You can also use a slightly modified version of these sub flows to mute all Alexas of which I do. I try to do everything local behind a firewall. There is just somethings like Spotify / speakers / mic that is really hard to do / very few products that work local only.
You can kind of do this with the mini-media-player and Spotify card. I guess I need to do a write up on this. It doesn't work the greatest unless you have audio groups setup. You could define a scene with the speakers and playlist you want to play as well.
Congrats! I wish we could do the same but with full time employee(s) being covered. But I'm terrible at marketing and just want to build the oss products. We have over 1k stars in multiple projects for .NET but the developer market is hard to get paying customers (we all want everything for free ;-)). It also doesn't help we have massively funded companies competing against us...
I have over 500 sensors integrated into my home and when all done will have over 1k. I find the biggest problem when you get big is that home assistant has no way to deal with back pressure of updating networked sensors. Once you have issues with say 5-6 sensors it starts cascading.. you hit 10 which could be just ecobee sensors or homekit it can bring down homeassistant.
Every room has a minimum of 7+ sensors (motion, occupancy, light sensor, nightlight, sound, mic, temp, humidity, smoke & carbon (and all related nest sensors like battery, last check, status..., path lighting (not included)). Then light switches are all smart + have occupancy + dimmer level etc... It adds up quick...
I have some crazy node red sub flows where I use state machines to turn on based on conditions and I can do whole home announcements as well as level audio and use it for custom security solution. I have smart bulbs too in some rooms and everything is 100% local with no external dependencies (if voice assistant stops working it doesn't affect anything and I'm going to probably move everything to almond). I built everything around sub flows and I made it easier to debug by adding friendly toggles that actually let me do some crazy accessibility stuff like announce motion or lights turning on etc.. If you were blind you'd know exactly where you were at.
I'm also doing local facial recognition (lots more todo) and I'm starting to integrate those into custom automations based on whose in the room (wife or me). It also greets us when we get home. It's been my dream since I was a little kid to have a smart home, and by smart I mean work for me and be quality of life improvement while being dumb for everyone (progressive enhancement). Home should alert you to dangerous weather or do the proper thing if a fire happens (open windows/ doors, turn on path lighting, announce what direction to go to the exit, etc).
If you’d be willing to open github issues for home assistant with your problems I guarantee you the dev team would want to know about and fix them. I’ve met Paulus in real life and he deeply cares about the community.
Can't speak for Azure as I don't use it, but I really have to call bullshit on AWS UI being better than GCP.
GCP is discoverable and follows the principle of least surprise quite consistently, and they show REST/CLI references for everything you're doing in the web UI too.
I mean imagine you're a new user and ask yourself "what does route53 do?". Then ask yourself "what does cloud DNS do?"
GCP's services are nicely named, but the console dashboard is still a mess, with them shifting around the items all the time. The half-foot into stackdriver is also kinda weird.
The AWS UI doesn't handle multiple subscription, it's somewhat annoying when you have multiple accounts (let say dev and prod ones). With Azure, you can federate several subscriptions.
Having no consolidated view for all regions is also somewhat annoying, as it is to search a specific resource since these can only be searched in their corresponding panel (ec2 for instances and ELB, RDS panel for DB, IAM for certificates, etc). In contrast, the Azure UI allows you to search accross all resource types in one search box.
There are other issues where I don't have a point of comparaison, but:
If you have a lot of certificates/domains, selecting the correct certificate in an ELB or a cloudfront distribution serving HTTPS is a pain. Same with selecting security groups for an RDS instance or an ELB.
The AWS UI is also painfully slow if you have a big account. The RDS panel used to timeout past 600 or 700 instances, it's a little better now (yay pagination \o/), but still quite slow (10 to 20 seconds to search for a specific instance). The EBS snapshot UI is not usable past 100000 snapshots.
I do agree that as long the API is working, the UI is not critical (don't get me started on API throttling, which are really painfull to track down and handle). It would be nice however if it worked reasonnably well, a UI is still quite efficient when you have to explore and understand what is going on.
I strongly disagree, they have a good service offering but using the ui is a horrible experience. Also azure has a CLI and a much better ui / experience esp around filtering and not being overwhelmed with a million different options / pages made by developers with no usability or design experience.
