Bug#207455: acknowledged by developer (Re: Bug#207455: packages.debian.org: HTML-encodes multi-byte characters as single bytes)
On Sat, Aug 30, 2003 at 03:22:30PM +0200, Josip Rodin wrote:
> On Sat, Aug 30, 2003 at 06:57:51AM +0200, Martin Godisch wrote:
> > > > UTF-8 multi-byte characters in package descriptions appear single-
> > > > character encoded on packages.d.o.
> > >
> > > Uh, yeah, because, like, the encoding of the description is not regulated?
> > >
> > > File bugs on the broken packages or something.
> >
> > Why is a package considered broken when its description is UTF-8 encoded?
>
> Because, like, there's no rule allowing it to use UTF-8?
If you're going to use top-bit-set characters at all, it's the only sane
encoding. We just need dpkg and apt to recode on output.
(It'll be a lot easier to get UTF-8 accepted everywhere if tools like
packages.debian.org start supporting it.)
> > Policy requires changelogs to be UTF-8 encoded
>
> Changelogs are not descriptions. Don't try to bend our arm with stupid red
> herrings like that.
He went on to point out that the Maintainer: in changelog and control
needs to match, and control contains the description so it makes sense
for it to be in the same encoding. Encodings tend to propagate.
(BTW, ITYM "twist our arms" ...)
--
Colin Watson [cjwatson@flatline.org.uk]
Reply to: