Randy Presuhn randy_presuhn at
Tue Apr 29 20:42:29 CEST 2008

Hi -

> From: "CE Whitehead" <cewcathar at>
> To: <ietf-languages at>
> Sent: Monday, April 28, 2008 5:40 PM
> Is it your opinion that indicating any script subtag
> (we must use [zxx] if we are to tag these languages
> properly I understand; [zxx] now means a bit more;
> it can be used any case a language tag is not applicable)
> for programming languages would violate RFC 4646
> insistence that tagging is primarily to indicate natural/human
> language???  Thus it's not o.k. in any way to indicate the script
> of programmatic languages?

I think putting a script subtag on zxx is really a bit of a non-sequitur.
It doesn't aid in the identification of the human language, since by
definition it is not a language for human communication.  Let's look
at three examples: C++, Z and APL.

Both representations of C++ (whether one uses trigraphs or not) end
up being the same script, at least as we're able to distinguish them.
For both Z and APL, the non-"Latin" (really non-ASCII) forms are things
for which we don't have appropriate subtags available to us.

In short, even if adding a script subtag made sense for tagging something
that is already defined as being outside the scope of our work, the script
subtags currently available don't make the right distinctions.

What *is* useful is the tagging of the portions of a program or formal
specification that are intended for human consumption.  Languages
tags as already defined handle that task just fine.


More information about the Ietf-languages mailing list