

And I’m still here with a 1080p TV. Because my previous ISP had data caps, I’m not convinced my current ISP won’t move to data caps, and there isn’t a lot of good 1080p content out there besides DVD/BluRay and my Rips, let alone 5k or 8k content.


And I’m still here with a 1080p TV. Because my previous ISP had data caps, I’m not convinced my current ISP won’t move to data caps, and there isn’t a lot of good 1080p content out there besides DVD/BluRay and my Rips, let alone 5k or 8k content.


For the Home Assistant install, I would suggest a mini-PC. I run mine on an old Intel NUC. The mini-PC will give you more flexibility as you add and expand automation. I had read of Raspberry Pis bogging down under load (this was pre version 4 and 5). But even the NUC bogs down with local voice assistant use. I’m thinking of building my own, local LLM computer to offload voice. Eventually.
As far as other hardware, I’d go with the ZBT-1 (formally called SkyConnect) for Thread support, which is the new kid on the protocol block and we’re starting to see a good set of devices.
After that you’ll have to make a choice between Zigbee and Z-Wave, and buy a dongle for one of them to attach to HA to get the functionality. Both work solidly. Z-Wave devices tend to be more expensive because of the rigorous testing Z-Wave certification requires. Zigbee is cheaper, but it works on the 2.4Ghz spectrum, same as Wifi. It can lead to some interference if you still have a lot of old Wifi devices. You’ll have to look and see what kind of devices you want to run, what they cost, and decide from there. Personally, I went Z-Wave and most of my light switches are Z-Wave. it’s been solid but pricey. Hoping to see more Matter/Thread switches soon.
You’ll see a lot of Matter talk. Matter comes in two flavors, Matter/Wifi and Matter/Thread. Matter is a framework for devices to talk to automation hubs to tell them what functionality/sensors are available, Wifi/Thread is the protocol it communicates to the hub and other devices. I look for Thread devices since it is more like Zigbee/Z-Wave (Thread from what I understand is based on Zigbee). It builds an internal mesh network and keeps everything local, and it can’t communicate over the internet by itself.
If you don’t mind a bit of tinkering, I can recommend the Home Assistant Voice Preview devices. It is a voice device that talks to HA to be the voice front end. It is not nearly as polished as Google Voice or Alexa, but since I regulated Google Voice to basically be a voice front end to HA, it has been a decent replacement.


I’ll second this. I have mine on an old NUC and it has been great.


One thing has nothing to do with the other. Just more, “Get a horse,” anti-AI energy.
I thought Govee2MQTT was for the non-Matter lights?


Meh, lots of them earned it.


People here actually believe companies should just accept fines and their execs accept jail on their behalf rather than the company comply with laws they don’t like.


Actually, there are doctors that want AI. Just they may not want it like most tech firms are currently pushing it. Each hospital system has a huge database or two or three with massive amounts of patient data. Doctors have talked about setting up data scientists to sort through that data for more effective outcomes to various health issues. It turns out it is too much data for a team of data scientists to sort through. AI might help with that. Just not the LLMs that are being pushed today.
Some of the challenges: How do you pull that data without personal identification or payment info. Keep in mind John Doe in Somewhere, somestate might be the only person in that state with Obscure Condition, so would be easily identifiable. Because once you have data that may support better outcomes, you’d definitely want to share that with other healthcare systems and government health agencies. Also, how do you use it ethically, something none of the current mainstream AI companies are really going to help you with. How do you share this with insurance companies without them punishing individual patients?


But, but, but, my science fiction reading says all AI is trying to kill us!!!
There is a lot of, “Get a horse!” out there.


Pixel phones are inferior to the Nexus phones. Google wanted in on the $1000 phone market and killed the best affordable phone line they ever offered to do it.


Problem is, it won’t stop with 4chan.


The fact this is the cheapest EV is entirely the problem.
If you had to wade through all the arguments in the Linux kernal “discussions” you’d be a jerk too.


I wonder which will kill you first?
Not the individual apps, no. Figured it might be something more systemic with Flatpak itself or my system. Was expecting to be pointed to logs to investigate.
The Deb installs mostly. Example, when I installed Lutris, it wouldn’t allow me to install games on the extra NVMe drive I have in the system. I uninstalled Lutris and installed the deb version from the Pop Shop (when you COULD do that) and it allowed me to install games where I liked. Just chalked it up to a Flatpak problem. Now I have Lutris installed, use it without issues, and I’m not going to try and solve Flatpak issues by re-installing it. It’s not worth my time.
Currently, any flatpak apps that don’t work properly are installed with other means, so I don’t have any non-working apps . If I run into that again, I’ll give it a try.
No use using flatpak if the installed apps won’t run properly. Be nice if they did, but they don’t.
Meta is training sexbots, isn’t it?