Friday Facts #373 - Factorio: Space Age

[Shuisman]

[...] Hard to see even for non colorblind people

…A different shape to the flask wouldn't help much either…

The flasks are actually slightly different. While the original were Erlenmeyer flasks, the new ones are round-bottom flasks.

[FutureSpec]

Great to finally see an update with real details. Like others, I do have a few concerns:

• Pushing cliff explosives and artillery back seems to make little sense: cliffs are notoriously so annoying that many people turn them off, and artillery has long felt too delayed already in the tech tree
• I wonder if there will be a meaningful win condition aside from just hopping to each new planet
• I'm bummed the DLC is "just" a refined version of the space exploration mod -- I was hoping for major new mechanics like tunnels, bridges, or making water more meaningful (boats, water enemies, etc.). Maybe I'll be proven wrong if we see big new features announced in coming FFFs. New enemies? New weapons? Biomes that are actually meaningful and not just cosmetic? Dyson Sphere Program does this well: e.g. you can setup thermal power plants on a lava world, solar on a tidally-locked planet, etc.

[Korgunaard]

Will the new features in the official Space Age expansion allow for more features in Space Exploration mod?

[Jesperhk]

I'm excited as always, can't wait to try Factorio: Space Age and i'm really happy for the FFF return, it was the best thing about friday's

[Mathematician]

This is a very good news indeed. I'm very happy to see Earendel on the team as SE is my choice when it comes to overhaul mods.
Some time ago Factorio has pushed beyond my nostalgia and claimed place of my favourite game and now I raise a glass for bright future of Factorio 2.0.

[kovarex]

but the player is quite omnipotent at the end of the vanila playthrough so we had to make some tough decisions

I do hope the expansion will give better things to play with though, maybe productivity 4 modules, power armor mk3, better spidertron or higher tier equipment. In vanilla player is very far from omnipotent when it's about megabase building, one has to rely on mods to be efficient. I don't think i could bear to play a MP server that is only vanilla. Can't fit enough exoskeletons to get decent movement speed, not enough roboports, having to swap armor if you want all functionality etc etc, vanilla has tons of room to add tools for player.

If you consider some sort of PvP balance, leave that out of it. Those scenarios can make their own rules and balance, what builders like is fun building. In combat side many more enemies can also be introduced to give more reason for the better tools. It has stagnated to the behemoth phase, people seriously don't even consider them a threat anymore.

Don't be worried. I'm also a progression addict, so the progression ceiling is extended for most parts of the game through one of my most favorite features. It will be explained in one of the future FFFs.

[wild_dog]

I'm wondering something about optimisation.

Factorio as a game is incredibly well made from an optimisation standppoint, and in previous blog posts, the difficulties about implementing multi-threading were discussed. The game already makes use of a remarkable amount of multi-threading, but there are still quite a few single threaded bottlenecks as far as I can tell.

With the space exploration mod, I once either asked or read a discussion with Earandel about if the different planets are also parallelized. That seemed like a perfect place to implement multi-threading to me, as different planets usually don't have that many interactions, so running everything to do with updating the game, even the things that normally have to be single-threaded, parallel per planet seemed logical to me.

The answer was that the different planets weren't multi-threaded. SE made use of built in surfaces functionality, it didn't implement running multiple plantes within the mod itself, and different surfaces weren't updated in parrallel. Since that is engine level, a mod couldn't implement parallelisation there even if it wanted.

Now with the vanilla game also introduciing multiple planets, and likely also making heavier use of the multiple surfaces, is parrallelising the surfaces for the game update something that is comming?

The last few years have seen an increase in core count (even if they are efficiency cores in Intel's case) and cache sizes (AMD 3D V-Cache) in even mid/low level consumer CPUs, even if the single threaded performance increase is stagnating. So taking even more advantage of parralellisation/multi-threading seems like a good optimisation srategy.

[Drury]

Awesome !

And the return of the weekly FFF to tide us over the wait until release !

And is that a cooling system that I see on those engines ?!

(But where are the radiators to get rid of the heat ?)
avatar_spaceship.jpg
EDIT : Oh, looks like the robotic arms are pulling in asteroids, which are then crushed and sorted, and the ice from them is then used for cooling in "chemical" plants ?!

Looks to me like the ice is melted into water which is then magicked into liquid hydrogen and liquid oxygen and fed into the engines.

[aklesey1]

Wnat to listen about new "better control train systems"
it will be interesting to be able to view the contents of train cars directly in a separate menu and not just through stations

[Atraps003]

The last few years have seen an increase in core count (even if they are efficiency cores in Intel's case) and cache sizes (AMD 3D V-Cache) in even mid/low level consumer CPUs, even if the single threaded performance increase is stagnating. So taking even more advantage of parralellisation/multi-threading seems like a good optimisation srategy.

https://www.factorio.com/blog/post/fff-324

The underlying algorithm didn't change but it just ran > 3x faster now by touching less memory. This is another nice example of "Factorio is not CPU bound, it's memory latency bound". More cores wasn't going to make this faster - because it was never limited by how fast the CPU ran.

I doubt there is any optimization available that is going to allow for much larger active entity counts when aggregated across all the new surfaces other than clusterio-like support.

[OldFart]

The only thing I actually need is train improvements:
1. Possibility of higher throughput for the single train
2. Bridges or tunnels or both.
The thing I want is a gate to teleport between distant parts of the base

But space... Biters will chew up my main base to bits while I'm prettying up a mini base on another planet...

