Factorio Forums

When using very large grid sizes such as 2147483647=2^31-1, placing blueprints at y=-1 or -2 causes jitter and places the blueprints incorrectly. The same does not happen for positive y-values or y-values that are further away from 0. Slightly smaller grid sizes such as 2147483645 cause the gaps to become larger. This looks like some kind of internal integer overflow.



In case you consider this not worth fixing, it would make sense to constrain the grid size to e.g. 2^30.

On a further note, the reason to use this large grid size is to make a blueprint tileable only in x-direction. Hence I request the feature to disable tileability in one direction more properly. Concretely, I propose to allow a grid size of 0, with the meaning that a blueprint is not tileable in that direction. In fact, this is semantically correct, since the only multiple of 0 is 0 itself; so allowing placement only at the integer multiples of the grid size means for a grid size of 0 exactly non-tileability in that direction.
Should this be posted as separate suggestion?

Best,
PhiPro

This doesn't happen for me, even with a (very laggy) 1569325055 size grid

the grid size need to be smaller then 2.000.000 in X and Y, any larger value makes no sence - because it is the max. Gamefield size

Wouldn't like 2000 be enough for the intended behavior?

If you intentionally cause an overflow... well, that's an overflow alright. Seems more like feature request than a bug report but I guess it is a tiny bug.

Thanks for the report however I'm going to just say "don't do that". Since it's not crashing the game and it's just more work for us to validate the values everywhere they can be changed for minimal/no gains.

Regarding my feature request for allowing a grid size of 0, should I post that again in the suggestions subforum?

Regarding the bug itself, values >= 2147483648 and <= 0 are already disallowed. Would it make sense to further restrict the upper bound to 2^30 or something for the time being?