> The objection last time was that, according to the SQL standard, LIKE
> patterns should use the locale-specific collation order for the fixed
> parts (instead of character-by-character comparisons),
Yes, I think that LIKE definitely needs to be character-by-character
and I can think of no example in the german language that would make
general sense (like ss and single byte sharp s (ß) would definitely be
nonsense if handled as equal for LIKE).
Andreas