Re: How much utf-8 do we accept in control files?
Siggy Brentrup <bsb@debian.org> schrieb/wrote:
> Let me try to reformulate: if we allow anything not representable in
> ascii in control fields, I see no valid argument to prohibit any
> alphabet (set of glyphs used to display names etc. in any language)
Control files are written in the English language. This already limits
the frequency of characters from non-Latin scripts.
It is not a good idea to impose a limit on the characters that are
allowed in control files. They can be useful, especially when bugs
are related to these characters.
However, changelog entries should be understandable when only
ASCII is displayed:
Wrong: Fixed display of "€" character. (Closes: #12345)
Right: Fixed display of Euro Currency Sign (U+20AC, "€"). (Closes: #12345)
> Without a complete set of fonts installed there is no way to even
> distinguish maintainer names written in their own language.
Even if you do have the fonts installed, most people don't
recognize the characters. For example, can you tell "陳阿瓦" and
"張阿里" apart if you don't see them next to each other[1]?
Names will have to be written in Latin (not ASCII, which lacks
diacritic marks necessary for many names from both Latin and
non-Latin scripts), possibly followed by the original name in
parenthesises.
Claus
[1] NB: These "names" were made up from samples. They may or may
not be proper Chinese names.
--
Reply to: