Rocket capacity is too inaccurate for some items

[gyorokpeter]

Example is the uranium ammo, which has a stack size of 100 and a rocket capacity of 25. This means the rocket capacity is 0.25 stacks, yet in the tooltip it shows up as 0.3 stacks. I guess too aggressive rounding is to blame for this. Some limits do need to be in place in case of mods adding items with a ridiculous discrepancy between their stack size and rocket capacity, but the inaccuracy of 0.3 vs 0.25 is just too disturbing.

[Koub]

[Koub] Movint this to Ideas and Suggestions : Pretty sure it's not a bug (as in working as intended), but there is room for improvement.

[Alwego]

i agree. Rocket capacity is too low. I can have a train in pocket, but i can send to space only 25 ammo. It's ridiculous.

[Muche]

Ref 119200 Rocket capacity stack counter rounding, 122554 Rail support bad math, 122889 Incorrect stack ratio in tooltip.

[Muche]

Now that v2.0.25 added recycling recipes to Factoriopedia, insufficient decimal places show up more prominently -

Tooltip-RecyclingRecipes.jpg (49.61 KiB) Viewed 436 times

[IsaacOscar]

Now that v2.0.25 added recycling recipes to Factoriopedia, insufficient decimal places show up more prominently -
Tooltip-RecyclingRecipes.jpg

That's because there's only a 13% chance of getting a copper plate:

12-13-2024, 13-11-47.png (86.53 KiB) Viewed 423 times

(The unrounded chance is exactly 12.5%, and the time 0.03125s, so the rounding here isn't bad).

[eugenekay]

My 2c Opinion: Changing the string display functions to consider Scientific-style “Significant Figures” would both improve this situation…. And make it a lot worse

Consider:
- “1/4” or “0.25” are rounded to “0.3” at present, with SigFigs=2 this becomes “0.25”
- “2” at SigFigs=2 becomes “2.0”, which could maybe be rounded….
- “139/s” (random figure from a Chemical Plant in my factory, the amount of Thruster Oxidizer produced) at SigFigs=2 becomes “140/s”; which is both factually misleading and insufficiently precise.

Using a higher value would lead to nonsense like “2.000 items/second”; which is less confusing but arguably worse-looking than the current situation.

I think that the state as-is is “Good Enough”, for a Video Game. If this was industrial or medical control software it would be a different matter.

[IsaacOscar]

My 2c Opinion: Changing the string display functions to consider Scientific-style “Significant Figures” would both improve this situation…. And make it a lot worse

Consider:
- “1/4” or “0.25” are rounded to “0.3” at present, with SigFigs=2 this becomes “0.25”
- “2” at SigFigs=2 becomes “2.0”, which could maybe be rounded….
- “139/s” (random figure from a Chemical Plant in my factory, the amount of Thruster Oxidizer produced) at SigFigs=2 becomes “140/s”; which is both factually misleading and insufficiently precise.

Using a higher value would lead to nonsense like “2.000 items/second”; which is less confusing but arguably worse-looking than the current situation.

I think that the state as-is is “Good Enough”, for a Video Game. If this was industrial or medical control software it would be a different matter.

I think the appropriate formatting depends on context, having 0.25 or 0.3 of a fluid doesn't make much difference, but 0.28 vs 0.3 stacks does (my 2c is that the tooltip should show total items/rocket & kg/item, and not bother with the number of stacks thing as it's often not an integer).

[IsaacOscar]

Hmm, looking at the '999' thing, it's actually at least 999.9kg, as you can't put 300 processing units and a blueprint in a rocket, despite a blueprint only weighing 0.1 kg. So it definitely should be showing 1,000 kg (as 999.99kg or something similar would be unnecessarily precise).

[dmikalova]

Why not do 2 sig figs past the decimal point, and no trailing zeroes? While not correct for scientific literature, imo it best represents the data for a game. Examples:

2, not 2.00
0.25, not 0.3
139.78, not 140
0.0034, not 0