On Sun, September 24, 2006 19:56, Dan Jacobson wrote: > MR> "About the CJK issue: there are only a few text->PS converters that > MR> support CJK correctly. Usually we Japanese use a2ps or a2ps-perl-ja > MR> with the encoding option, before printing (it means that we don't > MR> trust any PPD to correctly handle any CJK encoding)." > > Well that is nice bug all I know is that > $ w3m -dump http://jidanni.org/index.html|a2ps -o /tmp/f.ps; gv /tmp/f.ps > shows that a2ps is not ready for Unicode. OK, I'll file a bug against > a2ps too.
I wouldn't be surprised if it weren't compliant. Anyhow, as we're starting to notice, bug 387752 has nothing to do with cups-pdf per se, since we rely entirely on Ghostscript for producing the Postscript document and on CUPS for printing it. My gut feeling is that Ghostscript is beginning to show its age on two counts: improper UTF handling and default fonts that don't cover a wide enough segment of the Unicode namespace, resulting in missing glyphs. I have CC'ed the Ghostscript maintainer here for comment on that assumption. -- Martin-Éric Racine http://q-funk.iki.fi
