New multilingual text support

[forcer <forcer@mindless.com>]

I have a few questions about the planned multilingual text
support (files mltext.texi, mbapi.texi in guile-doc/ref/).

- How can i portably store text files? E.g. so that all versions
  of Guile will be able to read them equally?
- What encoding should i use to talk to C routines (e.g. MySQL
  api)?
  We won't be able to use any text munging functions from C code
  anymore, except if we convert to latin1 i guess? (or both use
  UTF-8 when it's finally supported)
- Where is the bytevector? We need a data type to store binary
  data in. (SRFI 4's u8vector)
- I suppose read-char will read scm_char_t (same for read-line,
  etc.) according to some charset (port-specific?). We need
  primitives to do binary IO.

(binary-read size bytes [port]) => u?vector
  This should just read the specified number of bytes (binary)
  and return a fitting u?vector.
  e.g. if size is 8, it should return u8vector, if size is 16 it
  should return u16 vector, etc.
(u8vector->number u8vector) => number
  This should transform this u8 vector to a number. The length of
  the u8vector should be 2, 4 or 8 (implementation defined).
[See also: SRFI 4]

Are there already any thoughts about this?

    -forcer

-- 
((email . "forcer@mindless.com")       (www . "http://forcix.cx/")
 (irc   . "forcer@#StarWars (IRCnet)") (gpg . "/other/forcer.gpg"))