you can not craft the components needed for CPU extenders. you must find them in pink or red loot crates or find the trader that sells them and grind the credits to buy them. Grind is in that either way. And we all know how easy it is to find things in space with the "fine" detection system we have in space. It is my understanding that is where the trader is.
I just want to have fun in a game that is balanced to itself and plays well with enough believable mechanics that I can suspend any other disbelief. I am simply immersed.
A ton of games do this with playing fast and loose with RL physics. There must be a school of thought in game design that this deliberate ignorance of physics and realism is somehow necessary for a balanced game and makes things better and more fun. For some games this is fine, but builder games tend to attract players who are more science minded and care more about logical consistency. I'm not saying exact RL values should always be used, a bit of fudging is fine so long as the overall feel is still correct and is implemented consistently and logically. Often it seems like devs like to use words like nuclear to convey a powerful energy source, but then aren't prepared to actually implement anywhere near realistic values for it since in RL it's over 1000x more enegy dense than hydrocarbon fuels, not to mention solar. Empyrion does this with biofuel and prometheum, normalizing them to the point that either biofuel is black magic, or prometheum/fusion cells are something else than the nuclear fuel they're implied to be. If a RL system/value breaks your game and is too far off to be fudged, then come up with a new pseudoscience explanation for the system instead of misusing and abusing RL terms. With the motion sensor example a few posts above, you've got an example of another immersion breaker where stuff in a sci fi futuristic game works MUCH worse than RL for Reasons. It would be far less immersion breaking to have basic logic like this use zero power with how little modern computers use relative to the amount of energy needed to move a ship. On the other side, thrusters use far too little energy for how much mass they're moving. This leads back to the topic of CPU. The fundamental problem IMHO is that it's an illogical system that doesn't really add anything to the game. Sure, I can still play the game and build functional ships with it, but I would much rather see the effort being put into balancing this system used instead to balance the mass/volume/energy/armor/damage systems we have now. The way CPU tiers works defies logic. Intuition tells us that if we don't have enough processing power, just add more processors until we have the performance we need. Not this silly thing of no gain until you put the magic extender blocks together just so. Not to mention the sheer amount of volume and energy these cpu blocks require for their performance defies intuition. Additionally, the way CPU penalties are applied (especially to thrusters) sounds more like what power would be expected to do instead of CPU, adding further logical confusion as to what this system actually is. From a gameplay perspective, CPU seems unnecessary with the much more logical mass/volume/energy systems already in place. Want to force specialization? Make armor and other components properly heavy and thrusters properly energy hungry so you'll need to pay attention to the mass you're adding to your ship. A warship will need to devote mass to armor, and won't be able to haul much. Combine that with the existing block limits and you've already got a framework that effectively does what CPU is clumsily trying to do. Anyways, TL;DR I would like it if game devs would do a better job with logic, internal consistency and capturing real physics in their game systems instead of just throwing out arbitrary values and incorrectly using RL terms.
I think the use of CPUs needs to be a bit more purposeful. Instead of just capping CPU units the CPU's should offset specific efficiencies. Perhaps different types of CPUs that steer the function of the ship. Eg. fighters should not be allowed to carry heavy weights. Haulers should not be allowed to have a large number of offensive weapons. I'm not sure how they can do this without making even more complicated.
CPU doesnt have a hard cap. The more you go over the more inefficient "everything" becomes (its gradual). So what you may be proposing is a way to 'designate efficiency priority' ... much like some sci-fi games let you choose "more power to thrust (reduced shields/weapons)" "more power to weapons (reduced thrust/shields)" ... etc? In otherwords, in THIS system, setting "priority thrust" would make sure the CPU efficiency for thrusters is 100%, and whatever the resulting inefficiency is, is divided between shields and weapons. etc etc etc
It HAS a hard cap -- it's just a bit further along than where the penalty starts. The hard cap currently is at 3 x CPU Limit.
It does? I've seen CVs at so far beyond the cpu limit (80mil pts on a t1 cpu lol), but the % drop stops at 50%. I guess their readouts are WRONG ... yet again. I wish they would get that **** under control. edit: .... or ive been jumping around in so many alphas I dont know end from up anymore on this cpu jizz ..... LOL. So, never-mind.
Yes, it does. Hummel confirmed the penalty is linear. It's just the method of description used ( penalty at +100% CPU limit is 50%) is really poor and misleading. That means up to 100% CPU limit the craft operates at 100%, at 200% it's fallen to 50% and at 300% it is at 0%. The first penalty of +100 CPU was ~90% pushed the hard cap way out to around CPU Limit x 10. The new proposed lmit of CPU +100% = 20% effectiveness drags it back to 2.25 * CPU Limit.
Well said. Yes this is the type of thing I was thinking. But not sure how to do something like this without making it too complex. I like the relative simplicity of Empyrion (vs another Space Engineering game). Prior to CPUs I focused on building the one SV for everything. Put 6 of each type of gun. fill full of storage. Lots of thrusters and RCS. After a few days playing with CPUs I have to think more about the function of my ship since i can't put too many high CPU value items. Hauler can really only load up on thrusters. I see the potential here to have specializations with the CPU concept. I dont know maybe adjust the item limits based on CPU specialization. "Thruster" specialization could increase how many thrusters you can add. "Offence" will increase the number of weapon types. (ok i will stop here. I'm not that creative. )
I hear ya. I used to create a "Raider" type SV, which was used to first attack POI at a distance, then serve as a mobile safe station (health, o2, constructor, clone) while I foot it inside, and then as a hauler for all the loot and materials as I would rob the POI blind. Not so easy to make a ship like that anymore
This idea could work quite well, and let you build ships that don't have enough energy/CPU to run everything at 100% (although it would also need a way to turn things off when not needed to really work well). By designating system priority you could control how performance degrades in a combat situation where you're pushing the vessel beyond its nominal limits.
