Discussion: long-range logistics revamp
-
OPERATE_POWER will become useless
-
I like the novel approach to warp container upkeep. If the repair route was more energy efficient it would create a more interesting tradeoff.
I think warp container upkeep is a bit too cheap (as others have pointed out). I suggest an upkeep cost of 3 energy per tick.
A sketch of the costs of offroad remote mining to use as a baseline:
- ignore reserver and drop miner costs since they don't change.
- .5 energy / tick for container upkeep (not 1 as stated elsewhere).
- 1.66 energy / tick for a 25C25M hauler up to 60 tiles travel
The upkeep of 2 warp containers must cost more than 2.16. I'd double the upkeep to be absolutely sure hauler remote mining is more efficient. So 2.16 energy per tick for each one, but round up to 2.5 or 3.
Keep the energyless decay where it is (needs 1 energy / tick of repair). Then players can juggle resources and repair the container spending more intents but improving energy efficiency.
It might make sense to keep the upkeep costs low for Claimed rooms to encourage players to incorporate them into base design.
-
One suggestion that came up on slack to benefit networked warp containers without making them too powerful for remoting was to decrease the upkeep significantly in highway rooms.
@zachiever said in Discussion: long-range logistics revamp:
OPERATE_POWER will become useless
The argument being that you can't use more than 50 e/tick in a room due to terminal limits? I guess you have to pair it with
OPERATE_TERMINALin this case, or use creeps/warp containers to move energy in.
-
Hi,
I dislike these changes, but would like to know the ideas and purpose behind these changes.
Changing two parts of the game that feel good, terminals and remote harvesting.
What is the design idea behind these terminal changes? Is the terminal change's purpose in order to nerf terminals while being under siege by players? GCL rooms? Or just upgrading bursting? Or all the above in hopes that we use power more?
For Terminals instead of creating a new variable 'bucket' that will limit transfers, why can't it just use cooldown. Recieving would add 10 cooldown to terminal, multiple receives would add 10 cooldown. Terminal operate can reduce by 50%. While terminal disable increase cooldown or stop cooldown countdown.
Warp Containers seem interesting until you said they can be built in non-controlled rooms. These seem like a best solution for remote harvesting with a downside of energy maintenance, making previous solutions obsolete. With the way you decided to balance this through energy maintenance - it'll either be too good or useless due to energy. Or balancing this by not allowing it to catch dropped resources(not a big deal, literally talking about just a creep.transfer as this being balanced/unbalanced?) . This doesn't feel like it will create code diversity, and creating a "single best solution" to remote harvesting.
By creating warp containers, what's the purpose of links now? Doesn't make links obsolete? Can't we just have operate link that makes it global? A operated link can send to any other link in any room? Or a operated link can receive from any link in any room?
We were talking about high player sieging/fights in other posts, and I don't think a single player mention terminals as an issue, and yet here we are changing it. Due to changes in terminal you want to balance it out by breaking remote harvesting, something that I think everyone currently enjoys as is.
-
I dislike all changes mentiond on first post, especially the terminal limit looks complete garbage to me.
-
@o4kapuk "a simple fix could be prioritizing deal transactions and intra-player send transactions over other transactions. Not a complete fix, however, but at least something. If this is a problem to be fixed (I'm not sure about it either)"
Attacking a terminal every tick by overwhelming, the counter you're suggesting is to always send materials to get priority sending.
But it doesn't fix the problem - where I can send over and over again every tick multiple times to attack the terminal.bucket, while you send what you need. You may get what you need in, but by then I've sent 10,000's of unusable resources.
This will make it more so I would rather attack the terminal through send than disrupt. One requires power enabling and be inside the room. The other just requires terminals. So creating a 'best' solution.
I have done this type of attack before and while it's a hugh drain on resources, if the target player isn't active it can be effective. .
