>From: Scott Brim <sbrim@cisco.com>
>Date: Fri, 12 Jan 2001 14:09:40 -0500
>To: Mark Nottingham <mnot@akamai.com>
>Cc: webi@lists.equinix.com
>Subject: Re: Proposed charter
>X-Mailer: VM 6.89 under Emacs 20.7.1
>Sender: owner-webi@equinix.com
>
>Just one question ...
>
>On 11 Jan 2001 at 18:16 -0800, Mark Nottingham apparently wrote:
> > 1) Gather requirements for, and specify, a Resource Update Protocol,
> > consisting of:
> >
> > a) A framework containing:
> >
> > 1. A channel mechanism with associated channel advertisement,
> > description subscription and reliability functions,
> >
> > 2. A message format for the channel payload, and
> >
> > 3. A scalable delivery model for messages.
> >
> > b) Specification of a Web Resource Invalidation Protocol as a
> > payload for this framework, and optionally other appropriate
> > payloads, as determined by the requirements.
>
>Could we possibly back up one step and treat channel mechanisms as a
>possible solution instead of a given? I mean, let's start with the real
>requirements. If you believe channels have already been justified
>somewhere, could you point me to that RFC? Channels imply basic
>assumptions about how intermediaries relate to their content sources.
>
>...Scott
This archive was generated by hypermail 2b29 : Thu Nov 18 2004 - 11:21:30 MST