Home Assistant has a new foundation and a goal to become a consumer brand
arstechnica.com
arstechnica.com
- It has improved a lot over the years but it is still not polished enough for end users who want the thing to just work. Every integration works in a different way. The recommended way of managing z-wave devices changed three times since I started using it.
- Similar hubs already exist commercially and all of them failed to reach a wide audience. The Apple approach of bundling the hub with the Apple TV is a great move in my opinion, consumers are excited to get the latest shinning Apple product, none of them would care about a home hub.
At least there’s a few geeks like me who’d be excited about a new Home Assistant update.
But literally no one cares about Apple TV, a product so uninteresting that Apple couldn’t even give it a unique name.
Because `iPhone` was a super unique name...
I would argue that people are not excited about Apple TV because it does its job. There isn't a major need for new features but if they find a good feature to add to it, it can cause excitement.
The OP is right though, the Apple TV being able to be a hub (which also means it processes Secure Video and how you access your HomeKit devices from outside your home) is a nice features that shouldn't be discounted. It is a Hub that is already in many people's homes and could be a gateway product.
I guarantee you that if Apple announces a new Apple TV with some fantastic new feature, more people will be excited and buy it than there are that follow HA updates.
The Apple TV is my favorite. It’s also the one that gets used the least. Because it’s one of the older models that has the glass remote and we drop the remotes all the time so it’s only used when I’m the only one watching and I generally control it from my iPhone (only iPhone in house). And it’s so much more expensive than the others. One Apple TV costs more than the others I have, combined.
While installing it initially might be technical, these days they also sell things with it preinstalled.
After tuning on HA for the first time, it just recognized our heating system (Plugwise, it's also zigbee but custom/closed), Christmas Tree lights, Smart Reader of our Electricity & Gas (as in district "propane"), Foscam and Reolink cameras and doorbell ...etc.
All those things just asked for username/password set for that service/device and everything just worked.
And actually controlling of things like TV (well Chromecast part of it that's built into) just work without authentication.
And while Zigbee Coordinators required some flashing of firmware updates. From there on it was very similar to any other smart device system.
I don't like how it's generally advised to use an SSD, I think Pi software should be SD optimized.
A lot of third party stuff uses some cloud API that breaks constantly, not much anyone but the manufacturer can do about that.
I think their internal APIs are a bit excessive. They have different entity types for everything as opposed to basic data types with compostable metadata.
That also lends itself well to constant tweaks and breaking changes...
I don't see why a garage door needs to be any different from a light switch. One might need an extra flag to require confirmation prompts, the icons are different, but... It's just a thing that can be in two different states.
It seems like a huge amount of the code just deals with different data types.
But then, I've never made a serious attempt to use HA, I just use YoLink's native platform plus my own system for CCTV.
I know I run HA at home and its generally good, but ive not benefitted from historical data b/c I re-build it from scratch semi regularly and cant be bothered to handle backups properly :)
As we've seen with 'OpenAI' there's never a guarantee, but I applaud this step.