[H-GEN] UTF-8 vs ISO-8859-1

Ted Percival ted at midg3t.net
Thu Jun 22 02:55:51 EDT 2006


Troy Piggins wrote:
> On most linux distros I've installed, if I enter my location as
> Brisbane, Australia during the install process it automatically
> seems to set locale to UTF-8.
> 
> $ locale
> LANG=en_AU.UTF-8
<snip>
> 
> I am not sure of /all/ of the implications of this, I think most of
> the applications I use from the terminal (mutt, slrn, elinks, 
> midnight-commander etc) all seemed to display characters like
> umlats (?spelling) correctly/better if locale was ISO-8859-1.
> 
> Is this correct?  Or was it a TERM type setting issue?
> 
> What /should/ I be using?

I recommend UTF-8 because it allows you to display a much greater range
of characters. Most programs these days support UTF-8 and often use it
as a default.

As for better display in terminals, installing the 'console-data' Debian
package adds fonts that display UTF-8 reasonably. They seem to be
missing a lot of non-latin characters, but at least missing characters
are shown as a single box rather than 2-4 "junk" characters, throwing
everything out of alignment.

I've heard a rumour that xterm still doesn't support UTF-8. The
console-data package will most likely only help if you're using a tty,
though Gnome's Terminal definitely supports UTF-8 properly, and I'd bet
on Konsole also working.

I'd guess Ubuntu's package is probably also called console-data, not
sure about other distros.

-Ted




More information about the General mailing list