Rocket mechanics feedback

[Coolthulhu]

The new rocketry system is fun, properly endgame, but rather rough at the moment. I won't be talking about obvious bugs, since everyone agrees that they should go.

Good:

• New stuff
• Makes use of many game elements: automation, gas mechanics, refined resources, resources created by phase change.
• Sink for multiple resources: steel, power (through cooling), water, petroleum.
• "Trading" renewable resources for non-renewable ones.
• New design challenges with massive heat sources, CO2 sources, huge vertical rocket silos.

Bad:

• Gantries. They are tedious, but do not have a large impact: high power cost means they have to be retracted one by one or be on many circuits, retraction is painful to automate (it's better to just attach clock sensors and start/stop those manually), they are more likely to get damaged by forgetfulness than genuine design faults. They should automatically retract before rocket starts and then extend once it lands, without the need for explicit automation, and the rocket should not start until they are all retracted.
• Clear progression of engines. Steam, steam+booster, petroleum, hydrogen. Once you have the next step, you discard the previous forever. There is no creativity involved here, no variation, no tradeoffs. Would be more interesting if there were reasons to use old engines even when new ones are available - for example, make engine mass efficiencies more similar, but give endgame engines better mass penalty exponent, so that weak engines can only power small rockets. Or have endgame engines use much more expensive fuel. Make weaker engines extend mission time etc.
• Rocket sensors have the same delay as meteor sensors. But the player can trivially check the rocket's arrival time on starmap screen and manually open the blast doors. This means that rocket sensors only represent convenience for the player. It is an objective error to ever attempt to balance a game (with active pause) by adding a cost to convenience - making a game less convenient for no gain can only ever make it worse.
• Forced order in asteroid selection until at least petroleum engines. Until then, it's all science grinding on few functionally identical rocks. Could easily skip this whole phase and start with petroleum.
• Bio rewards are very underwhelming. Bland stone hatch? I have 20 of those. A mutant hatchzilla might make it worth the effort, but at the moment, bio container is just something that you research when you have nothing else to research.
• Rocket fueling metagame. Getting the best range out of given rocket is a simple optimization problem that is very easy for computer to solve: it has only one variable, which is combined mass of fuel+oxidizer (or steam). Calculating it by hand is a pain, having the game maximize it and floor it to nearest multiple of 10k, then distribute the settings to tanks, would make it much more pleasant.

[Sasza22]

14 minutes ago, Coolthulhu said:

"Get on the rocket" is not a prioritized job like neural vac. The dupe may delay getting on for a long while.

Not sure about that. My guys go in the rocket instantly when assigned. After returning falls off the module and instantly returns to it (even if unassigned during it). Only scenario when they don`t board the rocket is when you disable the operate priority completely.

[Coolthulhu]

Just now, Sasza22 said:

Not sure about that. My guys go in the rocket instantly when assigned. After returning falls off the module and instantly returns to it (even if unassigned during it). Only scenario when they don`t board the rocket is when you disable the operate priority completely.

They go back to rocket often, but I had one dupe ignore the request completely. I think I fixed it by reassigning the dupe. Probably a bug then, I'll edit it out.