We’re still aiming for an ambitious patch next one in the next weeks: https://www.starfightergeneral.com/2022/11/ambitious-patch-coming/
Yet, we got a couple people on board that can work parallel with me and quite possibly get the galaxy done as well!



I’m not sure how big we want the galaxy… In my mind, I want to out do Elite Dangerous’ 400 billion star systems, and with proc gen, lets be honest, that’s just a number. You could make as many star systems as there are limitations in your random seed tech. If I’m lazy and don’t spend a few hours upgrading it, we could make 340282346638528859811704183484516925440 star systems… It’s not an exercise in anything relevant, just gamesmanship… It’s why Xbox did not sell XBOX2 to compete with Playstation 3, but made it Xbox360… (imagine grandparents buying for their kids,’Why by a 2 when I could buy a 3?’) Anyway, digressing…
I have a few choices when proc genning the galaxy, and I didn’t think through it since I wasn’t on that path, but when you have a team, you can travel many paths at once. Let’s logically look at the proc gen design of Starfighter General. There’s nothing in here, players can’t know, and no spoilers coming.



First Options of Design to look at:
If the galaxy is 10s of trillions of star systems, and I have 10,000 players, that means each player has about one in a billion chance to run across another player in a system…. Like you’d never see someone unless I made hub town systems, summoner/beacons to gather players, or a radar to see player activity…. Suboptimal
If the galaxy is too small, there won’t be anywhere to explore, you’d get bored of lack of stuff to look at.
Maybe I want to aim the galaxy size with the premise that you’ll see another player 1 in 100 randomly if the player base is 10,000. I just looked up World of Warcraft and they try and finangle it so each server only has 10,000. I have no cap on players in mine, so WOW has 1,000,000 concurrent players often, and I’ll have about 10 servers, my upper best case scenerio is 500,000 players per server. This would mean randomly players would see each other 1/2 per sector instead of 1/100, but this is not exactly right… Due to bell curve distribution of players not going to cruddy zones, a loner could probably find a system 1 in 20 no one there if desired. So yeah, all those numbers sound good if my premise is to want 1 in 100 to find a player randomly if player base is 10,000.
10,000 * 100= a nice cool boring 1million star systems for our first spit ball estimate



There’s one other reason I do no want quadrillions of star systems… While it costs no memory to create the system, I am allowing players to destroy/alter systems. Even store keepers and NPCS will be growing or shrinking their business or if they get too rich, building a fleet to cause trouble or guard stuff. So there’s a long term memory cost in each system though they require nothing to start. So lets do an estimate on memory costs, for this is the biggest thing with proc gen maps that change, in any genre, not just space.
My server… lets say has a capacity for 50 GB,not huge, maybe move servers if I can’t get it higher, but for starters,its there, def want 1 tb server disk,right? Yeah, lets go with 1 tb since this is a long run game and some hosting solution can drop 400$ on a 1 tb ssd forme, that’s silly if they can’t. Heck there’s a Sabrent 8 TB SSD going for only 2,000$. Alright, going to design around that guy, say we have 8 TB to work with… and I can ask my server team for custom solutions too, but we just spitballing here.
So we have 8 TB and a million systems. This means we can divide the million systems from the 8 trillion storage to get our storage per system, so you drop a trillion to a billion to a millon. So we have 8 meg to store info for each system… And to be honest, that sounds fair. 8 meg a system is more than enough to store CIvilization/Master of Orion style localized data for armies/fleets/economy per system. Great, so 1 million systems looks good on paper! Dr. Evil would approve I guess.
So we did some game design in this blog, I’m going back to coding.













































