On Mon, Sep 3, 2012 at 3:15 PM, Pádraig Brady <p...@draigbrady.com> wrote: > On 09/03/2012 04:58 AM, Nguyen Thai Ngoc Duy wrote: >> On Mon, Sep 3, 2012 at 6:36 AM, Pádraig Brady <p...@draigbrady.com> wrote: >>> Thanks for doing that. >>> I really like this, but wouldn't it invalidate all the existing >>> translations? >>> It seems there would be a lot of work in transitioning coreutils usage() >>> functions >>> to the scheme, but even more work munging existing translations? >> >> Existing translations should remain (though marked fuzzy). Translators >> just need to go through remove redundant parts. > > Yep it would be a fairly manual process, > and one that would need to be done en masse, > so that many translations weren't invalidated > on a new release. > > When I changed the df headings code to > abstract the translations in a similar way, > I updated all translations which Benno accepted > as a single tarball udate. It took a couple > of hours to do this though, so updating usage() > for all commands could be a full day's work?
I haven't thought of that. coreutils supports 40 languages. Yeah, maybe 1 day. > >> The thing is translations should be updated anyway because of >> misalignment (I only checked ja and vi). We could use this opportunity >> to avoid potential work for translators later (e.g. a new long option >> is added and the whole table needs to be realigned, which currently >> changes every help string, or a misalign in original strings >> propagated to all translations). > > Good point. > > Would you be willing to update the translations as a group, > if I concentrated on the coreutils side? I think I could. -- Duy
