iot should never accept proprietary closed standards. people should be able to know & openly freely interface with th devices of their home.
thread is an improvement over th worse status quo. but it still reserves all rights, still keeps humanity from being allowed to know how any of this stuff works. enslaving humanity to magic devices that only giant mega corps can understand is the opposite of the personal computing revolution we & our homes need.
there were some posts on the just announced "Matter" standard[1], which is basically the zigbee "dotdot" device profiles being ported to Thread. the new alliance is called Connectivity Standards Alliance. but given that consumers don't have access, it feels like an alliance against understandability & respect. just a land grab for companies to ally with each other upon.
Edit: I was hoping there would have been more discussion on that HN post, but your comment is an interesting one. Do you think we should simplify the lower level tech (or maybe rewrite it in more readable languages) rather that keep piling on abstractions & wrappers? Are you aware of any efforts attempting that strategy?
Apologies for whatever baiting I sprung on you! I was definitely very interested personally, have my Internet of Shit concerns, but also: I really had hoped there'd be more discussion in general! CSA/Matter was pretty huge news to me. This whole area is super interesting, and I super wish people could engage in it, I wish it was a frontier of tech that people could follow & participate in & converse about. But we seem to only have highly synthesized droppings to play to with; this is the most consumerized fore-front of computing the world has, as I see it.
I like your question, but it's exceedingly hard for me to form a judgement. Because Thread specifically keeps me from knowing/understanding/learning about it. The underlying technologies are what I want to see: 6LoWPAN, ipv6, udp, AES. But how these are combined & used is ultimately 100% opaque, unclear, deliberately withheld. Perhaps simplification would help at some level, but I tend to think whatever they have is probably on target enough that changing it wouldn't really matter; all the pressure is on how interesting the ecosystem can become, and with such a greedy iron fisted control over knowing what's happening, there's little chance for the tech to form an ecosystem, for it to become interesting, other than to the smallest batch of nerds some place far far far down in giant consumer-electronics company's totem poles.
And, just as damning, while there are standards here, it seems like the overriding push is still that of devices which talk to their 1st party cloud services, so the advantage of interoperability seems extra-abstract; this isn't a useful/usable technology to consumers in any way.
One can check out the OpenThread C API Reference to get a simplified view of what is at stake, what elements there are[1]. Conceivably we could reverse engineer & write a Threads protocol documentation without access to the specs from this, but what a grim sad task that would be. It highlights well though what abstractions there are. Frankly, I think they have about the right tools here. Trying to simplify is often a damning task that only causes problems. I'm far more about surfacing & exposing complexity, & letting the amazing adaptable complex human mind figure out how to harness & use the possibilities that are. Trying to narrow down, or re-conceive something simpler is, has rarely in my view paid off. Except, like, all those times it has of course. ;) Sometimes we can really find better takes. But seeing the OpenThread C API Reference, the exterior shape of this all looks about as expected... alas the interior of it is all murky shadowy darkness. Like every other internet-of-shit system on the planet.
Gruber mentioned Thread in a footnote on his review of the new Apple TV remote. First time I’ve heard of it. Here’s that article: https://daringfireball.net/2021/05/the_new_siri_remote_etc