I think the market has proved that it doesn't particularly think about anything past "Does it work right now?" And proprietary-in-a-box formats satisfy that just fine (usually at a cheaper price point).

We'll no doubt eventually get integration, as there's too much opportunity not to. But yes, the near future is probably going to be an ugly period for interoperability. Good time to start selling IoT bridges though...

In the connectivity sense, or the "I have a bridge to sell you" sense? :)

But generally my perception is that the evolution of the network before surplus CPU power made wonderboxes efficient (a "home router" that does everything!) is that network hardware evolved from simpler pieces into more capable but complicated (save on routers by using switches, save on switches by using hubs, save on hubs by using repeaters, etc etc).

So the original intent of my comment was that if we're faced with a network incompatibility issue (e.g. AppleTalk vs TCP/IP), then it'd be good to be in the business of making translation hardware.

Heh. Yeah.