Re: POSIX Regular Expression question

In article <200509051619.29210.peter_e@gmx.net>,
Peter Eisentraut <peter_e@gmx.net> writes:

> Presumably,

>     colname ~ '^[A-Za-z]*$'

> If you want to be independent of locale issues, then you'd have to enumerate 
> all the letters instead of using a range specification.

How so?  I thought character ranges are more an encoding than a locale
issue.  I dunno the details of the supported encodings, but is there
any encoding where 'a-z' includes more or less than 26 letters?
Fortunately we don't support EBCDIC ;-)