Why is "LANG=en_US.UTF-8" the default in Fedora

Alan Cox alan at redhat.com
Thu May 20 21:22:38 UTC 2004


On Thu, May 20, 2004 at 08:51:02AM -0600, Guy Fraser wrote:
> >Any time I use sed or grep on my script I make sure to set LANG=c first.
> >US.UTF-8 slows grep searches *way* down.  It was so slow that I thought I
> >had hardware problems - nope, just unicode slowing down my scripts.
> >
> >-eric wood
> >
> It also makes them case insensitive. #^%$$# ^&%*& !!!

Common confusion - it doesn't do this but the order of sorting is
dependant upon the language rather than numeric sort order of symbols.
This is correct because these standards are in most cases hundreds of 
years older than Posix C locale.

It does mean that in many languages [A-Z] is AaBbCcDd...Z, but you
should be using the proper posix upper/lower set for such searches.





More information about the fedora-test-list mailing list