FreeOrion

So i think i have now a design for the third iteration:
* just use the global LIMIT for input and for output
* the empire contributes overproduction up to the LIMIT to the stockpile. Excess overproduction is wasted.
* for calculating this limit, all planets contribute differently based on the _STOCKPILE species trait
* e.g. a GREAT_STOCKPILE species adds up to 0.75 PP to the limit per population
* only planets with industry or supply focus contribute
* in order to keep the slow rise and fall of meters, the stockpile limit gets multiplied by industry efficiency (local production divided by target-production)

For balancing
* nerf the technologies
* adding the GREAT_STOCKPILE species SP_SLY
* making the SP_LAENFA GOOD_STOCKPILE

_________________
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Sounds like there should be a "Stockpile" meter for planets (and perhaps ships), which determines how much they increase their empire's stockpile in/out transfer limit, and possibly a "Stockpiling" planet focus. I wouldn't call it "Supply" focus, as there is already a supply meter which determines how far supply networks extend. Having it be a planet/ship meter means lots of other content can interact with it if desired, like having a planet or ship in a black hole system be good for boosting the empire stockpile transfer limit.

Yes, i do agree. If we keep this design, we probably should give this own stockpileLimit and targetStockpileLimit meters instead of making it depending on the industry and targetIndustry meters.

For playtesting though this isnt really necessary. The code in the pull request is ready for collecting feedback I think.
Had a go at it with Laenfa until midgame and the stockpile helped. And I actually needed to research some of the techs for the extra boost. So balance is probably not too badly off. UI needs improvement and I have some concerns about micromanagement, but i think these are resolvable.
https://github.com/freeorion/freeorion/pull/1928.

I added the necessary info for this round to the _LIMIT entry in the encyclopedia, so should be helpful.

_________________
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

_________________
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Best I think would be actually to see how many PP get wasted. So instead of or complementary to the waste indicator the number written in red.

With the ratio it was like: ok, i have a ratio of 20% and the top row reads that I transfer 20PP to the stockpile. So i know i waste 80PP.


Problem:
If you want to use stockpile PP, you first have to put them there. At the moment you have to juggle your production so that you get excess to fill the stockpile but not much waste. This involves pausing/unpausing production items.

Some options I thought of:
* somehow explicitely specify how much you want to put into the stockpile
** add something like a pseudo-ship to the production queue (base cost 1PP, 1 turn; you can specify how many you want and how often using the controls for ship build items); this shouldnt create a ship though;
** it would be good if you see if you specify more than the input limit; best would be if it would be figured into the next turn stats
** Needs some fiddling with the build queue, good control of amounts

* somehow specify priorities of building
** e.g. items with low priority do not get PP until the stockpile input limit is full.
** For ways to specify priorities see UI Separators; complicated situation but easy to use UI
** Probably UI is hard to implement and needs some fiddling with the production queue

* remove the input limit
** no juggling necessary, dead easy to implement, KISS

_________________
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

I think the waste indicator being present or not is indication enough. The waste is a secondary statistic, not as important as the amounts and outputs of resources. There's only so much space in the GUI to show more numbers.


I'm inclined to try the no-input-limit approach first, and if that doesn't work, then a "stockpile" project on the production queue with adjustable cost / amount per turn. If there's no input limit, then I suspect an input efficiency will again be needed, though...

The thing is - you dont want to check every turn the top bar for waste. Better explanation of why I want this and how it worked with the input ratio: Imperial stockpile diff is good to see what you are wasting

As stockpile is based on excess, in many of the stockpile designs there is something like "accepted waste". So for those designs the problem of waste management is much more pronounced.
* Discussion of UI alternatives with dilvish which stopped because we removed the stockpile total limit: Imperial Stockpile -Indicate how much PP is wasted (%-based)



There are also approaches of high input limits or instead reintroducing a stockpile total limit; those would not solve the indicator problem, but make the problem occur less often.

We could also disconnect the stockpile from excess. I.e. stockpile input is independent from normal production, you dont transfer parts of excess, but directly "produce" stockpile PP. This would make the stockpile also waste-independent. I think that would also be very KISS.
This "stockpile-production" intrigues me the more I think of it. It probably redefines what those stockpile points are, they become more of a second "premium" type of production points than a way to store PP. We know very well what such production points are and how buildings and effects work on that, so balancing probably gets easier. On the other hand maybe its almost like production points just with a empire-wide supply network, so maybe the "stockpile-production" is more boring because its similar?

_________________
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

I write this here (not in the feedback thread), because I dont have really tried it.
There could be a micromanagement "exploit" of the basic stockpile tech via focus switching.

