Localization

Just a note: if you're writing documentation in two different languages (or 
more), there's no rule that says you can't just glom them all in one big 
file.  Clearly this becomes crazy when you have a dozen languages, but with 
just two or three, imagine:


=item foo($bar, $baz)

This function grunkulates $bar with $baz.

Cette fonction engroncule le $bar avec le $baz.


This seems the simplest way.  The trouble I can see is:
1) As I said, this doesn't scale well when you have a dozen languages.
2) It assumes that the languages you're documenting in in this one big 
file, are languages that you can express in a single encoding.  (Otherwise 
we have the near-Hell of a document in multiple encodings at once.)
3) It's fine for cases where the module is written by one person who knows 
all the languages he's documenting in.  In cases where there's more people, 
and they don't /all/ know /all/ the languages, my head explodes.

So far I've been consoling myself that the trouble spots in all three cases 
are actually pretty rare, and that the most common case is one where 
there's no trouble: one author maintaining documentation in English and 
Some-Other-Language.

So my nightmare is there's three authors of a complex module suite, they 
all speak English (hopefully); but besides that, one speaks Arabic, one 
speaks Chinese, and one speaks Vietnamese, and they're all trying to keep 
their quadlingual docset in synch (English, Viet, Arabic, Chinese); and 
NOBODY has a Unicode editor, everybody has only an editor that understands 
ASCII and their language-specific encoding, and dumps core when presented 
with any of the other encodings.
--
Sean M. Burke    http://search.cpan.org/~sburke/