[ale] IPv6 local devices with a prefix that may change

[Alex Carver]

It's not really a device of yesterday either. I don't see a particular 
need for devices having an entire OS running on them all the time. Why 
does a smartlock or a remote temperature sensor need an entire OS 
instead of just responding to low level network traffic?  Matter and 
Thread and all of those are very deep implementations but they all seem 
to want some very hefty computing power to keep them going on the end nodes.

On 2022-11-04 10:46, James Sumners wrote:
> Yeah, the devices of yesterday basically need to be left behind. What 
> frustrates me are all of the IoT devices that do not support IPv6 at 
> all. It sounds like the new "Matter" standard addresses that, but that's 
> another decade of device transition.
> 
> On 2022-11-04 13:41, Alex Carver via Ale wrote:
>> That's not quite the stability I was talking about. Using my example,
>> in the past ten years my prefix would have changed seven times (four
>> moves, three different ISPs, each happening independent of the other).
>>
>> We're also assuming that the devices and the software can understand
>> and use mDNS. That wasn't a given. In a private IPv4 setup, I just
>> know the IPs and some devices didn't have a provision to have a name
>> or define a name themselves so there would be nothing to ask of them
>> (think embedded microcontrollers)