-
@likeafox said in Discussion: long-range logistics revamp:
We were talking about high player sieging/fights in other posts, and I don't think a single player mention terminals as an issue, and yet here we are changing it. Due to changes in terminal you want to balance it out by breaking remote harvesting, something that I think everyone currently enjoys as is.
I dunno. I'd argue that terminals are part of the issue with seiges (ie. magically being able to circumvent any blockade), and that this change is an improvement. I think it moves disrupt terminal into a lot less of a knife-edge place, which is good (ie. previously if you lost the blockade for 100 ticks, there could be 100k+ resources moved, now it's 5000 without operate terminal). It does greatly limit repairer spam if you're limited to 70 (two source, 50 from terminal, rest blockaded) energy per room to defend with. It moves combat to more of a multi-room thing. All this can be nothing but good.
The limitations on civilian uses are a side-effect but not too onorous. I don't think many people are sucking in more than 50 resources/tick sustained in any room other than temple rooms. For temple rooms, there's
OPERATE_TERMINAL.
-
@likeafox said in Discussion: long-range logistics revamp:
@o4kapuk "a simple fix could be prioritizing deal transactions and intra-player send transactions over other transactions. Not a complete fix, however, but at least something. If this is a problem to be fixed (I'm not sure about it either)"
Attacking a terminal every tick by overwhelming, the counter you're suggesting is to always send materials to get priority sending.
But it doesn't fix the problem - where I can send over and over again every tick multiple times to attack the terminal.bucket, while you send what you need. You may get what you need in, but by then I've sent 10,000's of unusable resources.
This will make it more so I would rather attack the terminal through send than disrupt. One requires power enabling and be inside the room. The other just requires terminals. So creating a 'best' solution.
I have done this type of attack before and while it's a hugh drain on resources, if the target player isn't active it can be effective. .
It changes the defense from "50 per tick + 300k bucket" to "50 per tick". The best solution is to disrupt and spam raw minerals.
It's better than nothing (which it would be without the local priority) but it still kinda sucks.
-
I think we can implement something like:
StructureTerminal.setWhitelist(['MyRoom1', 'MyRoom2']);
-
@likeafox warp containers enable split bases. My experience with them has been that links don't provide enough energy bandwidth to balance energy needs of both halves. Warp containers also allow mineral balancing in a split base.
I think the intent is design warp containers so that they don't dominate the remote mining meta, but rather provide another tradeoff to consider. Fundamentally introducing a new strategy will disrupt the current remote mining meta. If it didn't the new feature would be a flop.The question is what disruption provides the best tradeoff for the most players.
-
@tigga said in Discussion: long-range logistics revamp:
The limitations on civilian uses are a side-effect but not too onorous. I don't think many people are sucking in more than 50 resources/tick sustained in any room other than temple rooms. For temple rooms, there's OPERATE_TERMINAL.
OPERATE_TERMINAL require a Power Creep though, and the temple rooms are not really spending a lot of time at RCL 8. I guess you could have the power creep travel back to a room with an operational Power Spawn once in a while to refresh.. Pretty sure i send more than 50 energy / tick when leveling new rooms as well. None of these are breaking changes, but still this does impact civilian use, where the intention is to limit combat sustainability.
-
@geir1983 said in Discussion: long-range logistics revamp:
@tigga said in Discussion: long-range logistics revamp:
The limitations on civilian uses are a side-effect but not too onorous. I don't think many people are sucking in more than 50 resources/tick sustained in any room other than temple rooms. For temple rooms, there's OPERATE_TERMINAL.
OPERATE_TERMINAL require a Power Creep though, and the temple rooms are not really spending a lot of time at RCL 8. I guess you could have the power creep travel back to a room with an operational Power Spawn once in a while to refresh..
I don't think a walking power creep is too tricky. Temple rooms are usually placed close to other rooms.
Pretty sure i send more than 50 energy / tick when leveling new rooms as well. None of these are breaking changes, but still this does impact civilian use, where the intention is to limit combat sustainability.
