mmmPI wrote: ↑
Mon Nov 13, 2023 5:47 pm
Yes , i tend to think it would be a fun puzzleGive us access to raw value like stack size, number of slot, weight of one item, and then we could do the math with combinator, it's not super advanced maths, ideally when the research is done it auto-update the value, but there is room for different level of automation/complexity.
Give us access to raw value like stack size, number of slot, weight of one item, and then we could do the math with combinator, it's not super advanced maths, ideally when the research is done it auto-update the value, but there is room for different level of automation/complexity.While I and you might agree on that to varying degrees, I don't think that's true of the general player base. And even if I could math it out I'd rather be told the exact number(s) if that's an option.
mmmPI wrote: ↑
Mon Nov 13, 2023 5:47 pm
if you try something like division to know how many rocket you can load or how many rocket you need for certain amount of material to be delivered, if you have enough material to load say 1.5 rockets it's part of the challenge to round up or down depending on your build i would say. But maybe it is not a fun challenge for other players and more like a source of endless pain
You can actually work that out. You get the division (which floors to 1) and then you can do the Modulus which if it's >0 then you get a partially full rocket, you can even find out the % by doing the math as Modulus*100 then /Rocket Capacity. Order matters when it truncates, and not everyone will be familar with that concept.
mmmPI wrote: ↑
Mon Nov 13, 2023 5:47 pm
Fluids are not integer in train wagon you can have 6.7 fluid, the "read train content" will truncate to 6 for the circuit, unless it's between 0 and 1 in case it will round up so that 0.2 fluid don't show as 0 and mess logic.
I hadn't noticed this... but it does show that Factorio works with decimals it's just circuits that do not.
mmmPI wrote: ↑
Mon Nov 13, 2023 5:47 pm
"at worse" 0.5 or 0.2 but not 0.33333333. If the weight is 1.3333333 kg, it will yield imprecision anyway if you have 1 plate in the rocket and try to know anything about the load level ( or how many you can fit) , same as you mention with 1 item that could represent 0.5% load, i think it could be made manageable but i may be wrong.The FFF says :
Output the rocket capacity of the input item (useful for Space Age logistics).
It could be because weights were made so that you can always put a integer number of 1 item in a rocket and fill it entirely. I understood this as you send "steel plate" and it tells you how many steel plate you can put into 1 rocket. In my mind you can already "math" how much weight a steel plate, if you consider the rocket can carry 1 ton it's a division which will yield a number that may be truncated, but one could then build to "weight" one stack of steel plate, in order to know how many iron plate you can put if you only fill the rocket half with steel plate. I was under the impression that this was how i'm supposed to use the "rocket capacity" function of the selector. Since by default it only tell you information on a rocket filled with 1 item to the max capacity.
unless it's /2^n then it'll be a continued fraction in binary with rounding errors eventually 0.2 isn't any more precise than 1/3 in that regard. The error is just so small that it's practically ignorable (and most implementations have more precision in binary and round off for decimal since that's what most people use).
Your suggestion of limiting weight to 1 Decimal place kg means that you can only have 1, 2, 4, 5, 8, 10, 16, 20, 25, 40, 50, 80, 100, 125, 200, 250, 400, 500, 625, 1000, 1250, 2000, 2500, 5000, and 10000 items in a rocket. It doesn't make sense to me that you couldn't have 750 items which is the 1.3333... item weight I mentioned, let alone a number like 70, which would be 14.2857142857... or any of the other 9975 possible whole number of items in a rocket (in this range) it skips 3, 6, 7, 9, 30, 60, 70, 90, 300, 600, 700, 800, 900, 3000, 4000, 6000, 7000, 8000 and 9000 which I'd consider round numbers of items relative to 16, 25, 125, 250, 625, 1250 and 2500.
You can calculate the mass but there's no guarantee it's gonna be a nice number, I'd multiply it by a million so I got some precision out of it in many cases but there's a limit cause 32bit integer.
We agree there will always be some level of imprecision in using the weight (and other stats... is gravity gonna be nice? Earth is 9.81... or Volume? Assemblers are 3x3xXm in volume is it gonna be 1000m^3 for 50 Assemblers giving us an estimated height for the assembler of 2.22222m) to manually determine the rocket cargo capacity with circuit networks, a fun challenge for people of engineering, math and programming backgrounds but not generally fun for a general audience, while internally the game can actually just tell you the right answer because the exact same logic that's used to come to that answer and determine the actual limit of the rocket can be accessed by the combinator logic. So why not let it just tell you the number?
And if we do actually define all these physical properties.... What does that mean for inventories? we can handwave away carrying a lot of weight... the space suit could have some level of assist for lifting heavy objects, but if the player can magically compress all this mass and volume into their pockets why aren't they using that technology for the space launches XD
mmmPI wrote: ↑
Mon Nov 13, 2023 5:47 pm
This is beyond my ability to suggest a fix ! The rocket was shown having 1 metric ton of payload capacity. If you use other units, it WILL create imprecision considering the circuit truncate decimals. Even in real life it happened
The rocket itself would need to be of a different payload capacity and the weights of all item reworked. I think it's not going to happen.My solution (and point) was "we don't need to worry about it" it doesn't matter if the Rocket has 345289 flugalcadences as the capacity and the weight of the item is 645038 prakens, the same cargo capacity limit can be output because it is a dimensionless number.
dstroth wrote: ↑
Tue Nov 14, 2023 2:37 am
These are good points! The engineer in me always wants to reduce solutions down to their simplest, most general building blocks. But, I agree that in the contex of Space Age, it probably makes more sense to just report the quantity that most people will want directly - i.e. how many items will fit into a rocket, given a potentially large set of external constraints like mass/volume/gravity/etc.
*nod nod* Simplicity is better, it's just how you interpret 'simplicity' is the issue, for me it feels like simplicity is being told the answer not told all the most basic information possible
dstroth wrote: ↑
Tue Nov 14, 2023 2:37 am
I trust that the developers know what they're doing - and if they think "rocket capacity" is the best property to expose, I can support that.
Me too in most cases 
dstroth wrote: ↑
Tue Nov 14, 2023 2:37 am
Agreed, I wouldn't mind seeing more metadata/stats/etc exposed via a special combinator.
If the player can know it and make decisions based upon data then it'd be nice if the Circuit network could have access to that same data to also make decisions. After all that seems to be the whole point of Circuit networks for me. It can do a lot of amazing and powerful stuff too but that's a side effect of why it exists in the game.
dstroth wrote: ↑
Tue Nov 14, 2023 2:37 am
Hah - as an American engineer that much prefers working in metric, I sincerely hope Factorio sticks to metric everywhere!
I prefer metric too but some people will complain "I don't understand KG" and "what is a meter anyway" if you give them the opportunity
dstroth wrote: ↑
Tue Nov 14, 2023 2:37 am
A separate mode makes more sense to me, though; overloading the meaning of Everything could be confusing, and certainly isn't very discoverable.
Agreed the number of niche combinator things I've found out by reading this thread is mind-bending. Adding more hidden features maybe isn't useful.
dstroth wrote: ↑
Tue Nov 14, 2023 2:37 am
"Each" would be even better! Maybe even just a separate mode/checkbox for "set stack size from filters", so it is more discoverable.
