From: bill_kempf (williamkempf_at_[hidden])
Date: 2002-01-31 12:53:25
--- In boost_at_y..., Darin Adler <darin_at_b...> wrote:
> On 1/31/02 7:31 AM, "bill_kempf" <williamkempf_at_h...> wrote:
> > I thought of that, but honestly see no problems here. The width
> > passed to the manipulator in my example and could as easily be
> > fetched out of a locale specific database as you'd fetch out the
> > format string.
> You seem to misunderstand the constraints here. There's already a
> locale-specific string database. Imagining another putative locale-
> database is a big step.
> Perhaps you see no problems here because you haven't dealt with this
> recently. I find that the translators have a much easier time
> single string than localizing a number of separate interdependent
> And it's non trivial to have more than one locale-specific database.
No, I've dealt with this issue a lot. Maybe we just have different
perspectives here though. Generally, I don't think that formatting
issues should be of any concern to a translator. Usually the
formatting concerns are either something specific to programming
needs (such as fitting the data within a given column size) that are
irrelevant to a translator, or are based on requirements for the type
and the locale, such as the width required to represent a date. I've
never encountered a situation where a translator would need to
specify any formatting information, and so including it in the string
makes the translator's job more difficult.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk