Re: [mmox] IETF policy question

[Lisa Dusseault <[e-mail]>]

Hi Heiner,

Your concern seems to be preventing fragmentation of work and
standards.  We are actually gaining experience in handling
fragmentation later, along with experience of what happens when we try
too hard to prevent fragmentation.  When IETF has tried to force there
to be "only one" answer we usually don't succeed, usually delay
progress, and often regret the effort later when enough time has
passed and multiple solutions emerge anyway.

I believe you're quite familiar with the SIP area, where the Session
Initiation and Session Description technologies are defined to let
agents negotiate which standard to use.  The higher we get in the
stack, the easier this approach is to use.  We don't get fussed about
how IMAP, HTTP/WebDAV, NFS and FTP access are still different
protocols even though if they were (over) engineered today we'd
probably have a giant discussion about how they're all fundamentally
access to a hierarchical static object store!

That said, we do have an increasing number of useful general-purpose
modules.  Some examples: for session security there's TLS, for
authentication there's SASL, for delegated authorization we're
building OAUTH, for resource download there's HTTP, for document
sharing/collab there's WebDAV, and URIs solve a number of shared
problems.  I'm hoping SDP and XMPP and Atom also prove to be useful
reusable components.

We don't task sub-groups.  Groups self-organize and if a group can
show a focused topic, a list of achievable deliverables, sufficient
volunteers, and IETF relevancy, we typically approve that charter
after making sure that subgroup is actually still talking to other
subgroups.  Split efforts are sometimes a failure of communication,
but sometimes a true differentiation of goals or tradeoffs.

Lisa

On Wed, Mar 25, 2009 at 4:20 PM, Heiner Wolf <[e-mail]> wrote:
> Hi,
>
> Assumed there are multiple systems with incompatible architecture,
> each sophisticated and with it's own protocols.
> Assumed that we as a WG do not know enough about the general problem space.
> Assumed that there are sub-groups which know enough about their
> problem and have working solutions.
>
> What would the IETF do?
> 1. task one of the sub-groups to standardize one of the
> protocol/architecture variants, although it might leave a large part
> of the community out of the loop, while having at least a hope to
> defeat fragmentation in the future
> ...or...
> 2. do not standardize at IETF level, which might be good for the
> competition of ideas and allows the community to learn more about the
> general problem space, but preserves fragmentation unless the market
> cleans it up.
>
> I am really undecided, but there is probably BCP in the IETF.
>
> Best
> --
> Dr. Heiner Wolf
> [e-mail]
> www.wolfspelz.de
> www.virtual-presence.org
> _______________________________________________
> mmox mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/mmox
>