[ruby-talk:00738] Re: RD with multi charset

Hi,

In message "[ruby-talk:00737] RD with multi charset"
    on 99/08/22, Minero Aoki <[e-mail]> writes:

|Now RD can contain only the one 
|If there is only Japanese document in ruby source code, it will
|cause trouble in non-Japanese environment. But, we (Japanese user)
|strongly want to use Japanese in documents.
|
|Then I propose the way to include more than one charset in RD
|document. The method is 4 step:
|
|(1) a writer writes "raw-RD" in English, and/or his own language
|(2) a writer encode "raw-RD" into "encoded-RD"
|(3) a reader decode encoded-RD
|((4)) a reader convert RD to html/tex/man/plain/...

Hmm, I don't like this idea much.  

I think the most valuable point about RD (or any embeded document) is
you can get script/documentation in one file.  You can read document,
along with code itself.  If you encode the document, say in Japanese,
you can't read it by mere browsing without decoding.

I'd rather choose separate files for other languages; e.g.

  foobar.rb     # main script, embedding document in English
  foobar.rd.ja  # the Japanese document
  foobar.rd.xx  # the document in language xx.  

BTW, what is the language code for German?

|You may think this is too complecated, and/or decrease RD's visibility.
|But both writer and reader need not encode/decode document by his
|hand. RD writer can encode his documents when he pack his software.
|For example, he can do encoding in Makefile.
|A reader, who is user of some package, can decode RD when he do
|"make install" (or setup.rb). If he doesn't need some charset
|(ex. iso-2022-jp charset for English user), he can simply erase it.

If you can consider that the code and the document are two separated
things, which are bound by RD just for convenience, it's OK to choose
this scheme.  But I, at least, still consider them as ONE thing.

                                                        matz.