Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
Discussions Rules and Guidelines
Report this post
The game has no way of knowing. I can't think of any generic criteria for an achievment, that would apply to everyone the same way.
Maybe it would require a different kind of rocket than the Dyson sphere Rockets, but launched with the same kind of rocket launcher. That would be the sort of project to really function as an endgame.
Remember, your civilization lives in a computer simulation (Center Brain). The matrices you are creating represents that computational power. The megastructure that uses Dyson Spheres to generate the energy for a star-sized super computer is typically called a Matrioshka Brain. This is technically what you are making, which for game purposes takes the form of creating Universe Matrices. Since your actual reality is simulated they don't care what happens to the actual universe so long as they can exploit it to power the Center Brain, which is what you are doing.
Now they are bringing in another civilization that lives in the actual physical universe, so they care a whole lot what happens to it. We don't know yet fully what this is going to bring to the game but I think this is going to give it a stronger end game and purpose. We will see.
Unfotunately, no, it is absolutely not 'easy'.
You have inadvertently stumbled upon THE existential and fundamental difference between manufactured computing devices and the computing device that exists in our skulls.
It is astonishingly easy for a HUMAN BRAIN to define or otherwise identify what a complete Dyson would look like. Programming that same things into a video game is something whole orders of magnitude different and difficult for a computer algorithm - even a complex one. Computers are extremely dumb, and can only do what they are told/programmed to do.
Because a Dyson Sphere in this game is completely custom to each individual player, there are, for the puposes of modern computational power, effectively INFINITE combinations of nodes and struts that can be assembled to create a 'complete' Dyson in this game. How do you program infinite combinations of possibilities into a non-infinite detection algorithm? Most people don't have an appreciation or even a real concept of what infinite truly means - especially when it comes to modern computing. How do you begin to define what is considered 'complete'? Is a simple ring around a star considered complete? Players don't even initially have access to the entire latitude of a star to build a whole 'sphere' top to bottom. Technically a simple 6 piece 'triangle' (3 nodes and 3 struts) on the side of a star is a complete shape/loop. If you stop to think seriously about it for a minute or two, you really start to get an idea just how complex this criteria is to define.
This complexity is exactly what the push in AI learning we've seen in the last 50 years has risen from - the ability for essentially a 'dumb' computer to be taught via a very complex set of changeable rules. The power of an AI is not in the complexity of the algorithm itself, as originally thought, rather the complexity of the network of processors that actually run it, as we have learned in this era of modern networking and parallelism.
The human brain has been evolving for millions of years and is an extremely sophisticated network of calculation engineering; allowing us to nearly instantly recognize simple shapes like a circle or a triangle, and almost nearly as instantly much more complex patterns like someone's face or the identity an entirely different species of animal on sight.
All of that to say that it is not 'easy' to create an algorithm to create that kind of detection you are asking for in a reasonable and timely manner - otherwise the Devs are in the wrong business and should be working in research and not video gaming :-p
Now if the Devs added very simple/symetric high-density premade blueprints that players could choose to build around a star (something I'd really like to see because drawing a complete high-density sphere frame in this game with a 'less than ideal' design tool has been my single biggest headache and complaint about this game) then THAT could be used as a 'marker' for an achievement. Plop the pre-made down, and when THAT is done, "Ding! Fries are done! Here's your achievement".
Since the Devs haven't given us that, we do have the next best thing, achievements for reaching certain power generation thresholds of current Dyson structures. Obviously, to reach these thresholds, you have to invest in a decent amount of Dyson construction over an amount of time. It's not the same, of course, as the last rocket of the last structure point of the last node of a fully planned sphere coming to completion. To be fair though, by the time you've reached the point of constructing entire spheres in the game, you've long surpassed the very linear (and thus easy to mark with achievements) part of the game and passed into what is essentially the open-world sandbox part of the game as it was intended by the Devs, which makes for creating additional achievements than what they've already added at least a little more difficult since they'd most likely require a bit of more linear play or intention to accomplish them.
So... yeah... *gestures* All of that.
One obvious part of the criteria would be that there be no more incomplete nodes, frames, or cells in the design -- since you self-evidently can't be complete when the design calls for more.
However making that the sole criteria would be deeply unsatisfying. Someone could design a 1 node "sphere" and 30 small rockets later get this proposed achievement.
But as you went on to say going beyond that gets harder to determine what should count as 'complete' -- though there are a few still some fairly simply guidelines that could be adopted. For example the game could fairly easily calculate total number of components used, percentage of coverage, or radius of the sphere.
So it would be fairly straightforward to design the achievement based on some combination of those. A sphere with no remaining designed items to fill in, possibly of some minimum radius, possible with some minimum number of components, possibly with some minimum percentage of coverage (aka some maximum percentage of "holes").
The small essay that you wrote ... that was exactly my point, but your explanation is much more detailed ;)
@josmith7
Please make room for the idea that a definition of "complete" is not the same for everybody. Simple as that.
For example: In my case, I only build the frame (not the shells), for 2 reasons.
1st: Even a "normal"-sized dyson sphere (I didn't even increased the size of that sphere to the max) with just frames around a blue giant provides way way more energy that I could ever use. So why would I bother to build the shells as well? I wouldn't need that.
2nd: The last time I build shells (solar sails), the size of my save game file expanded significantly and my computer couldn't handle the fact that the game had to deal with mountains of solar sails. And that was a very small dyson sphere around the star in the starter system. I don't even want to imagine the data and performance impacts if I do that with a blue-giant-sized dyson sphere. Sure, the devs had put in some optimizations into the game, but my PC is still shi-, so I'm not eager to go down that road right now again.
My Point is: Whatever kind of definition someone could come up with, it would probably include a certain amount of required solar sails as well. I would never be able to get that achievement, even if my dyson sphere is complete (in my book) and produces 1TW of power, just because I didn't shot any solar sails at it?!
And this is just my specific scenario. I'm sure, you can easily find lots of those. I by myself, can think of several ways how I could "complete" a dyson sphere, and I couldn't find a set of criteria to cover them all and which are also specific enough to qualify as definition. Now ask 2 people, or 10, or 1000. The number of possibilities for a criteria set just goes up, and with it the complexity. Good luck with finding a definition that everybody is happy with ;) And if you try anyway, you will probably end up with something that is too generic and therefore useless (like "finish a node" or something like that).
I wouldn't be surprised if the dev team actually talked about this, and they came up with the idea of this "finish the game" research, where you need 2k (or something like that) white cubes to complete it. A pragmatic and reasonable solution (in my opinion). For comparison, in Factorio you have completed the game in the very moment when you've launched your first rocket into space, and after that the actual game just begins. In Dyson Sphere Program, this is kind of the same thing.
Because "complete" is not the same for everybody my thought is that the devs shouldn't create such an unclear or imprecise achievement. But that if they wanted something along those lines they should instead create an achievement for building a sphere that met some some easy to measure objective stats (like the example I gave).
That sidesteps any arguments about what might count as "complete"; and yet with reasonable objectives absorbing the last rocket or sail on a compliant design would grant you this achievement. They could even set the threshold such that building out the default radius sphere in the starting system would be sufficient to trigger it.
(Though if they went with my idea of 'X numbers of components used' you would be able to get the achievement using a pure shell frame, no solar sails, but you'd need to build a much larger sphere to do so compared to one with sails)
And they kind of did that with We can make it if we try - "Build a Dyson Sphere around a Red Giant star with a power generation performance of more than 10GW." (That particular one doesn't care whether or not there's still room in your design to absorb more rockets or sails; doesn't care if it's just the shell or if you use sails; but it is a clear objective related to building a good sized sphere)
I claim it is easy. I am pretty sure I can write an algorithm that can determine a complete sphere. I can see multiple approaches to do that too.
I am well aware of things you are writing about.
Very beautiful. Now why do we care about it? It is easy to narrow down the problem and make it trivially solvable.
Thanks for a condescending explanation. It is funny that everyone always jump to assume that people they are talking to in the internet are some kids who knows nothing. I am a professional software engineer with many YOE.
I stopped for a minute and came up with a few approaches, that are easy to implement:
- [easy] Achievement for placing 100 nodes/structure points into the Dyson sphere.
- [a bit more difficult, still easy] Completely enclose a star in a Dyson sphere.
- [easy] Plan a Dyson sphere with 100+ nodes and complete it. As in no new rockets/sails needed.
I am sure I can come up with more computationally simple approaches.
This is just nonsense. You have no idea what you are talking about.
Very beautiful again, but pointless. This is just false as I showed above.
This is a working approach. Congrats on thinking about that at least. As you said devs did not go this route.
This is the entire point of having a really cool achievement for completion of a Dyson sphere. Because building the actual Dyson sphere goes way beyond the game completion as it currently is.
Exactly my point. It is not hard to come up with multiple approaches how to set up an achievement that is reasonable. Those people love to chat about computational complexity while having zero idea about it.
Outside of the achievement playthroughs any particular definition does not limit players in any way.
Those are decent examples, but those appear on the surface to be similar to what's already in-game. (Four achievements total: Build Dysons of 200MW, 1GW and 1TW, and my personal favorite, build a Dyson around a Red Giant of at least 10GW) Similar, in that you have to build a certain number of structure points in order to generate X amount of power around stars regardless of the shape of the structure or 'completeness of coverage' of the host star - I imagine the majority of newer players will build their first Dyson around the Home star in the initial game or two, which is an assumption that I imagine the devs made too when deciding those particular energy thresholds... or not? Who knows, the 1GW and 1TW could be completely arbitrary and just nice round numbers. /shrug
I dunno? Maybe? It seems a bit silly to me that we're trying to re-invent the wheel, as it were in this conversation, by adding additional algorithms that are similar in scope to and in danger of running afoul of the 'Traveling Salesman Problem'. I'm not saying it can't be done, rather, that it does leave a fairly large degree of open-ended...ness that then leads to the task of trying to anticipate (and otherwise deal with) a hefty pile of edge cases by whomever is in charge of that section of code later - which as a fellow SE, I'm sure you agree, dealing with edge cases is a PITA, especially if you don't know upfront how many may actually exist in such a theoretical case.
It appeared from your point-by-point response that I may have struck a nerve or that you took personal offense - if so, I apologize, as it wasn't my intent. The degree in Computer Science hanging on my wall (and also my resume), I feel, gives me at least a little bit leeway in (and license to) discussing issues of both computational complexity and also of modern machine learning in this context (and others). Anyone who's been active on the DSP Steam forums for more than a month or two knows that I tend to give longer explanations that most are used to. I try not to assume that everyone else around me has had the same experiences or background as me, so tend to be more thorough in my writing, so if it came off as condescending, again, that was not my intent.
I share information - a lot of it... all at once - it's what I do... heh.
You right those achievements are somewhat covering it as well. I see two issues tho:
- Jump from 1GW to 1TW is huge. 1GW is very easy on the starting planet. 1TW is very very hard and requires hundreds of hours in game. 1GW feels like not a real Dyson Sphere. 1TW is not something a lot of players will ever do.
- Achievements easier than 1TW can be done with Dyson Swarm. Which is not the same complexity and production wise at all.
If there was some middle ground, like 200GW Dyson Sphere, that would be perfect. It is hard but attainable for most players.
I don't see any Travelling Salesman equivalents here.
Even if we are trying to design an algorithm to determine a "complete" sphere, it is fairly straightforward. We just need to go over all enclosed shapes generated by the edges. Save file already has all shapes, edges and nodes listed. Then go over the list of shapes, and check that every shape is filled and completed. This is O(n) algorithm.
Yeah it is just annoying since it happens so often on the forums. People always assume everyone else is stupid in absence of any information.