Recent searches
Search options
Home automation and open source software acceptance is a usability problem.
Example: even if your vision is 20/20, when you’re turning on a light you are a low-vision user of the device.
Light switches can be found by sweeping your hand in any direction. They’re tactile.
Hue replaces this with a flat piece of plastic with three options you can’t distinguish in the dark. It’s a remote that comes out of the wall, so you can knock it to the floor while trying to find it. Needs batteries, too.
But you could download and install an app, then find wherever the hell the tinkerer put the little box with a button to press before the app works, then as long as you can find your phone in the dark and it’s charged, you can simulate the action of flicking a switch!
Listen to yourselves.
Unless this process solves a problem, you will face resistance.
Home automation can be more accessible. A light switch at wheelchair user height? Fabulous. Voice control for people with mobility issues? Fantastic.
The thing about accessibility tech is that, since disabilities vary between people, and can change from time to time, people who use the tech need to be consulted.
Even if everyone in your house is not disabled (yet), go through the same process. Your obstacle is not a lack of tech savvy, it’s a lack of product management.
Another use case:
It’s not that I *can’t* do the thing.
I *can’t be arsed* to do the thing.
Dearest tinkerers, my soldering is neater than yours. I can put on SMDs without flux, in a pinch. Get off my lawn.
“I’m happy for you that you want to solve a riddle to turn on a light. Follow your bliss. I want a godsdamned BUTTON” is a use case.
@akareilly and this is why the zigbee controlled lamps in this house still have a wall button where the inconvenience is that you may sometimes need to toggle it twice instead of once in exchange for color temp and intensity control. It's still a downgrade that may not survive the next revision because it fails the Basic Usability Test.
@akareilly
There's an annoying market trap here. In-wall stuff almost always means an electrician is installing it, so companies there target whole-house systems with custom wiring, which is only relevant for rich folks. Because this moves at architect speed, these systems are generally still dimming only, and if you want full color, you're looking at stuff built for professional event spaces. All of the nice wireless color stuff is targeting the lowest possible bar for end user installs, which means switches can't be attached by anything other than a command strip and have to be battery powered, hence removable.
There's no reason why someone couldn't make a box that replaces the in-wall switch, leaves the circuit always on, powers the transmitter off wall power, and gives the user nice firm predictable tactile buttons like they're familiar with, but it would need a lot more regulatory approval and in make markets a professional install, so it would cost ~8 times as much to the consumer, making it a specialty product relevant for maybe .5% of their customers, and thus not worth building. It's also a conceptually weird object, so it would take a lot of user education, making it a hard sell even for that .5%.
And thus, the world will continue to be awful.
@dymaxion @akareilly Yeah… I automated my home but no way it’s ready for final users.
Not only because of all you said, but it degrades (a litany of third parties changing stuff I guess) so it’s a moving target: who wants to debug when you’re on the go?
There is a market for a liberal professional to facilitate work… HA is too personal for big corporations to care, so this professional could suggest devices, workflows, assign roles, produce updated cheat sheets, and act as IT.
@dymaxion @akareilly Frankly just deciding devices (on one side navigating ever shifting corporate permissions, on another getting bulk discounts) make them cheaper than going on your own. Think accountant.
I know this professional exists for rich clients. But a dynamic of haute couture (highy personalized), pret a porter (tested and true practices, automated and more rigid) is possible.
@dymaxion @akareilly American corporations are on enshittification path… they can *promise* smart home personalization, but they can’t provide even for their existing markets, imagine new ones.
In time, in my experience US HA products were expensive and full of lock in traps, while Chinese ones were cheap, nimble, and offered all features and the kitchen sink.
@nonlinear
Building these systems isn't really my hobby, so I'm buying on the basis of software stack quality and reliability alone, which has meant nothing but Hue. I'm aware that at some point I'll be forced to replace all five bridges with a couple of home assistant installs, but this side of the house has already taken an order of magnitude more time than the professional side, so I'm really uninterested in sysadmining it harder. Otoh, the professional gear takes six weeks to get and costs a grand or more a fixture, while I can buy Hue bulbs at the hardware store.
