long term ietf evolution comments

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



I really applaud this, thank you.

https://www.ietf.org/blog/2016/06/long-term-ietf-evolution/

Fairly recently I gave a short preso [1] on what I think is a really
core problem every org involved in the internet has -

that bugs accumulate across layers of technology and social
interaction, and breaking down the barriers of language and
understanding across all layers of the stack - technical and social -
is one of the key unifying things the ietf can do.

I had a ton more direct comments than this, but these anecdotes are
what came out of me first:

...

When I first came (back) to ietf, I was making some serious progress
on the bufferbloat issue, dead broke, with just enough money for
airfare to canada. I got there, got a cot in jim getty's room for the
week... and

I appreciated very much the fact that badges were not checked at the
door, and I was able to breeze in behind jim and start to make a
difference. Had I not made that meeting, a bunch of things would have
gone down very differently.

...

Later on...

One of the most single (and most accidentally) effective things I've
ever done was personally cover simon kelly's (author of dnsmasq) costs
to come to an ietf meeting. I'd tired of people whining about DNSSEC
not getting deployed and hit upon the idea that maybe the guy
responsible for over a billion dns servers in the field should maybe
get a floor under him to get DNSSEC to work.

That triggered meetings with other dns folk, and a meeting with a core
ISP (comcast) that agreed to fund the work.

It's taken 3+ years of development and bugfixing since, but DNSSEC in
dnsmasq is now widely available and increasingly enabled. (and even
better, I haven't had to do anything more myself on it!)

Ideally I'd like there to be an ietf mechanism where "X" number of
folk doing valuable work could have their costs covered for travel and
attendance, to give a presentation, and intermix, for a value of X > 1
and less than 50.

Perhaps a registration method where you could do something like
register you and "guest" for regular ietf participants, would be more
broadly applicable?

...

Along these themes the costs of actual attendance are quite high in
travel costs and lost work, and what ietf covers in a week impossibly
broad, and yet at the same time overfocused on narrow issues at times,
while ignoring huge, glaring ones.

Security foci, overall, getting better, post snowden. :whew:

Highly dynamic ipv6 addresses and service discovery, not so much.

...

I strongly approve of more running code, before rough consensus, and
whatever it takes to get that, go for it.

...

In an age where you can always "find what you want" - aggressively
incorporating the element of surprise, the random meeting in the
hallway with a strange attractor that sends someone down a different
path, seems to be an increasingly valuable.

Getting different groups that are working to solve similar problems at
different layers, or technologies, to have a social, or to have one
meeting to describe each others work to each other, to try and
struggle for a bigger picture about how all the pieces fit together,
might be nice.

Something like that might need to be held in a mud wrestling ring
particularly when it comes to the various routing wgs!

...

I'd like it if there were more, shorter meetings, somewhat overlapping
and in the same locations as IEEE, ITU, and open source conferences
like the LF events, etc. I'd really like to see interop events like we
used to see in the old days, again, also.

My overall theme - as in this message - is more mixing of academics,
technologists, theorists, coders, isps, vendors, government, kids,
oldsters, hipsters, freaks, lawyer, males, females, cats, dogs, etc -
would help.

The biggest thing I've valued from ietf meetings since starting to
attend has been the chance to have a beer and short RTTs with people I
argue with regularly over email. We may never see eye to eye, but we
can usually agree on each others humanity, and that helps a lot.

We're all trying to bring spaceship earth, in for a safe landing.

That said... well... I think meetecho for handling remote attendance
has got *quite good*, and perhaps robots can suffice for more face to
face time.  :)

...

In my own case, I intend the berlin meeting to be the last one I make
for a while - if I make it! it's been too hard, financially and
physically, to make the last couple.

My principal focus is on fixing wifi now, and while I see some
interest ramping up in the ietf, most of its issues come from bad
layer 2-3 interactions and the lack of good fq/aqm technologies. [2].
I'd like it if the ietf poked into the possibilities inherent in usb-c
networking, also, and into exciting new technologies possible with the
distributed web.

I'm glad this upcoming meeting shares a town with some of the core
openwrt and freifunk developers, if you can't find me in the hallways,
you'll find me at c-base.org!


[1] https://www.youtube.com/watch?v=QWdL2Wu7M-8 "Why the
insecurities"? (the last 7-8 minutes primarily)

-- 
Dave Täht
Let's go make home routers and wifi faster! With better software!
[2] http://blog.cerowrt.org





[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]