Mats Lidell writes:
>>>>> Stephen wrote:
Stephen> "Someday" we'll move to a Lisp-based mechanism anyway because
Stephen> Xft can't handle those resources properly AFAIK (at least for
Stephen> non-Latin-1 characters), and those files will only be useful
Stephen> for backward compatibility.
Could you please explain? Would adding more languages to
mule-packages/locale be a waste of time/effort?
No, it would not be a waste. Although the mechanisms are rather
different, both the plausible technologies I know (X resources and GNU
gettext) are configured in one file per language, which provides a
mapping from an abstract message to the actual string in some
language.
The important thing to recognize is that the two methods implement the
same mapping. So since the raw data is organized the same way (ie,
file per language), it will be straightforward to extract translations
used in the X resource files, and automatically generate a pretty good
set of translation for another system such as gettext.
(Which could explain why so few languages are there at the
moment!?)
I think that this is not the reason. Rather, since only the menus and
startup screen are translated, we simply never put a lot of effort
into getting translations.
Specifically: I just started with adding support for Swedish. Should
I
stop?
No, I don't think so. Your effort won't be wasted as long as you know
people will use it. And it will provide a basis for further work even
if we change the implementation of localized messages.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta