Hi,
Holger Wansing <hwansing@mailbox.org> (2024-12-26):
> Hmm, according to https://d-i.debian.org/doc/i18n-guide/ch01s04.html#sublevels
> sublevel3 is for "expert strings (some low priority features such as
> RAID, encrypted partitions, etc.)" - aka not for default install. So
> sl3 might not be that wrong.
Thanks for checking…
> However, sublevel4 is "specific to less-popular arches (such as
> powerpc, mips or sparc) or used in experimental features".
> That leads to the decision, which arch is popular and which is not.
> But since arm64 is somewhat new, we could count that to the "not popular"
> ones currently, for sure. Thus moving the strings to sl4...
I think I'd just call that an “experimental feature”, without diving
into arch popularity considerations. :)
> But now we are at the point, where I ask myself, why that makes a
> difference?
>
> Are all strings from sl4 and sl5 generally not considered, when
> calculating the translation-status value?
>
> That's at least not documented in the above quoted doc...
Let's look at calc-release-status:
# Format of that file is '<lang>: <sublevel> <status>', where:
# - <sublevel> is the first sublevel that is not fully translated or,
# if the language /is/ fully translated, the highest sublevel;
# - <status> is: F (full), M (mostly: >90%), P (partial: 50-90%),
# L (limited: <50%).
# Note that sublevels 1 and 2 are taken together.
In the case of “fr”, that moved from “3 M” to “5 F” after the update.
Then, localechooser does this:
# Determine which template to display to warn for incomplete translations
# and fill in the variable contents
warning_template() {
local RET status template tbase twarn tabort
status=$1
tbase=localechooser/translation
case $status in
0) twarn=incomplete; tabort=abort ;;
1) twarn=normal-ok; tabort=abort ;;
2) twarn=partial; tabort=maybe-abort ;;
3) twarn=mostly-ok ;;
4) twarn=exceptions ;;
esac
if [ $status -le 2 ]; then
template=$tbase/warn-severe
db_metaget $tbase/text/$tabort description
db_subst $template TXT-ABORT "$RET"
else
template=$tbase/warn-light
fi
# Languages that have fallbacks may have special templates
if [ "$twarn" != exceptions ] && \
expr $LANGUAGELIST : ".*:" >/dev/null && \
db_metaget $tbase/text/warn_$twarn/$LANGUAGE description; then
:
else
db_metaget $tbase/text/warn_$twarn description
fi
db_subst $template TXT-WARN "$RET"
echo $template
}
so I'm not sure how much of a difference it would make to have those
templates in sl4.
> I have a walk-through for this task, that can be done without problem.
> Will do, when alpha1 is out.
> (And: No, there's no extra burden or extra work for translators here.)
Perfect, thanks!
> I should document how to do this in above doc BTW...
That'd be awesome. :)
Cheers,
--
Cyril Brulebois (kibi@debian.org) <https://debamax.com/>
D-I release manager -- Release team member -- Freelance Consultant
Attachment:
signature.asc
Description: PGP signature