Serious question: Why insist on forcing a cloud service connection, for air vents, that will not work when the internet is down (another issue with non-local control schemes), which most likely use simple basic MQTT protocols for passing messages from control points ("publishers"/the app) to end points ("subscribers"/the device, as in it subscribes to the data feed 'topic' on the brokering device...look up MQTT itll all make sense), and can easily be controlled by whatever MCU is running in that hub puck bridge thing?
I mean, i imagine its probably some proprietary or device filtered form of zigbee or some such layer 2/3 physical/data transport, and with third party firmware installed on ones router (like openwrt, most routers are fully compatible, and it unlocks the full capabilities of the hardware allowing for enterprise level configurations, also DNS filtering (this is what a pi-hole is) which means any nefarious DNS requests for a IP associated with a URL wont even make it out of the network, your router doesnt just drop the packet, it replies "NX_DOMAIN" which in DNS means "that url physically does not exist on the network) one can install an MQTT broker to handel message buffering, delivery, and status updates on the router itself, easily. Which through various forms of piggy backing on googles authentication backend, and some basic dynamic DNS to keep track of IP addresses (although honestly I dont see why the router couldn't detect its external IP change, then send every publisher device a push message which would update the apps to the new IP maintaining connection from outside the network) would mean an entirely local, secure, known, auditable, control system for ones home automation.....and I know, your company wants that juicy backend aggregated marketing data (dont lie to us, seriously. Its insulting. Look up, see how familiar I am with basic tech? Yeah, Im telling you, its insulting to even try, so dont) to cover operating costs, but here is the thing, I have a plus side for the company too.....by allowing users to NOT use your MQTT broker backend your required running costs go down! Less active connections to maintain. Of course...its MQTT...its a dead simple tiny packet size protocol, heck I'm willing to guess the user or the puck or the 'location'/home setting has a UUID which is part of the topic header in MQTT....Ill even put a dollar on it not being as secure as you might think (I wonder if I can send a command to someone else's device just by guessing the UUID string....) so its not like it costs very much to run, it probably runs in some sort of auto scaling microservice sub-contractor co-lo (basically a commercial hosting/data center, in case anyone was unfamiliar with what the back bone of the internet looks like), and yall get charged based on how much processor time your service used last month....less users having to publish to, and check for updates on, your backend, meaning less processor time, meaning less running costs. The marketing data is so innocuous and flawed anyways, just generate some garbage data, no one will ever know.
But you know who benefits? You, flair. You do. Because you will be the ONLY smart device on the market, in any segment, that would give the user control of their devices, locally. That means more tinkerers, more makers, more "say no to the cloud" types, more 'dont trust the man, man' types, more of the people all of the IoT manufacturers dont understand they are missing out on........it means sales. As a turbo tech nerd there are so many home automation products I simply refuse to purchase or impliment explicitly because of the cloud requirements on LITERALLY EVERYTHING when it is absolutely unnecessary and frankly unwanted. Yes, there are people who dont want to or do not have the time/skill to setup and configure a local MQTT server (although, really doesn't even have to run on a router, raspberry pis and clones are cheap and have tons of support and tutorials for setup) who all and all arent too bothered by a mystery organization with mystery black boxes inside their networks, running mystery code, reporting back to a mystery service backend, in some mystery country (I have $2 on china), or worse could be compromised at any moment and turned into a bot net with zero warning.... and yes it is nice there is a 'just plug it in a go' option for them. I am being serious, and honest there, there does need to be a service for those dipping their toes, or are simply not as technologically inclined as me and my ilk (which is valid, we all have different skill sets, knowledge, and experiences. Just because I can spec out a server rack doesnt mean I can coordinate an outfit, or know how to write an effective essay....Im just good with computers....and not being good with them does not make anyone lesser than. Ill fix your computer for you if you, idk, [insert thing you're good at] for me, deal?). However I also know with absolute certainty that you are certainly missing out on quite a bit of buzz and sales by not being one of the only products to offer an "experienced adult" option for fully customizing the control scheme and setting up local only control, and being the only one to do it from the factory, with the native firmware, no jail breaks, no restrictions.... Flair could be pioneers, starting the trend of just making decent, feature rich devices, that put the power in the hands of skilled knowledgeable owners, instead of just another faceless corporation pulling the same tricks as ring (vomit) awaiting its slow but inevitable swirl down the drain (when all its devices go belly up cuz they can only attach to a server that is no longer being ran or maintained, you know like nest did, which by the way also used MQTT as its command structure) by making loss leader products and hoping to make it back on the faux walled garden of forced cloud integration.
Case and point, Tuya. They used an ESP chip in their products, namely their smart sockets and light bulbs. Once someone wrote a third party firmware for their ESP chip controlled devices (tasmota) jailbreaking them from the forced cloud, they saw an appreciable rise in sales, and word of mouth. They were mentioned any time anyone was talking about home automation that was locally controlled. Wouldn't it be nice if Flair was on the tip of everyone's talking too? Tuya could have struck back, and they kind of did by using a different chip (which also has finally gotten a third party firmware) and using a different key for allowing firmware updates (which I'm not sure has been cracked yet, but it kind of doesn't matter because these are known microcontrollers, the ESP and the benken chips, and one can just write the control software themselves and completely obliterate whatever firmware is on the chips.... With a little R&D probably only lose 5 or 10 units before you got the limits of the commanded values right without frying something...) which did indeed hurt their sales. Instead of people saying, "You want rgb+cct bulbs? Dude, goto the orange hardware store a get a box of these. Trust me..." they're saying, "there are these, which are the same as a bunch of other ones. But you gotta be careful which ones you get, otherwise you cant jailbreak them..." Where there once was excitement and buzz in the HA community, there is not stories of caution and warnings that its not worth it. What good is a loss leader product you hope to leverage into income on the back end if you never sell it in the first place? There are wiki's, lists of model numbers and images to go along with them of the devices you want to get to ensure that you are able to flash them with a third party firmware and use them how the user wants to. But I don't see flair on any of them... Wouldn't it be nice to be one of those devices that was explicitly desired by a community of text savvy nerds, ones that are very likely to show off the system they have, talk about which companies to go with, and even agree to build out their friends home automation system who's not quite techy but really wants that cool stuff?
So, I'll ask one more time, why do you need the cloud? There are sales out there just waiting for you to do what you should be doing anyways.... The real question is, "...Will you? Or will you be the next nest...when you could be Tuya in its hayday..."
Comments
2 comments