I'm not sure who said:
> I think our concern is that we have finite resources here in the
> IETF. If you want a market decides standards, go set up an industry
> consortium or go to a market decides standards body.
> The IETF works best when people bringing technologies to it buy into
> the idea of building rough consensus.
My hypothesis is that (at today's scale of operation) tweaking
governance does not contribute to building concensus: governance
creates contention. Regarding IETF as a professional organization
rather than an industry consortium: when the governance question gets
hot, that seems to me to lead inevitably to turning the professional
organization into a de facto consortium when
industrial/governmental/NGO players start spending resources in
efforts to dominate it, interrupt it, etc.
The alternative to governance that I see is investment in communications
and trust building.
In one scenario: Alice working in her shop devises and implements a
new protocol that she would like to share and see implemented
elsewhere. The essential steps that seem to help are that Bob,
Candice, and Dave -- all recognized, leading professionals with
clearly contained and monitorable conflict of interest issues --
carefully review, comment on, and help refine her work. Also that the
attention of potential implementors is drawn to this effort. Finaly
that any party can observably endorse or object to the effort.
For that scenario, the actions of Alice, Bob, Candice, Dave and potential
implementors are entirely voluntary. It wouldn't matter if Esther, Fred, and
Grace are trying to "throw rocks", as one person put it -- the cooperators
have their own little moderated mailing list and how much more than that,
really, is necessary?
Aside from the most basic of protections against process spam,
*anyone* should be able to record their spec and call it a standard.
The interesting questions all have to do with "who agrees?" and, for
those questions, no central governance seems either necessary or
Wouldn't a system of mutual endorsements (a web of trust), suitably
loudly broadcast, be an alternative to elaborate committee procedures?
Wouldn't that allow at least the more careful consumers for these
documents to decide, individually, what kind of rough concensus, if any,
has been reached?
This would eliminate the "IETF" tag on a spec as a branding/marketing
tool but that just seems to me to be a return towards being pragmatically
In some sense, all working groups could be understood as voluntary
associations. Such associations could happen anywhere. The values
added by coming together under the IETF umbrella seem to me to be that
many people watch this central communications hub and that, at least
at this point in history, a subset of participants are clear
thought-leaders whose endoursement carries a merit-based weight.
If IETF continues down contentious paths, the value of clear
thought-leaders will fade amidst the noise. Therefore, it seems to me,
we are at a unique point in history where a shift to a system based
more on a web-of-trust approach is both an option and a potential solution
to a present problem.
Ietf mailing list