Have you heard of Homeassistant? It’s a self-hosted smart home solution that fills a lot of the gaps left by the most smart home tech. They’ve recently added and refined support for various different voice assistants, some of which run completely on your hardware. I have found they have great community support for this project and you can also buy their hardware if you don’t feel like tinkering on a Raspberry Pi or VM. The best thing (IMHO) about Homeassistant is that it is FOSS.
badlotus
Here’s an article with a bit more detail… but I’m still unclear whether these backdoor commands are hardware circuits or firmware logic.
Bleeping Computer: Undocumented "backdoor" found in Bluetooth chip used by a billion devices
Debian is one of the largest GNU/Linux variants out there. Who and what you trust are personal decisions but they’ve got a good reputation.
I’ve run across a couple sites.
https://www.opensourcealternative.to/
https://github.com/sereneblue/awesome-oss
https://opensourcesoftwaredirectory.com/
https://opensource.com/resources
https://www.techradar.com/best/best-open-source-software
Tangentially related: https://github.com/awesome-selfhosted/awesome-selfhosted
Pretty much the opposite of what you asked for but good to know: https://en.wikipedia.org/wiki/List_of_formerly_open-source_or_free_software
@chirospasm@lemmy.ml beat me to the punch with alternative.to
There are actually new RAM interfaces being developed. One I have heard of is CAMM2. This new interface improves on the performance that can be achieved with current DIMM interfaces by about 50%. Here’s an article from Corsair with more info: https://www.corsair.com/us/en/explorer/diy-builder/memory/what-are-camm2-and-lpcamm2/
It would have been nice to see this added to the new Framework desktop. The main argument against this would be cost/availability. New standards bring improved performance but may raise the price point of the computer to an unmarketable level. Also, Framework would have to source these modules, which are new and still niche. They would also have to adjust their motherboard design to accommodate the new interface which could introduce additional design considerations or cost.
Some of this was written by ChatGPT. Good eye.
Think of Docker containers like lightweight, portable mini-computers that run on your actual computer (the host). Each container has everything it needs to run an application—like code, libraries, and dependencies—but it shares the host’s OS kernel rather than running a full OS itself.
Containers vs. the Host System
• Not a full OS: Containers don’t have their own separate OS but use the host’s OS kernel. They do, however, have their own filesystem and isolated environment.
• Like a sandboxed app: A container is more like a self-contained app that has just enough system components to run but doesn’t affect the rest of your system.
Keeping Containers Updated
You do need to update containers separately—updating the host system isn’t enough. Here’s why:
-
Containers use images: Containers are created from images (like templates). If the image gets outdated, the container running from it will also be outdated.
-
Rebuilding is required: You can’t “patch” a running container like a normal program. Instead, you must:
• Pull the latest version of the image (docker pull my-image:latest).
• Stop and remove the old container (docker stop my-container && docker rm my-container).
• Start a new container with the updated image (docker run -d --name my-container my-image:latest).
Automating Updates
To simplify updates:
• Use a container management tool like Docker Compose, Portianer, or Kubernetes.
• Watch for updates to base images (docker images to list images and docker pull to update).
• Set up an automated pipeline to rebuild and deploy updated containers. There are tools like Watchtower that will automate this with minimal effort.
In short: Updating the host OS won’t update your containers. You need to rebuild and restart containers with updated images to keep them secure and up-to-date.
Note for comments below: If you are trying to customize a docker image, you must build a new image. This is done through “dockerfiles” that instruct the docker engine what commands to run on a base image to create a custom image. For instance, one could take a simple Linux image like Alpine and use a docker file to install NGINX and make an NGINX image to create a reverse proxy container. In many cases you can find images that have been published that meet most basic needs so building images is often only necessary for advanced docker implementations that require special customization.
OpenWebUI supports allowing web searches when submitting prompts to your self-hosted models. Usually I see three web searches per prompt if I enable this feature.
$4.39-$10.00 (USD) in Phoenix, AZ
New shoes. My previous pair developed holes behind the toe guard so I got a warranty replacement. My new pair is much lighter, softer, and a better design.
Have you heard of Ollama? It’s an LLM engine that you can run at home. The speed, model size, context length, etc. that you can achieve really depends on your hardware. I’m using a low-mid graphics card and 32GB of RAM and get decent performance. Not lightning quick like ChatGPT but fine for simple tasks.
Ollama