Re: [bug-gnu-libiconv] REQ: Encoding MARC-8

[Duncan Smith]

On Mon, Dec 27, 2010 at 11:02:16AM +0100, Bruno Haible wrote:
> Given the answer of the first question in libiconv/NOTES
>
>   "libiconv, as an internationalization library, supports those
>    character sets and encodings which are in wide-spread use in at
>    least one territory of the world."
>
> and the explanation in Wikipedia
>
>   "MARC-8 is rarely used outside of library records."
>
> I don't think MARC-8 support is useful or appropriate in libiconv.
>
> I see at <http://search.cpan.org/~esummers/> that whoever want MARC-8
> charset support also needs more functionality for MARC-8 records. In
> other words, it's really a special-purpose encoding which is best
> implemented together with MARC XML support and more.

Okay, so the purpose of libiconv is to support actual people talking
to other people on the contemporary Internet, as opposed to converting
all manner of strange things to Unicode and back.  That's a reasonable
place to draw the line.

Since I sent my earlier email, I got about halfway done implementing
support for UTF-1 <http://en.wikipedia.org/wiki/UTF-1>.  It's an ISO
standard but "never gained wide acceptance", so would that also not be
suitable for GNU libiconv?

-- 
Duncan Smith

signature.asc
Description: Digital signature