Skip to content

[ascii import] when reading data, assume it's UTF8 encoded and not Latin1.

Alexander Semke requested to merge work/fix_utf8_import into master

UTF8 is the commonly used encoding nowadays. We used it in the past and switched to Latin1 because of performance optimizations which was clearly a very bad decision. We go back to UTF8 now. This won't solve the problems with UTF16 as reported in #541 (closed) and probably other encodings that are not subset of UTF8 but this needs to be addressed differently if we want to support them properly in future.

BUG: 470338 FIXED-IN: 2.10.1

Merge request reports