Yeah. Batteries help at RCL 7. Might also want operate terminal and a longer walk for new claims. Or just walking creeps as you do pre-terminal.
-
Point is just that the intention is to change how high level combat works (by limiting terminal bandwitdth to reduce the effectiveness of repair), but at the same time works as nerfs to civilian use cases. Both have ways of solving, but is more complex / require the use of power creeps where none were needed before.
-
What if make a power to use on nukes so they will limit terminal throuput for some time. But maybe it is not good to mix this 2 things... Or imlement other two power creep classes so attacking with them will be more effective even vs good repairing. Don't forget that disrupt terminal can be done only in room where power enabled.
-
I think the intent is design warp containers so that they don't dominate the remote mining meta, but rather provide another tradeoff to consider. Fundamentally introducing a new strategy will disrupt the current remote mining meta. If it didn't the new feature would be a flop.The question is what disruption provides the best tradeoff for the most players.
Exactly! Thanks for summing it up in such a clear way.
-
@geir1983 No, it's not about high level combat only. The idea of providing an alternative way to long-distance haul instead of terminals has been debated for ages. We just need to find the correct costs so that it will be a proper CPU/energy/time tradeoff, balanced between terminals and creep haulers. And this is the reason why we created this thread - to figure out this math together with you. We'd appreciate if you not only express your (un)happiness with this feature, but also give us some math that you think would improve it.
-
@artch Thanks for the feedback and your thoughts on why.
-
Some rough numbers from me. There's a few unaccounted for factors, but I think it's mostly fair.
Right now on shard 2 I'm:
- Harvesting 181 remote sources across 108 rooms.
- Spending 406 energy/tick on haulers for this
- Spending 50 CPU/tick on haulers for this
With 3 warp containers per room I think I could replace this with 289 warp containers and 108 "edge bounce" creeps each of which would probably be ~0.2 energy/tick and ~0.1 CPU/tick. Could spent more energy and less CPU, not sure what the optimal ratio is. Removing a container at each source (saving 0.5 energy/tick) and with 1 energy/tick upkeep on warp containers replacing with warp containers would save me ~40 CPU, ~200 energy/tick and a lot of spawn time.
With 2 warp containers per room the simplest implementation would just be to halve my haulers in 2+ source rooms. Lets assume they're all 2 source rooms. That'd mean the replacement would save me ~20 CPU, and ~100 energy/tick. Now I can do better than this simple system in some rooms by placing warp containers between the sources and using one to feed two sources, but that's complicated to reason about.
I think the above shows that 1 energy/tick upkeep won't cut it. I also think that 2 energy/tick upkeep with 3 per room won't cut it either. 2 energy/tick with 2 per room would start to be alright, but still a little cheap IMO (in my case, saves 20 CPU, costs 100 energy, without any optimisation.. that's a pretty good ratio IMO).
I'm starting to reach 2.5->3 energy/tick as my price point. I quite like the idea of two warp containers/room as it makes solutions a lot less obvious, so I think I'd probably go with 3 energy/tick at 2 per room. Maybe a discount in highway rooms (back to 1 energy/tick)?
-
@artch Just curious why are creep haulers not a solution? Perhaps with some buffs to carry, uhm store, size. Increase battery to energy ratio, etc. Increase carry boost percentages.If the warp containers are going to sit somewhere in between, why would anyone use them? They will end up like tunnels barely getting any use due to being so expensive and CPU heavy. Needing to work with resource convoys or as an enemy trying to intercept them would add much more interesting dimensions to a siege then a 'pipe' system you can pummel anywhere at any time.
-
@tun9an0 I also thought about creep convoys with energy to help sieged colony. But depedning on the terrain your path will go through attacking squad or not. In first case it is very hard to keep haulers alive, sure there can be healers and attacking creeps or towers can heal them... But what if attacker just have more creeps and will intercept haulers easily
