25 LY, look for 25, 50, 75, 100, 150, 200, etc. It requires some math, but Alot reported the effect was reproducible.

Remember, if you plot a route of 1000 LY, your computer has to look at every star within that range to see which ones make the closest route. In the galactic core there are so many stars in that sphere that your computer can't keep up. Anything you can do to reduce the required processing power the better off you will be. But you might also just resign yourself to the fact that getting in and out of the core will take a lot of patience and discipline.

Good luck CMDRs!

Spot on answer, the faster your pc the better the results. I (am fortunate enough to) have a pretty high end pc and in the core clusters a 1000LY jump takes about 30-60 seconds to compute depending on how difficult the jump is.

 

[Redec]

Plan routes a little lesser than twice your jump range, then it will go fast. For example if your jump range is 33 ly, then plan routes which are 60-63 ly long.

/ CMDR Redec

 

[Ulysses_W]

Chalk it up to the shear density of stars. I'm surprised it can actually calculate a route at all around Sag A*.

 

[drkaii]

Not sure if it actually helps but someone said in another post that switching the map in Realistic helped.
You'll have to test for yourself though.

Yes this does help, as well as sticking to as few grids as possible

 

[acid_byte]

(...)2. Exit and restart the game. This seems to clear cached routes,(...)

The game stores cached routes?

 

[Titler]

Spot on answer, the faster your pc the better the results. I (am fortunate enough to) have a pretty high end pc and in the core clusters a 1000LY jump takes about 30-60 seconds to compute depending on how difficult the jump is.

I don't believe these answers are entirely correct; in particular the star map visuals, unless you're rotating the galactic map constantly, won't be computing anything so this won't affect base line performance. And although my system is way under minimum listed specs (Win7 64bit Home Premium, AMD Athlon X2 5600+ (2.9ghz), 4gb RAM, ATI 4670HD 512mb) it actually calculated the final 600ly or so, from Myrielk AN-S C17-3322 to Sagitarrius A* instantly. Then on the way back out it choked constantly on even 2 leap jumps until I reached around 5,000 LYs out again.

Likewise I thought I'd found a figure which seemed to work well with my jump range, around 987ly, but that changed completely later on and just didn't work, even over the same rough ground that it had once plotted instantly.

Rather it's likely to be a combination of cpu power (and whether you're doing anything else at the same time), the specific route you're on (moving across an axis increases the potential paths) and just a poorly optimised route planner which only occasionally quickly finds an obvious path. The variables are just way, way too many, not just including the star systems themselves, to be able to pin down one, or even a group of definite solutions. It's possible a monster PC is able to brute force it's way through faster, but on my wheezing system it can still plot instantly even in the core, but everything, not just the stars have to be aligned correctly. And I don't think we know what everything is, yet.

 

[Nilt]

My guess is it's a lot like brute forcing a password. Sometimes you get lucky and get the correct one right away and sometimes not so much. Same thing with the route. If it happens to find a workable solution quickly, great! Sometimes that is less likely, but as Titler said, we really don't know enough about it yet to say why that is.

 

[Whiterose]

Badger is right, the algorithm cant cope with all the stars in the core. I am there at the moment and manually jump, star to star.