Bug 204836
| Summary: | Fix CalcExpressionOperation::evaluate() to use doubles | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Simon Fraser (smfr) <simon.fraser> |
| Component: | CSS | Assignee: | Simon Fraser (smfr) <simon.fraser> |
| Status: | RESOLVED WONTFIX | ||
| Severity: | Normal | CC: | koivisto, sam, simon.fraser |
| Priority: | P2 | ||
| Version: | Safari Technology Preview | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
Simon Fraser (smfr)
We should use double-precision math.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Simon Fraser (smfr)
Actually Length stores a float, so we might as well do these calculations in floats. This is a bit weird, because math done at parsing time will happen in doubles, but math done at used value time will be done in floats.
Sam Weinig
(In reply to Simon Fraser (smfr) from comment #1)
> Actually Length stores a float, so we might as well do these calculations in
> floats. This is a bit weird, because math done at parsing time will happen
> in doubles, but math done at used value time will be done in floats.
Does the spec specify which should be used?
Simon Fraser (smfr)
No but there's talk of doing so: https://twitter.com/starsandrobots/status/1199757377286754309