Re: [hybi] Last Call: <draft-ietf-hybi-thewebsocketprotocol-10.txt> (The WebSocket protocol) to Proposed Standard

[Mykyta Yevstifeyev <[e-mail]>]

12.07.2011 8:39, Thomson, Martin wrote:
> On 2011-07-12 at 14:40:52, Mykyta Yevstifeyev wrote:
>> Section 4.2.  I have a concern with usage of ABNF here.  Let's see:
>>
>>>      frame-fin               = %x0 ; more frames of this message
>> follow
>>>                              / %x1 ; final frame of this message
>> Which means that ABNF represents characters only; no bits may be
>> represented by it.
> That's not true.  RFC 5234 merely uses ASCII in its example.  The terminal values for this particular application of ABNF could be a single bit in place of a character.  5234 doesn't prohibit this
ABNF is designed to deal with characters represented in octets.  The 
terminal value %xHH stands for ASCII character HH, so it's true that %x0 
stands for NUL; so does %b0.  The <BIT> production doesn't denote actual 
bits, but rather representation of bits.

> BIT            =  "0" / "1"

implies that <BIT> may be either ASCII 0x31 - "1" - or ASCII 0x30 - 
"0".  Once more, ABNF is for chars/octets, not bits.

Mykyta Yevstifeyev
> , though it would make most of the default rules meaningless and it would require a little bit of care.
>
> [ . . . ]
>
>
> I notice that the picture in 4.2 says only 63 bits if the first length is 127, which is an error, sort of.
>
> --Martin
>
>
>