You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Doing something like ("<I6"):pack(2 ^ 48 - 1) packs into the wrong output because the number gets cast into an int beforehand. Meanwhile, Cobalt casts their inputs into a 64-bit Java long, saturating if the value is too large, so it has the same issue for values larger than 2⁶⁴.
This issue isn't very frequent because most numbers are too small to come up into overflow territory, but since doubles can hold integers all the way over to 2⁵³, it's reasonable to expect to be able to pack these numbers too. Cobalt has some other peculiar choices of rounding, but I think implementing all of them would be too much to ask regardless.
The text was updated successfully, but these errors were encountered:
Doing something like
("<I6"):pack(2 ^ 48 - 1)
packs into the wrong output because the number gets cast into an int beforehand. Meanwhile, Cobalt casts their inputs into a 64-bit Java long, saturating if the value is too large, so it has the same issue for values larger than 2⁶⁴.This issue isn't very frequent because most numbers are too small to come up into overflow territory, but since doubles can hold integers all the way over to 2⁵³, it's reasonable to expect to be able to pack these numbers too. Cobalt has some other peculiar choices of rounding, but I think implementing all of them would be too much to ask regardless.
The text was updated successfully, but these errors were encountered: