Language tags in IPP (was: Re: [Suppress-Script] Initial list of 300 languages)

[McDonald, Ira]

Randy Presuhn wrote:
>
> Ira McDonald wrote:
> > IEEE-ISTO PWG 5100.5 "IPP Document Object" (October 2003)
> > defined 'document-natural-language' and 'document-charset'
> > metadata attributes to tag the _content_ of the print data.
> > Because newer documents transferred in a Unicode charset
> > (e.g., UTF-8) don't allow any meaningful guesses about
> > character repertoire from the charset tag, most printers
> > and application software for word processing have been
> > guessing repertoire based on the language tag.
> ...
> 
> Why is there a need to determine the repertoire for printing?
> Would the guess made change how anything is actually rendered?
> What happens when the printer guesses incorrectly?
> 
> To a naif like me, the question whether a particular code point
> can be rendered in a particular font would seem to have a binary
> answer, quite independent of the repertoire of the rest of the text.
> 

There is a need to determine the repertoire for printing
in order to ACCEPT or REJECT the entire print job (before
scheduling the print job and examining the possibly large
amount of print data).

Most printers actually render using _multiple_ fonts,
selected on a best-effort basis (just like modern word
processing software does on the client side).

Users sometimes send documents requesting specific fonts 
(for which there are NO naming standards, by the way) and 
printers and print spoolers do best-effort font matching.

Printers do NOT simply print what they're told to print,
Because what they're told to print is almost always 
underspecified.

That said, I repeat - I don't want to try to teach this
mailing list how printing really works.

Ned - thanks for your illuminating note on email - but
Mark seems to have missed the point entirely - to whit,
support for UTF-8 in email readers has NOTHING to do with
administrators and users actually throwing away their
"legacy" email systems and moving to pure Unicode.  

There are lots of benefits to Unicode - thus IPP specs have
always REQUIRED support for UTF-8 by all printers.  But it
will be many years before most _deployed_ applications make
it reasonable to shift to Unicode everywhere, if ever.

Cheers,
- Ira

Ira McDonald (Musician / Software Architect)
Blue Roof Music / High North Inc
PO Box 221  Grand Marais, MI  49839
phone: +1-906-494-2434
email: imcdonald at sharplabs.com