The thing is, you don't notice properly-implemented physics. But you do notice poorly-implemented physics (acceleration-dependent max speed cap, anyone?). Yes, a priority system for CPU (and power, for that matter) as well as the ability to shut down subsystems to decrease CPU consumption would give a lot of flexibility. One could also imagine a kind of overclocking mechanism that would allow devices to exceed their normal operating parameters in exchange for additional CPU and power consumption (but either only for a limited time or causing the affected devices to take damage over time). So you could, for example, shut down all weapons and non-essential systems, divert 150% power/CPU to thrusters and shields, and make your escape.
I have a similar approach, but get away with just vessels using a Tier 1 Core: 1. Use an SV to eliminate a POI's external defenses and return to CV. 2. Return with a light HV, park outside of the POI's patrol radius. 3. Approach POI on foot and raid it. 4. Return to HV and then to CV. 5. Park the CV by the POI, loot and salvage the POI. Variations: If I don't have a CV, I loot the POI later after I get one so that I can loot "big stuff." The HV has a little bit of cargo that I use to strategically loot crashed ships to get the needed CV materials. I could use a motorcycle instead of an HV. Notes: My starter CV is basically just Cargo, Constructor, Fuel and Thrusters. It doesn't even really have a hull beyond a little to dock the HV and SV.
Question for the devs: if 90% of players refuse to use the CPU system, will you admit you royally screwed up & do a major overhaul to how it works (& this time you mean it?), or will you just make that other 10% your target demographic? I mean, the story being pushed is that the system is supposed to encourage specialization. The problem is it doesn't really do that. On a T2 vessel, it's a choice between a warp drive, a constructor, & an extra gun. On a T3 vessel, it's a choice between a hull shield & every other basic functionality in a pile. That's it. There's no other specialization. The only other things that fail so spectacularly at their manifest goal are american federal laws, but at least in the case of those laws, it's because the person pushing them is lying. Here, it's just failure. Then we move past manifest goals & just look at what it actually accomplishes. Let me just say, I've seen more direction from a college student whose changed their major for the 4th time. The name & description imply a vehicle-based skill point system. The mechanics look more like a power system. The fact that hull blocks have a CPU cost throws that theory out & replaces it with "block count limit". For small vehicles, it's a tool limiter. For CVs, it's a thruster limiter. Bases are just victims of the bureaucratic process. HVs are more expensive than SVs, but have a lower CPU limit. & I can't test any of this outside of creative because there aren't enough hours in the day for the grind needed to make use of T3 constructions. Has anyone actually made a T4 CV in survival mode since the patch came out without using console commands? Asking for a friend. None of it makes sense, it's trying to cover too many bases at once, it doesn't fix any problems in the game, & the more I try to describe it, the more I think it came from the kid on the playground who makes up the rules as he goes. I'd strongly considered making a guide on steam to help players learn CPU, but the more I build it in my head, the more I just want to tell them to turn it off. Then I'd spend the rest of the guide ranting about it. But here's the thing, I was willing to look past all of this & try to work with the system in play, except for one glaring issue: it isn't fun. Even when I try to ignore the lack of sense or structure, I can't have fun with it. Survival has become a chore. Creative is even worse; instead of building to an idea in my head, I'm building to a CPU limit. It's a seven-layer dip, except 3 of the layers are sand. Even if I try to ignore limits & design a T4 CV, my computer starts overheating when CPU gets too high, probably because the numbers are too big. So there it is, I'm turning off CPU until it starts to have some redeeming quality. Right now, it is utterly pointless & less than useless. If you wrapped it up in a gift box & sent it to me in the mail, I'd throw it out & thank you for the bubble wrap. Do better.
I try the same, but it works a bit differently for me: Use an SV to eliminate a POI's external defenses. Take too much fire, retreat, repair. Return only to lose the front of the craft because I missed a spot. Respawn at home base. Grab a backup SV and complete the job. Return victorious Return with a turret-mounted HV that includes a clone chamber and park outside of the POI's patrol radius. When I hear the turrets fire, start to realize the HV is in a drone patrol area / the patrol area is larger than I thought. Rush back to try and save the HV. Move it to a safer spot. Approach POI on foot and raid it. Respawn at the HV when I get shot in the head from the side by an unseen enemy. Respawn at the HV when my legs dissolve from spewed acid as I take too long to kill some form of nasty. Respawn at the HV when I realize some of the spawners are invisible as enemies take me from behind despite my very careful clearance of the area. Respawn at the HV when I trigger a pop-up sentry gun trap and freeze in place trying to see where the shots are coming from. Find the core and kill it. Return to HV and then to CV. Park the CV by the POI, loot and salvage the POI.
1. Use combat SV to snipe external defenses 2. Return to CV 3. Fly CV to POI 4. Have cannon turrets target the base to show where the core is (or look it up) 5. Park CV on base 6. Use SV to drill to the core and destroy it 7. Core the base and claim it for yourself 8. Loot, salvage and profit
Didn't they take away core targeting as an option on turrets? I used to use triangulation to at a minimum reduce the amount of searching, but when I approached a POI in 10.5, I didn't see "core" in the HV turret choices.
Maybe, haven't done a full survival game recently, although now that we have any to any docking and project eden I might try it out again and turn off CPU until the dust settles on that system.
@sillyrobot try building one of these for base infiltration newly updated for cpu, 98% efficient. be warned, the single rcs is required but makes rolling very twitchy. I don't like this new flight model. https://steamcommunity.com/sharedfiles/filedetails/?id=1917882105
