Back to the original point, for a moment:
Stephane Bortzmeyer <bortzmeyer(_at_)nic(_dot_)fr> Wed, 11 May 2016 12:58 UTCShow
header
A very interesting paper (I said "intesresting", I didn't say I
agree!) on open networks where independant nodes with independently
developed programs interoperate thanks to standards. The author claims
closed and centralized systemes are better, because they allow faster
evolution (he uses security as an example).
Many IETF cases mentioned (XMPP, IPv6, email...)
https://whispersystems.org/blog/the-ecosystem-is-moving/
My long-standing observation is that the climate has changed. In the
early days, there was both "demand pull" for new protocols, and an
environment that encouraged (and to an extent) funded new protocol
development and deployment.
Since then, the climate has changed:
- it's very hard to get a new protocol into the ecosystem (there are
quite a few useful protocols, that simply are not supported)
- the drivers have changed from greater interconnection and
interoperability (back to the original ARPANET drivers of resource
sharing and collaboration) - to "can it be monetized?"
It's simply a lot easier to deploy a new SaaS, behind an API, and to
charge for it, than it is to deploy new protocol infrastructure.
The exception seems to be when there is a strong "forcing function" -
applied top-down. DoD Force Transformation & the Command & Control
Research Program drove new operational models into the military
environment - into networks, into system specifications, and into
doctrine. Examples that come to mind:
- XMPP is widely used for tactical chat
- DIS is widely used to support distributed simulation and training -
including deployment of persistent training federations
- Tactical Data Links (e.g., Link-16) are all over the place
- DDS is widely used for sensor-weapon linkages
Also of note - NNTP remains widely used on the SIPRNET, at the top of
the MDMP (Military Decision Making Process)
Another example that comes to mind is the Digital Libraries Initiative -
which forced a lot standards and protocols for library system
interoperability.
IMHO, without such forcing functions, the natural tendency is toward
centralized, proprietary services - and back toward a world of walled
gardens. Even in areas where we have a measure of widespread
interoperability - such as calendaring - we see things like Google
pulling iCal support - making it ever so much more tedious to schedule a
meeting.
Miles Fidelman
--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra