W3C home > Mailing lists > Public > ietf-http-wg@w3.org > April to June 2012

Re: Make HTTP 2.0 message/transport format agnostic

From: Tim Bray <tbray@textuality.com>
Date: Sat, 31 Mar 2012 17:44:06 -0700
Message-ID: <CAHBU6ivKQypwvHPf09FjENm57DsPnH=qehbRSknHW+jpyo75dw@mail.gmail.com>
To: "Adrien W. de Croy" <adrien@qbik.com>
Cc: Poul-Henning Kamp <phk@phk.freebsd.dk>, Mark Nottingham <mnot@mnot.net>, Kevin Cathcart <kevincathcart@gmail.com>, "ietf-http-wg@w3.org" <ietf-http-wg@w3.org>
Funny you should mention that.  ASN.1 has historically been a
miserable failure; the vast majority of net traffic today is a media
format (jpg/mp4/whatever) or text encoded in HTML/JSON/XML, in most
cases wrapped in an HTTP envelope.  None of these (media formats,
markup language, HTTP) have an underlying “abstract model”;
interoperability is achieved at the level of message syntax and
interchange patterns.

Which is to say, syntax-based-interoperability wins, and attempts to
have an abstract model with a secondary syntax is a trail of tears.
HTTP2.0 is a little unusual in that the operational semantics and
payload-carrying capabilities are predefined, but at the end of the
day it’s a syntax/message-interchange design problem; attempts to do
it at a higher level of abstraction are a waste of time -T

On Sat, Mar 31, 2012 at 4:53 PM, Adrien W. de Croy <adrien@qbik.com> wrote:
>  p.s.
>  I wasn't for a second suggesting we should actually _use_ ASN.1 or any of
> its encodings.
>  It's just the only other system I know of which uses an abstract notation,
> and has numerous transfer encodings, so I wondered if any people knew
> whether that turned out to just be a waste of time because everyone just
> ended up using one of them.
>
>
> ------ Original Message ------
> From: "Adrien W. de Croy" <adrien@qbik.com>
> To: "Poul-Henning Kamp" <phk@phk.freebsd.dk>
> Cc: "Mark Nottingham" <mnot@mnot.net>;"Kevin Cathcart"
> <kevincathcart@gmail.com>;"ietf-http-wg@w3.org" <ietf-http-wg@w3.org>
> Sent: 1/04/2012 11:36:45 a.m.
> Subject: Re: Make HTTP 2.0 message/transport format agnostic
>>
>>
>> ------ Original Message ------ From: "Poul-Henning Kamp"
>> <phk@phk.freebsd.dk>
>>>
>>> In message <em8582f3b4-bdc2-4f0b-ad2f-0dccfd9729fb@boist>, "Adrien W. de
>>> Croy" writes:
>>>
>>>>>>
>>>>>> We should be able to learn from experience here... ASN.1 encoding
>>>>>> rules... have any become dominant?
>>>>>
>>>>>
>>>>>
>>>>> Have any ever broken the 1Gbit/s barrier ?
>>>>>
>>>>
>>>> no idea. What I meant was dominant amongst themselves... e.g. one
>>>> preferred encoding (like DER) rather than the other 6 or 7.
>>>
>>>
>>>
>>> I don't care how dominant, if it doesn't do at least 40Gbit/sec with less
>>> than 10% of a contemporary machine, it's not relevant.
>>
>>
>> that's not the point.
>> I'm trying to see if there is any knowledge out there about whether there
>> is any benefit to having multiple encodings or not, whether experience has
>> shown that it was a pointless exercise or not.
>> So throughput is meaningless.
>>
>>>
>>>
>>> --
>>> Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 phk@FreeBSD.ORG | TCP/IP
>>> since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to
>>> malice what can adequately be explained by incompetence.
>>>
>>
>>
>
>
Received on Sunday, 1 April 2012 00:44:34 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:59 GMT