Feedback
Document Decimal rounding at precision limits
- Status
- Fixed
- Fixed in
- v26.9.4
- Area
- docs
- Last updated
- 2026-09-28
Upvotes
0 upvotes
Uses your Objo forum account.
Public summary
Clarify that Decimal addition and subtraction can round when their exact mathematical result cannot fit Decimal's 96-bit coefficient and scale. Document how representable spacing grows with magnitude, that arithmetic uses ties-to-even rounding when precision must be reduced, and that positive and negative values have the same precision.