Re: Rechartering HTTPbis

On 25/01/2012, at 7:51 PM, Poul-Henning Kamp wrote:

> In message <0C615921-7EE0-4E53-93F9-8B406D1561A1@mnot.net>, Mark Nottingham wri
> tes:
> 
>> Thank you. My *personal* responses below.
> 
> Appreciated.
> 
> I have two problems with the proposed project:  The objective and the
> timescale.
> 
> The timescale is just rubbish, it's not going to happen and we know it.
> Pretending otherwise just makes everybody look like ceremonial fools.

In my experience, standards groups given generous schedules will fill those schedules (and then ask for more time). At this point it's impossible to know what the timescale is, because we don't know how much work there is to do. However, implementers are already deploying a candidate, and the longer they do so, the less likely we'll be able to change it to incorporate other concerns.

As such, I'd rather put pressure on the effort to deliver, and throw an exception when it needs more time, rather than let it pontificate for three years and come up with nothing. I'm happy to re-charter the effort if and when it demonstrates that it needs more time to perform productive work.


> But worse: the objetive is almost guaranteed to become a failure,
> unless managed very ruthlessly.

Go on...


> If there ever were a protocol subject to Second Systems Syndrome,
> it would be HTTP/2.0.  Everybody and his web-programmer is going
> to have opinions and we'll never get through their "input" in finite
> time.

Yes. We've managed that in HTTPbis, and we'll continue to do so. New features go elsewhere, unless they're specifically chartered in.


> My suggestion:
> 
> Make a public call for HTTP/2.0 protocol proposals.
> 
> Rules:
> 
>    1.  Each proposal SHALL be described in a single ID.
> 
>    2.  That ID SHALL be 29 pages or less.
> 
>    3.  The ID SHALL be an RFC-ready description of the protocol.
> 
>    4.  Deadline is 2012-06-01 00:00:00 UTC
> 
>    5.  We decide what to do next after the deadline.

We're actually not too far from this, discounting the size requirement. The current plan is to gather proposals and evaluate in a few months; if we can't get consensus, the WG will stop HTTP/2.0 work.

I'm happy to adjust the charter proposal to make this more clear.

BTW, Mike's current proto-draft of SPDY weighs in at 44 pages.


> Unless we get at good proposal that way, HTTP/2.0 is not worth our time.


"Do it my way, or it's not worth our time?" Really?



--
Mark Nottingham   http://www.mnot.net/

Received on Friday, 27 January 2012 23:32:09 UTC