Shut up!
I would snatch that so quickly I'd get friction burns...
Canis Major Dwarf Galaxy... More fun for explorers!
- Thread starter Greasetrap42
- Start date
I would snatch that so quickly I'd get friction burns...
We seem to be working at cross purposes, the galaxy is proceduraly generated from a seed, you can't just "add stuff," outside it because, despite what you see in the game, there is no "outside" to the ED galaxy, what you are looking at when you are at the edge of the galaxy isn't a void you can fill with stuff, it's a picture! This is a game, there is no "outside"!
Of course maybe if you created an entirely new galaxy on a different serve with no connection to the current galaxy, like original Elite with their 256 galaxies. But that was an entirely different model and was built like that because the computers of the time couldn't load any more data due to current microprocessor limitations, the idea of this ED is that the galaxy is entirely one object with everybody exploring the same thing.
Doing it that way would mean essentially duplicating the entire game and all the infrastructure associated with it, it would esentially be two different games that FDEV would have to support but only the same number of players spread between them, and moving from one galaxy to the other would mean being entirely seperated from the existing player base, which completely destroys the intention of ED that everybody is playing in and affects the same galaxy. That would be a costly and ultimately ineffective way of doing things.
Maybe a different game, the next version of Elite, could include things like the dwarf galaxies and other new stuff we have found, but that simply isn't going to happen here because the galaxy model in ED isn't designed to work like that and nothing you can say will change it to work like that. The seed that generates the galaxy generates the galaxy and that's it, if you want more stuff generated such as dwarf galaxies then you need to use a different seed designed to generate said dwarf galaxies as part of the galaxy model.
Of course it should also be pointed out that the reason the galaxy is the size it is is simply because of the limitations of our microprocessors, the number of possible stellar objects in the galaxy is limited by the fact that we are using a 64bit number to assign a unique individual ID to every stellar object in the galaxy, there are none left to create a seperate galaxy anyway, so creating dwarf galaxies as part of the model would actually mean having a smaller, much smaller, main galaxy.
"Why you want to get stars from Canis Major when we've got stars at home"
- Space Mom
Seriously though, that's a nice idea for an expansion. Maybe not for the near future, but pretty cool. That diagram (A) that CMDR Maylor Rom found inspires imagintion. (And it also helps address the FSD range-creep that comes with each expansion needing to use ship upgrades as a reason for us to Do The Thing)
- Space Mom
Seriously though, that's a nice idea for an expansion. Maybe not for the near future, but pretty cool. That diagram (A) that CMDR Maylor Rom found inspires imagintion. (And it also helps address the FSD range-creep that comes with each expansion needing to use ship upgrades as a reason for us to Do The Thing)
Deleted member 38366
D
Galaxies 2 through 9 (increasing weirdness and SciFi stuff)... confirmed
Manual FTL jumps with Carriers... a natural must-have
(hey, one can dream
)I don't think that's a huge problem (which is not to suggest it wouldn't be a lot of work). Procedural generation at the level of E-D already has many steps between seed and output. Adding additional steps that do not change the established output but add additional points to the output is possible, and likely possible at galactic scale too. We don't know exactly how FDev has architected their system or how time-consuming it would be to shuffle in that kind of system, but as you point out we do know that they can and do make manual edits, which indicates that the seeded output is already being combined and/or adjusted by the output of other systems.
For a new section of galaxy, you'd need it to be procedural rather than database, but I don't see any reason why certain sectors of the galaxy edge (that intersect with a new galaxy) couldn't run two proc-gen star-generators (established galaxy and expansion one) and combine the result. The extra process would presumably be less processor-work than the existing proc-gen at the core of the regular galaxy where the star density is insane
(It wouldn't be as astronomy-perfect to do it that way, but I'm just describing a method that favors developers. Whether hypothetical devs might want to set themselves a higher bar is up to them and their hypothetical resources
)You could probably even make it work like horizons where the new galaxy doesn't appear for players without the expansion
(Changes and additions will definitely involve lots of problems and lots of work, but I expect the problems are solvable for something like this if the result justified enough time to do the work. That said, there might be a deeper issues like with coordinate space - I'd guess that it's cubical but if it's optimized more like a pizza-box shape of the milky-way, then you might need a new coordinate space and that would probably be a nightmare unless it was made with a lot more future-proofing than they probably had time for)
Last edited:
What magic are you going to use to pull extra unique ID's from the 64bit integer that supplies all the ID's of all the stellar objects in the galaxy?
Perhaps similar magic to how space games like Elite already exceed 64bits of coordinate space. Eg perhaps you compartmentalize. Perhaps the second system runs in parallel. Perhaps there is enough unused address space. Perhaps it's a nightmare best left alone, perhaps it's already compartmentalized. Hard to guess the best way without being in the trenches, but these things are generally an issue of whether it's worth the dev time or tech debt rather than an issue of possibility.
Last edited:
Discussion by others so far has been emphasised as "just adding a bit to our galaxy." I have already discussed the compartmentalising issue where you essentially double FDEV's support cost but add no new players to the game, I can't see that as a supportable way to go since they are trying to keep costs down to allow a subscription free model without the undesirable adding of play-to-win mechanics other F2P games employ.
The entire philosophy of the ED galaxy is that it is an entire modeled galaxy for players to explore, once you start compartmentalising you lose that basic design philosophy and risk breaking the game up into lots of smaller play areas each of which need extra support. You also lose the unique ID for all stellar objects which means many of our tools will no longer work since inevitably ID's will be duplicated and cause confusion.