The hardware layer won't get less crap until the business model/consumer expectations change, and thus the only real place for usability improvements is at the software layer, aka where the enshitification is happening.
@akareilly
@dymaxion @akareilly Yeah. It’s impossible to decide on which best device to add to your ecosystem based on *current* specs and *future* business decisions that may enshittify your purchase.
Best is to choose based on less dependencies, “dumb” appliances that don’t need to talk with mothership to operate, so your smart home doesn’t deprecate as billionaires fight.
Or don’t try it at all, because it’s a costly hobby, both in time, money and patience.
It’s all very bittersweet.
@nonlinear
I've very carefully only touched lighting when it comes to consumer home automation — I run our house as a professional event venue, but not at a level where the professional venue lighting options would be accessible. Of the three landlord-installed smart devices, one immediately broke to a degree where it had to be removed and the other two have had intermittent failures. Everything else is either as dumb as it came, or automated with decoupled professional control layers (e.g. Dante and Artnet). It still takes an inconvenient amount of at unplanned times.
@akareilly
@dymaxion @akareilly There are actually Hue/Zigbee wall switches that use energy harvesting. Very nifty: That satisfying "click" is thus the very thing that briefly powers up the radio element to send the signal.
The main challenge in this household to switch out all the legacy wall switches with that system is a) the ridiculous cost of them and b) I'd still need to hotwire the circuit behind the switch and then have no easy option to power off the lights when needed (most zigbee lights get reset by switching them on and off 6 times in fast order.. )
Which is why of the 16ish light strips on my ceiling, half are controlled by the Hue basestation, half by the Tradfri one and half by the Home Assistant - and a few by nothing at all. And to change that, I'll have to basically dismantle the whole thing to reset each light strip manually...
@jollyorc
Yes, but afaik they're all still third party, which means intentional compatibility issues from the Hue side. Also, in my particular weird-ass use case I need to be able to lock out all the lights witches every time we have a party, so if switched to those, I'd also have to move to home assistant — at least the annoying come off the wall switches let me just take them all and shove them in a drawer.
@akareilly
@dymaxion @akareilly I'm basically waiting a while for Home Assistant to have a bit nicer UI and less scripting, then I might actually take the plunge and move everything in the flat to that.
Of course, that won't solve the Sonos-Enshittification problem, but that's another can of worms...
@jollyorc
@akareilly I'll go HA once Hue is broken, but not before — I was going to have to do it to get integration with the DMX side, but now that I'm moving to MA3, I can just write Lua against the hue API library, which is actually going to be better in the end when it comes to representing what the hues can actually do.
I just chucked my Sonos pair because they were irretrievably shit. I can't actually recommend buying an Allen & Heath SQ6, the Dante card, and eight two channel Audinate Dante dongles to anyone who isn't either insane or actually running a venue, but it has solved that problem in a sufficiently final way. Of course, I still have a Wiim box hooked up to all that to get audio in 
The €15 waterproof Bluetooth speaker in the shower plus an old phone can reliably play BBC Radio Scotland and NPR without giving me ten error messages, random radio stations, and demands for subscriptions or a different app.
I do not go into negotiations with my @#&*!!! speakers.
@dymaxion @akareilly just looked at the pricing for the Dante stuff... I think my Sonos speakers will do for a while longer :D
@jollyorc
Yeah no it's entirely unreasonable. On the other hand, 1ms latency audio everywhere in the house that's (almost) as reliable as a wire.
@akareilly
@dymaxion
I have exactly these switches tho. With Zigbee.
They are in the wall behind the switch and use the original switch.
They are there. Maybe people just don't find them?
@akareilly
@Menel
Ah, right — looks like a newish release from Hue at least
@akareilly