indeed ! I also noticed strange behaviour by using en_US.UTF-8 or any *.UTF-8 in LANG. Not really noticed performances problems, but rather misformated files after using perl scripts, or mouse cut&paste, it is very anoying !. So, what is the benefit of using UTF8 in shells ?
Almost none if you never deal with anyone or anything outside the US! However, for those of us that have been struggling with varied character sets for _years_ the benefits outway the disadvantages. My advice (for what it's worth) is learn to live with UTF-8: before long you're going to find it's more important than you think.
The only significant performance problem I have is grepping around a million lines of source code (this may well be covered in the FAQ Eric Wood referenced, so apologies in advance for the any duplication -- I wouldn't want to upset anyone :-)
Since I have a little script that picks out the right source files, I have embedded in it:
... | LC_ALL=C xargs grep ...
which fixes the performance problem. You'll find a similar thing scattered through a lot of the RH init-scripts.
jch
-- Shrike-list mailing list Shrike-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/shrike-list