The "just adding a bit to the galaxy" suggestion can't be done and the "compartmentalising" idea, while doable, is probably undesirable both financially and philosophically for the devs.
I don't actually think any of them do! You will need to provide an example, ED is unique in it's modeling and method of simulating the entire galaxy.
You suggest that a significant difficulty is that over 99.6% of Stellar Forge's 64bit ID addresses would have already been claimed by the Milky-Way, but if I were to guess I would think a lot more than 0.4% of address space is still available. Is there a reason to think otherwise? Did Stellar Forge already run out of IDs or something? I'm inclined to assume there are enough IDs outright for a dwarf galaxy, but that said I'm also inclined to agree that it's unlikely the effort of adding one would be seen as worthwhile by FDev with the existing galaxy already so vast. (And there would definitely be players who view it as misdirected dev-time and get angry)
Last edited:
The answer to that is, yes. The 64 bit address space is divided up into sections such that every solar system in the galaxy has a reserved 256 unique addresses for all cosmological objects in that system, current and proposed future object of course. They aren't all used, but they are all reserved for current and future content. Every single star, planet, moon, gas giant ring, asteroid ring and cluster has a unique id in that 64bit address space, which comes out to over 100 quadrillion unique ID's. required to model the galaxy.
Sounds like there's plenty of addresses for enough extra stars and solar systems then, so the difficulty is spatial? Do you mean the addresses are divided into grid sectors (ie spatially-located), and you think the sectors are arranged too much like a pizza-box and don't stack high enough to allow a very good depiction of a dwarf galaxy? Something else?
I'd guess the unused sectors in the corners could be "relocated" to where they're more useful, but yeah, time and effort vs payoff.
I'd guess the unused sectors in the corners could be "relocated" to where they're more useful, but yeah, time and effort vs payoff.
Last edited:
Not it's heriarchical, it starts from the top down, galaxy, region, sector, star, etc. You can't pull unused ID's from a solar system to use in a different section of the galaxy and you can't pull unused ID's from a sector to use in another sector because there aren't any unused ID's in a sector.
I will drop this here:
With the thread it's from;
Myth busting on Stellar Forge and the generation of everything from stars to rocks.
It became apparent at Lavecon that a lot of players are still under the misconception that there's some kind of database of discovered systems and worlds on a Frontier server somewhere and that when another player goes to visit these places the game client downloads this information so they see...
FWIW I know intimately how to do proceedural-generation (I've been in the game industry for ~20 years, the most recent project is even based on a seeded proc-gen world-builder so I'm dealing with some of these issues currently), I'm trying to figure out more stellar forge detail from your description, since it sounds like you've examined an aspect of the structure (for 3rd party tools I assume). It actually sounds like you're describing what I'm envisaging / trying to describe, but text forums are just kind of bad for this sort of thing. The messages you're quoting aren't referring to the sort of approaches I had in mind, though I can see how it might seem like it. My hunch is that the hierarchy could be used so that changes within a sector propagate further down the hierarchy with limited propagation sideways or up it, which can be used to silo off an override that could open the kind of approaches I had in mind. Though as your quotes also highlight, the system is live (and unusually complex) and it's more than understandable that they wouldn't want to touch it. In those shoes I can totally can see me saying "I think we can do that, I believe we are capable, but just nope!"
Last edited:
I don't really see the interest of another galaxy.
I think Frontier has a lot more important things to do.
Lots of work for Frontier for just a panoramic view ?Well, imagine the vistas.
The highest I've been "above" the Milky Way was some 2500LY, I think. And the view from up there is already stunning. Imagine 25k.
I fully appreciate your reasoning however that is focusing on what can't be done, not what can be done. In my experience pretty much anything is possible given time, money, and expertise. Limiting our thoughts to what we "think" is possible/impossible is very... well... limiting. Making assumptions based on how things are currently done or our own personal current experiences. Saying "it can't be done" is a very limiting statement. In the last 30 years I have worked on numerous projects that were supposedly not possible. But when the right people with the right ideas are put onto a project with a ton of money its pretty amazing what can be done. Outcomes can be very surprising. I have also found there are people who ideas and abilities far surpass my own or others in my projects.
I'm not saying FDev will add another galaxy. Maybe they will or maybe not. And I have no idea if its worth the cost/effort. But I am quite certain that if this game survives another 5 years they will certainly be adding significant things to the existing galaxy map. If this game looks exactly the same in 5 years it will be dead.
Are there really enough stellar objects as to run out of possible 64 bit tags? Mind you, with 64 bits you have 18.446.744.073.709.551.616 possible combinations (i.e. tags), that's well above the 400 billion star systems and the objects we currently have.
I doubt it, numerically speaking each of the 400 billion star systems could have well over a million objects each and we would still have room left over in the 64-bit integer.
It's not so much "It can't be done", but "it's really, really hard to do, and there's no motivation to do it".
There's plenty of stuff in the Real-world galaxy and galactic neighbourhood that haven't been added to the game. Globular clusters, for example. They're like miniature dwarf galaxies and there's about two hundred of them, out there in the Halo. Not one of them has been added to ED.
There's plenty of stuff in the Real-world galaxy and galactic neighbourhood that haven't been added to the game. Globular clusters, for example. They're like miniature dwarf galaxies and there's about two hundred of them, out there in the Halo. Not one of them has been added to ED.
Last edited: