There have been and will continue to be experiments in the "best" way to
construct distributed applications. There are many ways to implement
distributed applications already and, over time, the paradigm may well
change. But for the foreseeable future the way that has been endorsed by
the companies that will deliver the products to their customers is web
services. And that endorsement isn't just in the form of a set of emerging
standards, it's in the form of many millions of dollars invested in
software, in infrastructure and in marketing. There's a whole industry out
there!
so what? why is that a justification for demanding IETF's imprimatur on
protocols and architecture that don't make good engineering sense?
the E in IETF stands for "Engineering". why then, do so many folks believe
that it's IETF's purpose to endorse their business models?
Sacred models do not serve a fast paced industry; informed thought and
engineering work do.
true enough. and this applies equally to sacred business models.
but it's naive to dismiss an underlying architectural principle such as
end-to-end transparency, or clean separation of layers, as 'sacred'.
to one who understands why they are beneficial, they are not sacred -
they are (in most cases) essential.
That thought and work occur in the WG, not in a meta-discussion.
not if that WG is focused on promoting a poorly chosen architecture.
we have to have architectural discussions somewhere, and this is as good a
place as we currently have.
We're also not going to make interception proxies go away no matter what we
do.
true. but we don't have to endorse them.
But we can contribute to providing Internet-level scalability and improved
user experience for the web applications that are the current and future
Internet.
indeed we can, and we should. but perhaps we should start by examining the
barriers to scalability that exist in present-day web applications.
If people have specific suggestions for improving the charter, I personally
welcome them. If not, I'd like the IETF leadership to make the decision
about whether we're in the business of doing web application protocol work
or not and then decide what to do with the OPES charter request.
you will, I hope, recognize that these are separate questions.
Keith