non-ascii characters and 8-bit Hurd console

[Samuel Thibault]

Hi,

When typing non-ascii characters (for instance 'ç') on an 8-bit Hurd
console, one gets:
ç
which is obviously wrong :) But this is actually a double utf-8 encoding
of 'ç'. Here is the fixup:

[hurd]/console/ChangeLog
2006-01-23  Samuel Thibault  <samuel.thibault@ens-lyon.org>

        Fix encoding of console input.

        * input.c (input_create): Fix arguments order of iconv_open()
        call.

--- console/input-orig.c        2006-01-23 20:33:30.000000000 +0100
+++ console/input.c     2006-01-23 20:33:41.000000000 +0100
@@ -62,7 +62,7 @@ error_t input_create (input_t *r_input, 
   condition_init (&input->data_available);
   condition_init (&input->space_available);
 
-  input->cd = iconv_open ("UTF-8", encoding);
+  input->cd = iconv_open (encoding, "UTF-8");
   if (input->cd == (iconv_t) -1)
     {
       free (input);

Indeed, input should convert input (which is always UTF-8, as
documentation says) into the application encoding. Not the contrary :)

Regards,
Samuel