SwizzleStick

Honestly have not bothered too much on the internal security side. Everything is in a melting pot on the same subnet, with pfsense managing what’s allowed out. At the very least, the cams and any other accessible internal devices do not run default/duplicated credentials.

Only two users on the network, and the occasional trusted guest. I don’t see the need to go further quite yet.

Hi fellow HA user in the wild.

Whatever cams you go with, most important thing is that they support a direct rtsp connection. Frigate is an excellent add-on for recording, and for object detection if you want to do that.

We have some generic IP cameras here that have local access only, and a couple of Arduino camera PCBs in printed housings.

A coral TPU is essential if you want to get into object detection on more than one camera. Can use the CPU for testing, but it’s very easy to tap it out.

Use a different service, or encrypt your data before upload & share password separately. For anything remotely private, should be doing that already.

For chaotic good, upload LLM poison to fuck with the training data.

Oh nee :((

It’s a Synology NAS so will work with the Synology DSM integration: https://www.home-assistant.io/integrations/synology_dsm

There’s definitely ways to poll the sensors of other devices though. I had some janky sensors set up before for monitoring a standard Linux box.

We’re doing fine here:

Aber Deutschland ist geschmolzen :(

Whatever way you go for setting up the systems themselves, I’ve found dwservice.net to be perfect for accessing systems with only a browser.

The host component is Mac, Windows and Linux compatible. The clients need only an account at DW. Hosts tied to your own account can be shared with others.

Depending on host OS, you get screen, terminal and fire transfer access. Sessions are logged if you need to review who’s accessed what.

Free. Donation optional.

I condensed down from a power hungry tower server to a couple of thinkstations and a nas. Much nicer on the power.

It’s better to name known safe options rather than leave it up to user search. The entities that work against extensions like uBO are already well aware of their existence, so hiding their names has no benefit.

Case in point - uBlock and uBlock Origin are not the same, with the former being a bastardised version that does ‘acceptable ads’. There are plenty of other poor blocking options out there for the unsuspecting to stumble into besides that.

Personal setup is Librewolf/uBO on the client and pfBlockerNG/Snort for network level blocking/additional security layer.

And welcome to .zip :) Hope you enjoy the new home!

Well we can do you the dragon at least: 🏴󠁧󠁢󠁷󠁬󠁳󠁿

🧠 + a few slapdash notes in a password manager. It’s very organic, very human.

Occasionally leads to situations like this.

At that point I would expect control of it, or at least for it to respect the configuration it is given. If neither are true, then it just doesn’t go online at all. If that’s part of the main function, then I find an alternative or live without it.

Nothing on the inside should be sending anything to the outside that can’t be inspected before it leaves, with the exception of stuff that is directly driven by a human (guests browsing, etc).

This is the best way, really. Generally, you have much more control over what you plug into it.

A display shouldn’t have anything even approaching what can be called an ‘OS’ on it. Yet here we are.

Sometimes even that’s not enough. I’ve had some questionable kit before that would just ignore the DNS settings fed to it if it thought they were no good, and fall back to something else preconfigured.

pfSense is a wonderful tool for situations like that. Anything intended for local use only here just doesn’t get outside at all. Handy for stuff like a fire stick that only needs to be calling up a local media library.

It can also mangle any DNS requests going out to a different server and redirect them to itself instead. You could do this without it with iptables/nftables on a generic Linux box, but pfSense makes it much friendlier.

There are other packages that can do the same, but physically all you need is one piece of hardware as a bouncer that manages connections between inside/outside.

On Dell server hardware with the right cards/licensing, you can remove the need for physical access to the server to input an FDE password by leaning on iDRAC. This provides access to the console remotely during the boot process (and thereafter).

Alternatives exist that supposedly do the same thing, but I’ve never had to try them. Airconsole, pikvm, blikvm etc.

You can keep this interface unexposed by using wireguard to dial in when you’re away, as per your original thinking. Just make sure the endpoint isn’t on the server you’re rebooting…

Half the shit I actually want I just run directly these days, rather than nosing through either.

• ncpa.cpl
• diskmgmt.msc
• devmgmt.msc
• control userpasswords2
• cmd
• mstsc
• regedit
• taskmgr

Just to name a few.

It’s utter bollocks. It used to be the OEM crap that had to be removed or clean installed over. Now you have to spend time unfucking fresh installs.

My 11 image is just about usable, but only after a lot of gutting, reg entries, powershell scripts and openshell.

The railroading to sign in with an MS account has become worse too, but still just about bypassable.

Not out of the goodness of their own hearts mind. It’s probably more because Euro NCAP are going to be deducting score for not having physical essentials in 2026.

Used FF forever, even though the birth and rise of Chrome.

We’re done. The company I IT for therefore is also done. As are friends and family I sort computers for.

The shit now stinks and must be taken out.