[Suggestion] PLEASE make these basic yet critical improvements (v1.01)

[tanelorn]

If the requests are queued, rather than filled/denied immediately, then this not work to measure deficit. That structure could allow deficit to be partially measured by queue length and the time a request has spent in the queue.

Measuring the number of power-hungry processes going on could help, but with custom modules thrown in, it becomes endlessly complicated.

Queueing wasn't on my mind for this. Just throttling the max amount of energy that any one mine can dispense at one time.

[GoodMorning]

I used the word "are" to refer to the current state.

Throttling the mine rte may involve extra work to ensure that ship priority for packet requests is even.

[planetfall]

Throttling the mine rte may involve extra work to ensure that ship priority for packet requests is even.

Why would it? Packets are already evenly distributed when there is less than 30/s being produced. My understanding is there's a packet queue, but each thing requiring packets can only be in the queue once at a time. (That's how I implemented the EE, and it seems to round-robin similar to CW3/PF packets.)

[GoodMorning]

That will likely be a minor point, but suppose two large ships. One is in range of two mines, and the other in range of one of these. Either will consume the entire throttled output. Now suppose that the one in range of two mines is asking for power first, and is supplied from the mine which both ships are in range of. The other ship loses power.

In essence, you have to allocate the mines supply capacity individually, taking note of any and all other ships in overlapping range. EE, as far as I could tell, does not have a chokepoint. CNs can supply power as the current mines do.

Round-robining from each mine would work, to a degree, but then packet distance comes into play. Balance that with the 'earthquake' consumption graph... Possibly it could partially be brought to heel if ships were preferentially resupplied from the mine that most recently provided power, meaning that an underworked mine will 'hold onto' ships in range while an overworked mine will 'lose' them.

Changing to this could also require significant rebalancing of every map to date, as ship construction times and patterns go out the window.

Returning to when I first mentioned a queue, if the length of time a request has been queued for was measured, and summed over the fleet, then we could scale that to a reasonable measure of drain, which will spike when deficit conditions are reached, while remaining meaningful.

At this rate, I'll end up making a toy system to bounce ideas off. But what we have now seems to work, it just needs a measurement shown.

We've pushed this thread away from the original point... Oh well.