Sectors Can't Count

[ZomgK3tchup]

Either that, or I can't count. Not sure which.

I don't know if this is a bug specific to full hyperlanes, but there are 5 systems shown below that are 4 jumps from my capital that are considered frontier worlds:

Spoiler

Am I missing something here?

 

[HFY]

It's not limited to full-hyperlanes, it's present in the default hyperlane density too.

If you have a map where this happened without mods, please do make a bug report with the save.

 

[ZomgK3tchup]

Made a bug report with an entirely vanilla save, so upvote to get this bug fixed:

Stellaris - Sectors Miscount Hyperlane Jumps [3.4.5] [a361]

Description Sectors Miscount Hyperlane Jumps [3.4.5] [a361] Game Version 3.4.5 What version do you use? Steam What expansions do you have installed? Do you have mods enabled? No Please explain your issue is in as much detail as...

forum.paradoxplaza.com

In this game I ran on observer, the AI created 7 examples of this:

Spoiler

The AI also sucks at specializing its planets, but that's another conversation, I guess.

 

[SirBlackAxe]

I know this bug occurs when there are hyperlane routes of different lengths from the sector capital to the excluded system, so higher hyperlane settings do make it more likely.

Without seeing the algorithm itself I can't say exactly what the problem is, though.

----

I think something like this would work without needing to calculate each possible route for each possible system:

draw_sector = {
    sector_capital_system = {
        if = {
            limit = { is_in_sector = yes }
            clear_current_sector = yes
        }
        add_to_sector = yes
        
        every_neighbor_system = {
            limit = {
                owner = sector_capital_system.owner
                is_in_sector = no
            }
            set_star_flag = distance_1@sector_capital_system
            add_to_sector = sector_capital_system.sector
        }
        
        this.sector = {
            every_system_within_border = {
                limit = { has_star_flag = distance_1@sector_capital_system }
                every_neighbor_system = {
                    limit = {
                        owner = sector_capital_system.owner
                        is_in_sector = no
                    }
                    set_star_flag = distance_2@sector_capital_system
                    add_to_sector = sector_capital_system.sector
                }
            }
            
            every_system_within_border = {
                limit = { has_star_flag = distance_2@sector_capital_system }
                every_neighbor_system = {
                    limit = {
                        owner = sector_capital_system.owner
                        is_in_sector = no
                    }
                    set_star_flag = distance_3@sector_capital_system
                    add_to_sector = sector_capital_system.sector
                }
            }
            
            every_system_within_border = {
                limit = { has_star_flag = distance_3@sector_capital_system }
                every_neighbor_system = {
                    limit = {
                        owner = sector_capital_system.owner
                        is_in_sector = no
                    }
                    add_to_sector = sector_capital_system.sector
                }
            }
        }
    }
}

 

[Starfury]

Is this the same problem messing up sectors in the L-Cluster?

 

[ZomgK3tchup]

Is this the same problem messing up sectors in the L-Cluster?

Sorta.

Both problems are about how sectors count jumps from the capital, but the L-Cluster issue seems unique to systems generated after game start.

 

[taltamir]

Either that, or I can't count. Not sure which.

I don't know if this is a bug specific to full hyperlanes, but there are 5 systems shown below that are 4 jumps from my capital that are considered frontier worlds:

Spoiler

Am I missing something here?

I think it has a problem with considering multiple routes.
Take [Churassix] for example. it is 4 jumps away if we go:
sol -> chiminol -> bessia -> oshimir -> Churassix
but it is 5 jumps away if we got:
sol -> lolala -> eptrabon -> bessia -> oshimir -> Churassix

The 5 jumps path is "shorter" in terms of absolutely distance travelled, but is slower due to more jumps.

I also noticed the tooltip for sector says 3 jumps away when that is flat out wrong. it is 4 jumps away. except for when it breaks due to multiple bugs.

 

[taltamir]

I think the logic to handle this should be something like the following pseudocode:

new array X; //creates a new array with most existing data copied but distances all being set to -1.
bool change_made = false;
while(!change_made)
{
change_made = false;
for each s in systems{
{
  if capital() {s.distance = 0; change_made = true;}
  else for each h in hyperlanes(s)
  {
    if(target_system(h).distance == -1) continue; //skip for now until that system gets populated
    if((s.distance == -1) || (target_system(h).distance + 1 < s.distance))
    {
      s.best_path = target_system(h);
      s.distance = target_system(h).distance + 1;
      change_made = true;
    }
  }
}

First loop it will go through all systems and mark capitals with 0. Subsequent loops can hit more than once system at once.
But if it encounters a scenario where a system ends up being closer, it will lower its value and switch the path to the more optimal path.

for example. let us say it has those paths
sol -> lolala -> eptrabon -> bessia -> oshimir -> Churassix
sol -> chiminol -> bessia -> oshimir -> Churassix

loop 1: sol marked as distance 0
loop 2: lolala and chiminol marked as distance 1. through coincidence eptrabon was checked after chiminol and marked as distance 2. then bessia was checked and marked as distance 3 through eptrabon.
loop 3: bessia path through chiminol is noted to be distance 2 and bessia is corrected to the new quicker path. etc

basically it will go through it and each time lower possible loops as much as possible until it figures the optimal path