Kris,
Your point is well made. For historical reasons, I tend to focus on
the much more limited industrial market (as is apparent in my
reference to IIoT). 6TiSCH attempts to cover a much broader market
segment, and thus may gain enough traction to be attractive for SoC
development by major semi companies all on its own.
What would be useful in this discussion would be for someone to
enumerate the specific features of 802.15.4-2011 and 802.15.4e-2012
about which people are most concerned. which is to say those
features for which a fear of change has been driving this
discussion. If such a standalone document existed, even if it
references 802.15.4-2011 and 802.15.4e-2012, then it becomes
possible to have a more nuanced discussion of the generic, latest
802.15.4 vs. 802.15.4-2011 and 802.15.4e-2012. It also becomes
feasible to tell chip vendors that "These are the specific features
that we need to remain supported even if the generic 802.15.4
standard evolves away from them." All of that can survive reference
to a spec that will be obsolete and may become unpurchasable when a
later edition is available.
-Tom
====
On 2015.05.01 08:45, Kris Pister wrote:
Tom - at the risk of appearing dense, I just don't get it.
If minimal is expected to be of use for more than a year or two,
then it needs to
stay constant, like 802.11b (or USB, or UDP, or ...). You don't
integrate new
ideas by changing the underlying spec otherwise there is no
interop. You integrate
new ideas by adding (802.11g, USB2.0, TCP, ...), but by saying
"now when we write
'802.11b' on the outside of the box, it means OFDM not DSSS".
How can you call something a standard if it can change and break
interop at any time?
With regard to silicon vendors, either we are successful with
6tisch and it
becomes widely adopted outside of the existing industrial process
automation market,
or we fail and it isn't. If we're successful, silicon vendors
will include 4e capability in
all chips going forward until the market stops buying it (just
like 11b capability is still
present in chips that speak much more advanced protocols). If we
fail at getting
adoption now, then it doesn't matter what silicon vendors do in
the future.
ksjp
On 5/1/2015 8:18 AM, Tom Phinney
wrote:
Hi all,
For me, resolution of point 2) depends on how long we expect
"minimal" to be of use. In the short term, reference to
802.15.4e makes sense. However, if "minimal" is expected to be
of use for more than a year or two, then it needs to be written
in a way that tracks the most recently published IEEE 802.15.4
standard, as that is the one for which chip vendors will invest.
In the short term, many available chips will support
802.15.4-2011 and 802.15.4e-2012, but in the future that
support, for low-cost low-power high-capability SoC IoT devices,
will track the most-recent version of IEEE 802.15.4.
Industrial IoT is not a large enough sub-market to drive
variance from the greater IoT market. Thus it is foolish (in my
opinion) to insist on adherence to what the dominant market will
consider an obsolete spec.
-Tom
=====
On 2015.05.01 08:05, Qin Wang
wrote:
Hi Pascal,
After
the discussion in the thread "Removing the "e" in the
charter", do you still ask agreement about solution on the
second point?
Personally,
I prefer the text suggested by Pat, because it can
decouple 6TiSCH from IEEE802.15.4 MAC. But, if the minimal
draft is not final standard, I will agree the current
solution for point 2.
Thanks
Qin
Pascal,
+1 on both points
Thomas
_______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch
_______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch
_______________________________________________
6tisch mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/6tisch
|
