Comment on PEP 263 - Defining Python Source Code Encodings

[Stephen J. Turnbull]

>>>>> "Martin" == Martin v Loewis <martin at v.loewis.de> writes:

    Martin> Can you quote chapter and verse where it states that? In
    Martin> respect of this text, how do you interpret

No, I can't.  It was just wishful thinking on my part.

You have my apologies for not rechecking.

According to Annex F of ISO 10646-1, the use of the signature is
optional, to be decided by the application.  Amendment 2 specifically
updates that for UTF-8, so you are right.

However, I think it very unlikely that many applications on Unix will
ever adopt the UTF-8 signature, because it would break cat(1) (among
others).

And these signatures certainly don't belong in internal objects, like
Python strings.  They're purely for interapplication communication (as
you point out elsewhere, Python surely knows implicitly whether it's
BE or LE).

XEmacs will support UTF-8 signatures, but (by default) it will
complain as loudly as I can make it do without actually interrupting
execution of a function.

-- 
Institute of Policy and Planning Sciences     http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba                    Tennodai 1-1-1 Tsukuba 305-8573 JAPAN