Meeting: ASDF WG, 17:00 – 18:15 CET (1600 UTC), February 25, 2025
Location: Meetecho
Chairs: Michael Richardson and Lorenzo Corneo
Etherpad/CodiMD: notes
These notes at:
https://github.com/ietf-wg-asdf/asdf-working-group-notes.git
draft-lee-asdf-digital-twin-05
This meeting will be a working meeting.
More of a design team meeting than presentations and the like.
Please come with your github logged in, and your best kramdown-fu.
Our goal is to take the digital-twin document, get it into a new github
repo,
and break it up into two documents allowing the non-Affordance work to
be
clearly broken out.
(Please don't fear that if you do some text lifting that you'll be the
stuckee forever for it. Unless you want to be.)
Our goal here is to accelerate the efforts around these documents, so
that we aren't just working three times a year.
https://notes.ietf.org/Ru9l7vxJQZSYg-eoiBvOlw?view=
Carsten presents slides:
NOTES about
examples of attributes: the colour of the device.
e.g., SOME devices know their serial number, and some devices do not.
Bart: if one talks about a serial number, SCIM is pretty good for that.
If you are talking about the colour of the chair, or how many legs the
chair has, then SCIM is not as good.
Serial numbers are good, because we have a good idea of what it means.
BB: Lets say a digital twin of a building, there may be many items which
we can't interact with, such as a meeting room has chairs, and they have
a count, and they are hard or soft.
Some devices might know their serial number.
AK: how you interact with the information... when you get a file that
describes an instance or a class of instances. But there seemed to be a
further level of indirection.
CB: What is the lifecycle that we use the information. SDF can be used a
compile/devel time, and does not describe run-time information. When we
want to talk about instance information, we need to talk about when the
instance information turns up... when we do communicate about the
instance? Like an outlet in a building is static location, but an outlet
in a ship might be moving. We must not assume all this interaction will
occur within SDF.
AK: wouldn't we have an affordance?
CB: on what thing is that affordance?
AK: moving ship example... a normal affordance information on the
object, giving a relative location, with a sdfRelation to the ship
itself.
CB: the item that has a location might still be the same. A diesel
engine on a ship, in a factory, on a train. So a specific valve in that
engine has a location ("third from the left" ... that static). But the
surrounding structure is moving.
AK: containers would model where things are.
CB: we have avoided doing an access control model. We might have to do
that later on.
MK: this other category of non-affordance needs a binding. Literal in an
SDF file, but that's kind of an instance.
Back to discussion about colour vs diameter of cylinder. Here is a
property.
HL: afforance is connected to the Internet, and non-affordance is not
connected to the Internet. Only meaningful objects are presented in the
digital twin. Information that can not be represented are
non-affordance.
Dr. Hong will work on non-affordance document!
We will use the term "sdfNonAffordance" as the really really bad term
which nobody likes and clearly is not the final term.
CB: 47 hours to the oneDM meeting, where there might be more discussion.
2-3 hours earlier on a Tuesday would work.
April 2025 Su Mo Tu We Th Fr Sa
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Wednesday April 9th at 12:00 UTC
Wednesday, April 9, 2025 at 8:00:00 am EDT UTC-4 hours
Helsinki (Finland) Wednesday, April 9, 2025 at 3:00:00 pm EEST UTC+3
hours
Seoul (South Korea) Wednesday, April 9, 2025 at 9:00:00 pm KST UTC+9
hours
Los Angeles (USA – California) Wednesday, April 9, 2025 at 5:00:00 am
PDT UTC-7 hours
Corresponding UTC (GMT) Wednesday, April 9, 2025 at 12:00:00