[NastyFlytrap]

Isnt him being the creator of Space Exploration a conflict of interest? @Devs? Anyone?

[Sigma1]

haven't been here in a while, this was definitely interesting enough that i have some thoughts however

first off, the expansion looks great, seems like it fits the Factorio visual style well, and is concise enough to not have my main issue with the SE mod, which is that it's incredibly long-winded for my taste, whereas Space Age seems to go more directly to the space parts. overall, i'm definitely excited, and will likely get a ton out of it, especially when combined with some smaller mods

however the balancing changes i do have some concerns about. especially cliff explosives being moved further up the tech tree is not a great move in my opinion, as they're a QoL feature more than anything, which i don't think should be locked to the late game. especially with how often cliffs end up in the way of your factory even now. artillery meanwhile i have mixed feelings about, it feels like a sort of win condition even, so in that sense i definitely support moving it further. however from an immersion standpoint it has always felt weird, and this'd make it feel even more weird, as i often end up having nukes before artillery, which is just very strange-feeling progression, because in the end artillery is really just an enormous gun, and it's not like the engineer hasn't built guns before. meanwhile, nukes since being added have felt a bit underwhelming, so i think a better idea would be rebalancing such that artillery is a good mid-to-endgame defense rather than a proper offensive weapon, and nukes become the high end offensive tech you need to reach space for, which'd make the progression feel a bit more natural in my opinion. now, granted, that's only something i thought of while making a coffee, and probably has some issues i haven't thought of, but point stands

oh, also, i do hope we hear about the train control soon, those are among my favourite part of the game and i'm interested in how they can become even better (speaking of which, on the topic of rebalancing, imo use of longer and mixed freight trains could use more incentives )

[sciencemile]

Isnt him being the creator of Space Exploration a conflict of interest? @Devs? Anyone?

A conflux of interest more like. I don't see the problem?

[PssX]

Engine Support: There are so many things that Space Exploration just can't do because it's a mod.

But after release of the expansion, would other mods be able to use these new features added to engine because of the expansion? Or it will be expansion-exclusive features, i.e. expansion will have unique way of interacting with the engine? Not like a general mod.

[Nidan]

Engine Support: There are so many things that Space Exploration just can't do because it's a mod.

But after release of the expansion, would other mods be able to use these new features added to engine because of the expansion? Or it will be expansion-exclusive features, i.e. expansion will have unique way of interacting with the engine? Not like a general mod.

That has already been answered in an earlier FFF: If mods want to use expansion exclusive features, they need to depend on the expansion.

[bayesianmind]

Anything that is part of the 2.0 engine and not tied to expansion content will be used by SE.

SA won't be a hard dependency, i.e. you won't NEED it to play SE.

SA will probably be an optional dependency, so extra features be used if they are unlocked. This will mean that SE will be a bit different based on if you have SA or not. I don't really like having 2 versions like this, at least not while I'm still changing so much in the mod, but it seems unavoidable unless the SE playerbase overwhelmingly already has the expansion, but I don't know what the acceptable % would be for it to be ok to make it required.

Either way, things are changing on both sides and I have a long time before I need to start making these decisions.

Space Exploration's core audience is advanced players who have put in a lot of energy on Factorio and who want ~150-500 hours of challenge. Surely the expansion cost isn't much to ask given that many hours of quality entertainment? Many games provide far less for more money. Given that everything is a tradeoff it seems more valuable to be able to focus on more features sooner rather than spending a lot of time trying to make SE work well without the expansion. If they still don't want to buy the expansion, they can play an older SE version.
Given how painful timing and compatibility mod conflicts already are requiring the expansion seems like an easy way to reduce development complexity to me.

[Chindraba]

Engine Support: There are so many things that Space Exploration just can't do because it's a mod.

But after release of the expansion, would other mods be able to use these new features added to engine because of the expansion? Or it will be expansion-exclusive features, i.e. expansion will have unique way of interacting with the engine? Not like a general mod.

That has already been answered in an earlier FFF: If mods want to use expansion exclusive features, they need to depend on the expansion.

Best as I recall, the intention is/was to make the "engine" the same everywhere. Only one set of code for the devs to keep optimizing like nobody else can or has. Some kind of switch or setting in the code of the expansion to indicate it "is" the expansion and that it's turned on. Allows SA to load (it's probably going to be completely possible to get the SA mod content, just won't load without that flag) and that other mods can read to load compatible features with SA or stick with Vanilla features.

Frankly, I don't expect they will even have any of the engine's abilities gated behind such a flag, just the loading and running of the SA "mod".

[Sopenas]

Oh god, just tell us if you are going to implement train tunnels or ramps. That would be the best feature i could wish for

[kovarex]

I wonder if there will be a meaningful win condition aside from just hopping to each new planet

Yes, there is a win condition, and as in vanilla, you can just stay and continue playing and doing infinite research if you wish.

I'm bummed the DLC is "just" a refined version of the space exploration mod -- I was hoping for major new mechanics like tunnels, bridges, or making water more meaningful (boats, water enemies, etc.). Maybe I'll be proven wrong if we see big new features announced in coming FFFs. New enemies? New weapons? Biomes that are actually meaningful and not just cosmetic? Dyson Sphere Program does this well: e.g. you can setup thermal power plants on a lava world, solar on a tidally-locked planet, etc.

The space part is the cornerstone/anchor of the expansion, but there are features which extend the game regardless the space part as well. (In the current state, the expansion is actually 3 mods, and you can have 2 of them without having the main space-age part turned on.)