On Jan 15, 2004, at 4:41 AM, Peter Eisentraut wrote:
> Dave Page wrote:
>> That's not exactly the case. It can be easily redesigned, however we
>> have other priorities right now which are more difficult - eg.
>> Internationalisation.
<snip/>
> The same mistakes have
> been made before, see catgets, Java properties, various similar
> key=>string attempts. Check out WML and gettext for better approaches.
> Not only do they fulfill the criterion of not separating text from
> context, they also have tool support and are actually used in
> situations similar to ours.
Have you looked at xml:tm (translation memory) or XIFF at all? There's
a recent article at xml.com about it
(<http://www.xml.com/pub/a/2004/01/07/xmltm.html>). I don't know a lot
about it, but am wondering if it wouldn't be another option. I don't
know much about WML either, but am off to learn!
Michael Glaesemann
grzm myrealbox com