ietf
[Top] [All Lists]

Re: Long-term IETF evolution thoughts

2016-06-15 11:28:34
On Wed, Jun 15, 2016 at 9:01 AM, Michael Richardson
<mcr+ietf(_at_)sandelman(_dot_)ca> wrote:

Randy Bush <randy(_at_)psg(_dot_)com> wrote:
    >>> All the cool kids are using slack, and what matrix.org has pulled off
    >>> is quite impressive.
    >>
    >> which is just an HTML interface to an XMPP server.

    > so xmpp done good.  and your point was?

The "cool kids" are just using XMPP, they just aren't cool enough to know it.
I think that slack is a walled-garden, but maybe I'm wrong.

Well, it federates. And honestly, I was pointing more to what matrix
was pulling off rather than slack. Actual micro-home-servers, for
example, written in more accessible language than erlang. apis in json
rather than xml. security and webrtc integration, so messaging, longer
documents, and voice/video are integrated.

And so on...

http://matrix.org/docs/guides/faq.html#what-is-the-difference-between-matrix-and-xmpp

Maybe it could be told to send XMPP traffic to jabber.ietf.org.
So, before someone says we should use "slack", they might want to just
realize that we already "do"

Vs with "email" which the ietf is only getting around to sort of
fixing the security of, which, in particular, looks always to be too
hard to be able to run inside the home, starting with the reverse dns
requirement and the basic need for public ipv4 addresses, and ending
with controlling spam (I would like to see "email over some better
transport to natted nodes" happen somehow. https://darkmail.info/
getting anywhere?)

Given the state of the law nowadays, I do not think enough protections
in the cloud exist or can ever exist, and efforts to move data back
closer to the actual owner of it, I applaud, in addition to good
crypto of it wherever it may rest or be exchanged.

Security starts with physical security.

...

Please note that I started this portion of the thread as one example
of how the next generation is or could be interacting, where a major
touchdown point is not a keyboard and screen at a desk, but over a
phone or tablet, where the increase in interactivity makes typing a
seem lame, and the typewriter plain text spec of a RFC, less useful
than html and links to actual, complex code.

There are other encouraging things out there (like tox) - in addition
to the discouraging issues with xmpp (example:
https://wiki.bitlbee.org/HowtoFacebookMQTT )

Isoc uses "fuze", which has a handy ability for participants to record
(and notify others you are recording). Can you do that with webrtc?

someday, perhaps, we'll be able to ask our personal copy of siri to
search our text and personal audio/video logs to find whenever the
heck we'd discussed some thorny issue or another, and/or be able to
query the sousveillance and surveillance infrastructure for it.

Or we can just wait for clippy to do all that for us.

https://twitter.com/dantz/status/742336999084961792



--
Michael Richardson <mcr+IETF(_at_)sandelman(_dot_)ca>, Sandelman Software 
Works
 -= IPv6 IoT consulting =-






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