system generation

[Q (OP)]

Hi,
I was wondering if it's possible to have a rather small 'closed' universe. 
I tried to make a game with max.   number of systems set to 4, but gave up after finding more than a dozen.    (numbered from -6 to 19.  )
Also, i tried to make a more serious game with max.   number of system set to 80 and if i go to the "system summary" menu of the spacemaster menu, i see two systems, mine is number 0 and i guess the number 286 is the NPR, does that mean i'll never see him (since max systems is 80) or is the max number of system ignored completely? 
I hope i make sense, i'd like to be able to set up a game with less than 1000 systems that would still feel like a normal game since it took me a looong time to find 40 in my first game.   

[DatAlien]

I never spend any attention to this point so i tested this with 5 systems max and 4 jp in Sol (as maybe a bigger number may force the game to spawn more systems), the only effect that I see is that all systems are heavily connected and I get a few times almost a closed network, but then there is an system that have an jp that leads to an new system

[voknaar]

I know you can modify the chance the system has a connection but I don't understand why it would generate more actual systems than the number you set it at. I'm trying to figure out basic gameplay still rather than reconfiguring the "norm" game scenarios. It might be a bug unless I'm not getting the system generations logic. 

[laz]

I do know steve has said there isn't a hard coded limit on the number of star systems. I think once you reach the max number it will still generate but i don't know if when you start a new game there is a pre-concieved number of star systems generated that it creates as the initial number of starting systems. Then generates the rest randomly?

Also I think steve said the last patch broke the map generation part. So you can't change the system generation modifier and other figures at the moment. Well you can change the numbers but they won't translate into the game as its broken till 5.30

[sloanjh]

The way it is supposed to work is the following:

Unexplored jump points aren't bound to a new system until you explore them.  When you explore them, Aurora randomly chooses a system ID (where the number of possible system IDs is the max number of systems we're talking about).  If there is no system with that ID, Aurora generates one and uses it as the target system; if there is a system with that ID it uses the system in question as the target system.  If the target system has any unexplored jump points, then it randomly picks one as the other side of the jump.  If the target system has no unexplored jump points, it creates a new "dormant" (not sure if this is the right name) jump point and uses it as the other side of the jump.  In Starfire, these new jump points were a special type called "hidden", which was technobabble to explain why they weren't detected when the system was surveyed.  Hidden jump points could only be detected by going through them or watching someone go through them, not by survey.  In Aurora, the dormant jump points are normal jump points and can by detected by (another) survey after they've been created (by a ship going through them).

So the way it's supposed to work is that the odds of generating a new system are roughly (Nmax - Nexplored)/Nmax.  Whether or not the mechanisms are bugged is another question - based on the descriptions in this thread it sounds like they are.

John

[ZimRathbone]

I suspect that the problem may lie in the local system chance which steve introduced to produce clustering effects.

 If your initial roll indicates that this is a local system then the program randomly assigns a number within the local range, ie say you have local system chance at 50% and  the range as 15, and you're in system 004 and explore a WP then there is a 50% chance that the destination system number will be in the range -011 to 19 EVEN IF you have max number of systems set to say 10 - basically there is no limit check in the local system code.  This only really becomes noticeable where you have a very small number of systems and a wide local range.

solution is to put the local system chance to 0

[Beersatron]

I am fairly certain that the local system chance does not work with a Real Stars game - that could be a factor.

[Q (OP)]

ah thanks, i could try with 0 spread and "not real" stars.  There would still be the problem of NPR spawning in systems that are to far to ever reach (like 200 - 300) however.  Should this be in bug reports?

[ZimRathbone]

ah thanks, i could try with 0 spread and "not real" stars.  There would still be the problem of NPR spawning in systems that are to far to ever reach (like 200 - 300) however.  Should this be in bug reports?

I think that if you set up the max no of systems before starting the game then the inital NPR spawns within that limit - its only if you start a game and then subsequently reduce max systems that the problem you describe will occur.
Even if you did that the odds are that the NPR would explore and find you (as their initial explorations would link to systems within your max range). 

Actually, thinking about it, thats not a bad thing - you could never open a WP straight into the NPR homeworld, but would only be able to reach "ring-1" systems - Steve you might want to consider this for homeworld systems

The issue of the local system generation code not checking that it generates a valid system no is  one you can reasonably raise as a potential bug - although given steves free time lately don't necesarrily expect a code fix soon - especially since there is an easy workaround

[Steve Walmsley]

I am fairly certain that the local system chance does not work with a Real Stars game - that could be a factor.

Yes, real stars has its own clustering mechanism. The local chance and spread only apply to random rather than real stars games

Steve