OK that is fair, though that is not self hosted…

VPS machines are a completely different beast than self hosting. But I guess I only said home use, not specifically self-hosting though we are in a self-hosted community. There are 1000 guides for setting up a VPN on your home network.

To be fair, if something is open by default or very easy to enable without informing about the risks, tons of people will have it exposed without thinking.

It isn’t that “tons of people do it so it is normal and perfectly fine” but more “people don’t realize.” It also uses some nontrivial amount of resources to process and block those attempts, even if they never have a chance of getting in.

There is yet a reason I can find to have it forwarded for home use. Need to ssh into a machine to fix it? VPN.

There are plenty of secure web-based tools to manage your server without a VPN also.

Does inkscape have diagram connecting? One of the best draw.io features is the wide array of premade shapes, styles, and auto connecting for flow visualization

You definitely don’t use CUDA then. That is hardware accelerated machine learning basically.

For you usecase then it doesn’t make much of a difference. DLSS 3.0 is indeed better than FSR, but there are few games that use it i guess. DLSS 2.x and FSR are about on par with each other and FSR is enabled in all games. . Many/most of people don’t even realize that DLSS/FSR is disabled when gaming as the vast vast majority of games don’t even have it and and most don’t think about it, I have no idea if you are in the same boat, but then it makes no sense to base a decision based off of features you don’t use, in my opinion.

Games simply don’t benefit enough for the cost of a new processor, let alone new motherboard and ram.

A new GPU will almost always the best bang for your buck improvement in games.

Then you should definitely go AMD. There is literally no reason not to unless you are already using cuda or ray tracing a ton. AMD is the best value for the money by far, has a MUCH better software interface (never thought I would say that), comparable or less driver issues than nvidia now, and it also works flawlessly on Linux, including full undervolting support (important on any GPU, but on AMD it is much easier).

That being said, if comparable performance GPUs are the same price in your region and you use windows, nvidia is also fine to grab.

Always undervolt your GPU. My 5700XT that ran on 200W before now maxes out at 150W and usually is at 140W with a 1% performance difference. That is like a 9C temp difference.

Hey, I developed some of these for a different use case at my job.

Generally they use BLE servers to query and update every BLE device one after the other.

I would venture to guess that these tags have a ublox BMD-3xx variant controlling them and you can connect via Bluetooth low energy every X hours.

If you just want the displays they use, here are some kits:

https://www.tinytronics.nl/nl/displays/e-ink

You want a prebuilt NUC. That is pretty much it. A Pi generally doesn’t cut it, plus by the time you get all the accessories to make it usable as a server, you are at the price of a much more powerful small PC.

I run a ryzen 3600 with 32GB RAM in a Node 304 and it is very quiet with 12TB red pro helium drives. 4 HDD capable with a GPU and 6 without. However, it still draws 20-30W idle or 50W with an Arc A380 installed.

There is one neat trick: don’t expose SSH.

There is still not a reason anyone has been able to give for 99% of self-hosters to expose SSH.

If you need to access your machine via ssh while on the go. Wireguard to your local network, use SSH. Done. Unless you are running an always-up public facing site, the amount of times you have to access your machine that can’t wait until after work is very low anyway.

Bots will scan all ports. That is just how it works. Less than 22, but you will still get spammed. Why force your computer to go through the fail2ban loop and take up resources when it is simply not needed at all and you can block it on another machine?

If you want to build it yourself, you have to decide on size.

Are you trying to keep it as small as possible?

Do you want a dedicated GPU for multiple jellyfin streams? (Definitely get the Intel A380, cheap and an encoding beast)

If you don’t want to start a rack and don’t want to go with a prebuilt NUC, there are 2 PC cases I would recommend.

Node 304 and Node 804.

Node 304 is mini-ITX (1 PCIe slot, 1 M.2 slot for boot OS, 4 HDDs, SFX-L PSU, and great cooling)

Node 804 is micro-ATX (2 PCIe slots, 2 M.2 slots, 8-10 HDDs, ATX PSU, and 2 chambers for the HDDs to stay cool)

Why do you want a N100? Is electricity very expensive where you are that idle power is a big factor? Because desktop CPUs are more powerful and the CPUs can idle down to 10W or so without a GPU and they can have way more RAM.

