On Wed, May 29, 2002 at 02:38:19PM +0100, Nick Ing-Simmons wrote:
> Roca Carrio <[EMAIL PROTECTED]> writes:
> >Hi,
> >
> >BS2000/POSIX-BC platforms use the OSD_EBCDIC_DF04_1 character set which
> >differs from the IBM's MVS Code Page 1047.
> >UTF-EBCDIC is based on Code Page 1047 
> 
> Not exactly. UTF-EBCDIC _Must_ use tr16 translation table or resulting code 
> is not UTF-EBCDIC as defined by the report but some "other" rendering 
> of Unicode of your own devising.
> 
> This may mean that which chars are "invariant" are different for other EBCDIC 
> pages though, but that does not seem to be what your patch is affecting.
> 
> I think the patch is wrong.

Uhh, okay.  So retracting the patch...

Ignasi, could you show a minimal example of a "Malformed" test?
And what does for example this show?

../perl -Ilib -MDevel::Peek -e 'Dump(chr(256))'

> -- 
> Nick Ing-Simmons
> http://www.ni-s.u-net.com/

-- 
$jhi++; # http://www.iki.fi/jhi/
        # There is this special biologist word we use for 'stable'.
        # It is 'dead'. -- Jack Cohen

Reply via email to