[rfc-i] For v3: language tagging, but only where useful

Nico Williams nico at cryptonector.com
Mon Jan 13 13:15:46 PST 2014

On Mon, Jan 13, 2014 at 3:01 PM, Leonard Rosenthol <lrosenth at adobe.com> wrote:
> However, deprecated does NOT mean that you can't use them if you have a
> use case where no other alternative is possible.

To me "deprecated" means "don't add new uses".  I agree that if
there's no alternative it's better than nothing.  But I'm guessing
it's deprecated because of lack of support in many systems for Unicode
language tags.  (e.g., I'm guessing that many Unicode string
comparison implementations lack support for ignoring language tags).

> I had a long talk with the Unicode folks about this issue as their choice
> to deprecate it impacts a variety of existing standards (incl. PDF) that
> utilize this feature.

I find the answer to "Should I be using those language tag
characters?" from the languagetagging FAQ to be somewhat


It's clear that one should prefer language tagging markup wherever
it's available, and it's often true that where there's no such markup
it's best to not use the Unicode language tagging codepoints at all
for lack of support for ignoring them (e.g., filesystem object names).
 But it's not obvious to me that there are no cases where one would
have no other choice.  Perhaps I should assume others have had enough
experience to think that that is obvious, but it'd be nice if the FAQ
were more fleshed out.


More information about the rfc-interest mailing list