Re: (OT) vim, osx, text encoding (was: magic diacritic markup?)

[Trevor Baca]

> Are you using 10.4?  (see below)

Yes, 10.4.2.
 
> > What do these values give you:
> >
> >   :set encoding?
> 
> Entered in command mode?

Yes, command mode.


> E519: Option not supported: encoding?
> (same result for ":set encoding")

OK, that's just odd. vim definitely takes an encoding and for unicode
what you're looking for is encoding=utf-8. Maybe your previous version
of vim wasn't compiled with the multi_byte option set (more below), or
maybe someone else can comment on how on earth vim wouldn't support
the encoding setting.

> ":set" just gives me this:
> :set
> --- Options ---
>    backspace=2         scroll=14           ttyfast
> ttymouse=xterm
> Hit ENTER or type command to continue

Right, that would be consistent with the no-unicode situation you're
describing. I'd be willing to be that your encoding is defaulting to
"latin1" (ie, no unicode), which in turn is probably deriving from
your value for "fileencoding" not being set.


> I finally tried using vim from fink.  Fink has version 6.3-4; OSX
> 10.3.9 uses
> 6.2.  I didn't change my terminal settings or ~/.vimrc... and now utf-8
> stuff
> works.  I can see digraph chars, I can see the smart quotes in the
> lyrics
> section, etc.
> 
> I don't know if you're using 10.3 or 10.4, but I know that Han-Wen uses
> 10.3.9, and I'm pretty certain that he can use utf-8, so I find this
> quite
> curious... but at least it's working for me now.

It definitely works with Apple's build, though, so you shouldn't have
to go through fink (unless, of course, you prefer). Asking for the
version of vim in command mode

  :ve

gives me

  VIM - Vi IMproved 6.2 (2003 Jun 1, compiled Mar 21 2005 02:12:42)
  Compiled by address@hidden

and the unicode stuff all works great *after* having set
"fileencoding" and "encoding" in .vimrc. I really think if you just
had the following two lines in your ~/.vimrc that everything would
probably work great:

  set encoding=utf-8
  set fileencodings=utf-8,latin1

but it's also possible that your previous version of vim wasn't
compiled with the multibyte option set (which you can see from the
same version screen mentioned above at :ve in command mode ... where
you can look for the +multi_byte option; + means the option was set at
compiletime, - means it wasn't).

Anyway, glad the fink version is working for you! Note that *very*
high unicode values (like the hiragana and katakana characters near
the bottom of the :digraph cheat sheet) will print with an extra
trailing character in most vim sessions; probably not an issue unless
you specifically need those characters for lyrics or something.

Trevor.