Re: [hybi] Changed framing

Greg Wilkins <gregw@webtide.com> Fri, 06 August 2010 01:20 UTC

MIME-Version: 1.0
In-Reply-To: <Pine.LNX.4.64.1008060038340.5947@ps20323.dreamhostps.com>
References: <Pine.LNX.4.64.1008060038340.5947@ps20323.dreamhostps.com>
Date: Fri, 06 Aug 2010 11:21:13 +1000
Message-ID: <AANLkTimSFH1so6Ku4cFV1F0sU74ZE2SbwDoYQ7+0Lesj@mail.gmail.com>
From: Greg Wilkins <gregw@webtide.com>
To: Ian Hickson <ian@hixie.ch>
Content-Type: text/plain; charset="UTF-8"
Cc: hybi@ietf.org
Subject: Re: [hybi] Changed framing
Precedence: list

On 6 August 2010 10:41, Ian Hickson <ian@hixie.ch> wrote:
>
> I've changed the WebSocket protocol's framing to use a 64 bit length
> prefix instead of the combination of sentinel-based framing and
> variable-width lengths that we had before.

Ha! just as I was almost convinced of  variable length!

But unfortunately I have a quibble.  While I appreciate the sentiment
of making to so big that it should not be a problem, I think 64 bits
is perhaps a little too big.  The language I use (java) doesn't have
an unsigned 64 bit integer and I suspect there are other environments
which will also have difficulties with such a value.

Perhaps it should be 63 (or 31) bits, and the high bit is used for the
fragmentation bit.  This would help avoid errors associated with
signed types being used for the signed value.

regards

Re: [hybi] Changed framing Greg Wilkins
[hybi] Changed framing Ian Hickson
Re: [hybi] Changed framing Douglas Otis
Re: [hybi] Changed framing Maciej Stachowiak
Re: [hybi] Changed framing Ian Hickson
Re: [hybi] Changed framing L.Wood
Re: [hybi] Changed framing Ian Fette (イアンフェッティ)