--- Comment #3 from johnlumley <[hidden email]> ---
normalize-unicode-01 problems are caused by switching to a serialization
assertion (as indicated by the TODO) which fails. This is because in the
reference tests decimal numeric character entities (e.g. ̈) are used, but
(at least in our implementation) the results use hex numerics (e.g. ̈)
and hence a character-by-character serialization check fails.
How (or whether) this can be controlled needs me to investigate. Temporarily
these tests have been reverted to use assert-xml
You are receiving this mail because:
You are the QA Contact for the bug.