Tldr; go with prebuilt NUC or go with a desktop CPU for a custom build.

I have no extra drive of the capacity of my system. Maybe I should just go with hetzner or backblaze

At that point, you might as well get Mealie.

• Recipe manager (with online recipe parsing so you don’t have to read everyone’s life story)
• Equipment and ingredient lists per recipe
• Meal planner so you can plan out a week of meals
• Shopping lists by adding item by item or actually linking recipes and automatically importing all of the items
• Different users and access control, OIDC, backups, and most modern features.

shopping list demo

I am making an esphome system for my old Yamaha AV receiver. RCA out via a cheap i2s chip, mems I2S mic in the enclosure that can hear well enough in the room, and an IR blaster output to change the AV receiver channel to the esp when voice is activated and change it back after.

Cool side effect is that it can also be used as an HA media player for streaming Spotify and local music to the speaker system.

Sorry, I misunderstood what you are trying to do here. I thought you were trying to use the Atom Echo itself as a media player. Disregard that arduino library comment, it isn’t relevant. I just watched the video since I couldn’t earlier.

Indeed what you are doing should work. Are you certain that the upload was successful? With GPIO21 set as the speaker output, the speaker data should absolutely not work. The fact that it does means that somewhere along the line, the GPIO22 is set as the speaker output.

on_tts_end: you have a media player component while you define a speaker instead. They are not interchangeable. It is likely trying to grab default values from somewhere because of that. Media player is better if you want the device to also play music or alerts through home assistant instead of voice assistant or some preset wav files.

Media player is also a speaker using an arduino library (not compatible with esp_adf as that uses the esp-idf framework and not arduino). If you want to use the media player, you have to get rid of vad_threshold and the esp_adf.

Please follow this yaml for an esp-idf config using the speaker instead of the media player: https://github.com/esphome/firmware/blob/main/voice-assistant/m5stack-atom-echo.yaml

Then use Wireguard to get into your local network. Simple as. All security risks that don’t need to be accessed by the public (document servers, ssh, internal tools, etc…) can be accessed via VPN while the port forwarded servers are behind a reverse proxy, TLS, and an authentication layer like Authelia/authentik for things that only a small group needs to access.

Sorry, but there is 1 case in 10000 where a home user would have to have publicly exposed SSH and 9999 cases of 10000 where it is not needed at all and would only be done out of laziness or lack of knowledge of options.

Woah, let’s not be hasty. A few big tech companies are really good at their jobs…

Let’s not forget the dozens of big tech companies run by absolute morons that bring products that nobody wants or needs and only stay afloat due to legacy, stealing data & selling it, and/or venture capital.

I love the absurdity of game reviewers 😂

• “This game is the pinnicle of its genre”

• “This is the 1 game I would bring to a desert island”

• “one of the most captivating puzzle games ever”

85%, 88%, 70%, a C to a B. That is just above average.

Meanwhile you get an absolute broken AAA piece of crap that barely functions, incoherant story, generic and boring and those same reviewers say “70-80%”. So there is a <10% difference between absolutely mastering a genre and releasing straight garbage?

Do people know if he is using one of the yaml configs on github or has his own?

I am currently developing a PoE ESP-C3 bridge for the LD2410 and 2450 (don’t know if I will sell it or just have all of the files available to make yourself) do I am doing a bit of research. When I make mine open source anyway, I guess the license will be similar to the skreek version anyway so I can work off of his.

Definitely Red plus. They are quiet as hell and 12TB+ are helium filled.

Just got a 12TB a while ago and it is as quiet as my 4TB drives.

But for OP, just use software raid instead of hardware raid. There is very little point for homelabbers using hardware raid at this point without an existing setup.

Leantime for sure! Because it is very feature complete for project management