[Suppress-Script] Initial list of 300 languages

Peter Constable petercon at microsoft.com
Tue Mar 14 18:35:10 CET 2006

> From: McDonald, Ira [mailto:imcdonald at sharplabs.com]

> Umm - all existing network printers are going to fail to
> "do something reasonable" if you send them ANY language
> tag that includes a script subtag - because they don't
> have parsers to take it apart - so they'll fall back to
> the administrator/manufacturer configured default document
> language.  This applies BOTH to external tags (in print
> protocols) and internal tags (in print datastreams, like
> PostScript or XML instance documents).
> Script subtags are important, but they're new - existing
> software (in the firmware of printers or spoolers or file
> systems or client applications) won't know what to do with
> them - and existing matching algorithms will often fail.

Then that's not an issue I think should be driving what we do except for
cases like en-US. At some point, people will be printing content that
was tagged with language tags like az-Cyrl or el-Latn. If existing
printers don't do the optimal thing, that's not our problem -- it's
simply a scenario that wasn't supported in older implementations. All we
need to worry about is that scenarios that *were* supported in those
implementations, things like en-US, continue to work.

If someone tags data now as "sga", are you telling me that printers will
do something different than if there was no tag or if the tag was "en"?
(I find that surprising in this particular instance.) If someone has
sga-Latn and sga-Latg content, what bad result will occur when they try
to print this content?

Peter Constable

More information about the Ietf-languages mailing list