Re: Register EBCDIC Character Set "OSD_EBCDIC_DF04_1"

[Markus Scherer <markus.scherer@jtcsv.com>]

Martin Duerst wrote:
> I think these are interesting questions, but strictly speaking,
> charset definitions are from bytes to characters, i.e. from
> the legacy encoding to Unicode, and not necessarily back.

I realize that my questions go beyond what the registration strictly requires. However, I am 
interested in getting the information published if it is available, because most implementations 
will not just convert from such a charset to Unicode but also the other way around.

Also, many converters use one-way mappings not just for Unicode->charset but also for 
charset->Unicode, and these are relevant even just considering the narrow requirements in RFC 2978.

Viele Grüße,
markus