Lets say if you switch your planets from production to stockpile for a turn and back to production. PP goes down by 2PP,
turn 0: 5 PP, 0 Stockpile Limit
turn 1: 3 PP, 1 Stockpile Limit
turn 2: 4 PP, 0 Stockpile Limit
turn 3: 5 PP, 0 Stockpile Limit (same as turn 0)
So you trade 3PP for one Stockpile limit. So 4PP become one stockpile PP. Thats an implicit stockpile ratio of 25%.
(If you switch for two turn without hitting limits, its trading 12PP for 4 stockpile limit. So same ratio)
For the force energy structures tech (3 points up and 5 points down per turn) the ratio is more like 33% (1 turn switch: 7PP vs 3 limit; 2 turn switch: 22 PP vs 10 limit).
Also I think sometimes meters just seem to drop 1 point instead of the expected 2 (in that case ratio would be a perfect 100%).

But that is not really much different from production vs research.
Egassem are able to convert their 300% industry into a 75% research this way (until hitting limits, that is).

More importantly: If you are in need of stockpile limit, you switch all your e.g. 50 planets for e.g. 2 turns to stockpile, extract 200 PP and switch back to production. Not sure if that can be counted as exploit or an acceptable strategy. You just need the first tech to do this.

The basic problem is that the stockpile limit is only useful if you input/output points. Else its just wasting time. It makes sense to put planets constantly to stockpile if you use it at least regularly. If you have planets dedicated to stockpile it might make sense to research later tech, else not.

edit: fix details about force-energy-structures

_________________
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Why is this a problem...?

Theoretically it could be a problem if it leads to incident-driven micromanagement. Ohh shit, i need some stockpile. I switch all my planets to stockpile, besides some. Ten turns later, oh shit i am wasting resources on the stockpile. Now i need to decide which planets to switch focus to what.
Im not sure if it is a problem, as I am mostly too lazy for micromanagement, so i wont even try for testing.

_________________
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Is is more problematic / micromanagement prone than other existing requirements to occasionally change focus?

I think it is a bit, and that gets pronounced for non Laenfa/Sly empires that try to use IS, the change of focus will have to affect more planets and probably more often.

With industry/research focus I seldom have to change focus of a given planet more than once in a whole game.
I only use defence focus for the occasional choke-point vs a stronger fleet, not in every game.
I never use supply focus (I prefer orbital elevator the rare occasions I need it).
The special growth focus only once per special I found.
All these changes require, to select which planet to change and which not, easy reasoning and all information is available in the objects window.

On the other hand, to select which planets change to stockpile focus you often need to see in which supply groups they are and what are the waste ratios. This requires more time with the GUI. And the changes my be required several times during a game, 10+. So in the end I think it really takes time from you.

So I would be happy to find a way that makes stockpile focus unnecessary and remove it from game.

A distributed empire without a stockpiling species is not intended to be a worthwhile enterprise. So I guess the necessity to keep an eye on stockpiling is not so high. I hope somebody gives playtesting feedback for this.

_________________
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

So design for fourth iteration:

• no input limits and stockpile use enabled by default (geoff already implemented as rule; already in git master)
• sly growth nerf; size 4 adjustment for gas giants (Oberlus?ophiuchus branch & cherry-picked AIDependenciees change from dilvish )
• flux bubble nerf (ophiuchus branch; remove unlock, more expensive/hitpoints)
• sly AI testing (already a branch by dilvish)
• local refueling for hidden ships or special refueling for sly (ophiuchus branch)
• gas giant + space elevator nerf (ophiuchus branch; only +3 supply instead +4; sly minus one supply)
• populated gas giant nerf (ophiuchus; only +5 for generator; AI ???)
• no terraforming for sly

related topics

• a way to define stockpile input amount (ophiuchus; ATM probably as supply pods in a ship; long term as a "project" build item)
• gaseous metabolism (probably only if more distinct)

edit: update

_________________
Any code or patches in anything posted here is released under the CC and GPL licences in use for the FO project.

Consider me just a dumb playtester. I'm still digging through tutorials to be able to fetch and try changes/branches done by others (right now it is crossing my mind to just type in Dilvish changes on my local repo to try them already!).
So feel free to add that simple change (viewtopic.php?f=28&t=10859#p91346) I'm testing:
https://github.com/freeorion/freeorion/ ... t.cpp#L205
There, change 6 by 4.

And I don't know, but I assume that including also Dilvish's PR is a good idea? Those are few changes that won't broke anything I think.