[whatwg] [hybi] US-ASCII vs. ASCII in Web Socket Protocol

[NARUSE, Yui]

(2010/01/31 2:05), Julian Reschke wrote:
> Ian Hickson wrote:
>> On Fri, 4 Dec 2009, WeBMartians wrote:
>>> Hmmm... Maybe it would be better to say ISO-646US rather than ASCII.
>>> There is a lot of impreciseness about the very low value characters
>>> (less than 0x20 space) in the ASCII "specifications." The same can be
>>> said about the higher end.
>>
>> Where the interpretation was normative, I've used the term
>> "ANSI_X3.4-1968 (US-ASCII)" and referenced RFC1345.
> 
> I think you just lost both readability and precision.
> 
> Please keep saying "ASCII" or "US-ASCII", and then have a reference to
> the ANSI or ISO spec that actually defines ASCII, such as
> 
>    [ANSI.X3-4.1986]  American National Standards Institute, "Coded
>                      Character Set - 7-bit American Standard Code for
>                      Information Interchange", ANSI X3.4, 1986.
> 
> (taken from the relatively recent RFC 5322).
> 
> RFC 1345 is a non-maintained, historic informational RFC that's nit
> really a good definition for ASCII. If you disagree, please name a
> single RFC that has been published in the last 20 years that uses RFC
> 1345 to reference ASCII (I just searched, and couldn't find any).

The use of US-ASCII and ASCII in draft-hixie-thewebsocketprotocol-54 is correct.
Changing all to ASCII or ANSI_X3.4-1968 is not correct.

In draft-hixie-thewebsocketprotocol-54, allthe term "US-ASCII" are used as
"encoded as US-ASCII". This use is as encoding name.
So the prefered MIME name, "US-ASCII" is correct.

"ASCII" is used as
* ASCII case-insensitive
* ASCII lowercase
* ASCII serialization.
* ASCII <a char> like "ASCII :" or "ASCII CR" or "ASCII space"
* If /code/, interpreted as ASCII, is "407"
* upper-case ASCII letters
* Unicode to ASCII
* the IDNA ToASCII algorithm
* UseSTD3ASCIIRules flags
They looks refer to so-called ASCII, not definitions in the spec of ASCII.
So the nickname "ASCII" is suitable for them.


Anyway,
latest so-called "ASCII" definition is named "ANSI INCITS 4-1986 (R2007)".
http://webstore.ansi.org/RecordDetail.aspx?sku=ANSI+INCITS+4-1986+(R2007)

And its ISO version is "ISO/IEC 646:1991 IRV".
http://www.iso.org/iso/catalogue_detail.htm?csnumber=4777

-- 
NARUSE, Yui  <naruse at airemix.jp>