On Jun 08, Roger Leigh <rleigh@whinlatter.ukfsn.org> wrote: > >> > Wrong. The problem is packages which need to interact with text files, > >> > mail and usenet messages generated by broken software, and for which > >> > assuming UTF-8 would be totally wrong. > >> This is completely orthogonal to making UTF-8 the default locale > >> codeset. > > No, it's not because most applications do not allow setting a different > > "default charset". > > Please could you re-read what I wrote? What you are saying does not > follow from that. > > By default locale charset, I'm referring to the defaults in the > locales package, which are used to generate /etc/locale.gen. If you I'm not. I'm referring to the charset which is used by applications to interpret unlabeled text streams. > GNU/Linux has been slowly moving to UCS since the late '90s. We are > now well past the point where it's mostly usable and ready for proper > use. Debian is well behind the times here. UTF-8 support is good. A default may be good for some locales and wrong for others. > As something to ponder: with all current gcc's in Debian, UTF-8 and > UCS-4 are used as the internal narrow and wide string literal encoding > in all binaries, independent of the C source encoding. See I can't see why this would be relevant. I'm not arguing about the merits of Unicode. > > Unsurprisingly, looks you live in a country where anything else than > > US-ASCII was rarely used in the past. > ISO-8859-1 actually. But this is not really topical. It is, it explains why you do not understand the issue. -- ciao, Marco
Attachment:
signature.asc
Description: Digital signature