23 Feb
2024
23 Feb
'24
3:55 a.m.
On Fri Feb 23 03:55:16 2024 +0000, Giovanni Mascellani wrote:
Would it make sense to print `XFAIL` instead of `PASS` when there are expected failures inside a test case, even if the are no hard failures and therefore the full report is not emitted? It used to be that way before `todo` support was added in the shader runner.
I tried it but it makes it complicated, at least for me, to visually parse red errors if the normal output becomes an amalgamation of "PASS" and "XFAIL" which have different length, and color.
For that reason I propose the idea of handling these cases by keeping them as PASS but displaying the number of XF in parenthesis.