Proton Bridge is not an SMTP interface, so this almost surely will not work. Libraries expect requests and responses to the services they speak to ensure they are working correctly.

You need an actual SMTP interface. Can’t recommend running one out on the public internet if you don’t know what you’re doing.

Don’t know why this is being downvoted. Sensible answers, but again, this is not “reasoning”, it’s all still sorting and validation. It’s not running simulations or counter balancing it’s answers based on anything.

In that case, as someone else said, you may want to take a look at some non-CMS specific style frameworks like Hugo or Jekyll who have tons of these available, but it might be a bit hard to say if you’re not super familiar with markdown for templates and such.

If you’re used to drag and drop and selectors for creating your views, may want to stick with the CMS stuff.

Might be more helpful if you mention what you’re looking for that Grav and Automad don’t cover. There’s a lot of options out there, and you seem to be looking for something pretty specific.

That may look fine from your router’s perspective, but if your network clients don’t get an updated routing tables, they won’t know or possibly accept traffic from the new subnet on the VPN.

OpenVPN does routing a tad differently, but same point applies. Network clients need to know where to go to find a route that isn’t part of the home subnet they are joined too. With containers in the mix without bridging, the host needs to know that the WG subnet can be found at the router.

Wireguard only gets you to the endpoint. You need extra routes from there.

Post your wg config, and possibly the staticnroute table on the router. What kind of OS the router is using might help as well to understand possible firewall rules being a problem.

Solution may be as simple as adding a static route to the Wireguard subnet so your other hosts can find it.

It’s certainly not capable of flashing in one step to a chip already on board. That’s the hill you’re trying to climb.

GNU projects often are sucky. Implementations of things that already exist as they happen upon them and after their uptake. Then people like you come in and defend them. 95% of the time they fail because the existing solution works fine.

I couldn’t spare an extra thought about using the GNU brand over another FOSS solution. Anyone in here stomping their feet about this is an absolute moron.

Yeah, I get that, but my points still stand.

As you can read from the code, LoRa is many frequencies, and software is not the most efficient way to deal with them. Hell, the chips now even take a ton of power in comparison to BLE or WiFi.

It’s cool, and I’m glad they did it, but there are already chips out there that do this. This work doesn’t make the chip makers any more incentivized to use this over cheap chips that already exist.

Groups of any kind are echo chambers. That’s why they exist.

Sure seems dead if you can’t get it to detect via two different routes.

Yeah, it’s super attractive until you realize the current conditions we can technologically handle that. It’s dangerous slave labor.

Good luck selling that to the era of kids working a MacDonald’s.

Depends on what you’re doing.

If you want a low-power setup, get a shell with a backplane for 4 drives and an n100 board.

If you want to host games, get an AMD APU on a mini-itx and do the same.

Maybe a 20-40W difference, but the AMD is going to outperform the n100 like a hot knife through butter.

It’s not a “surprising turn of events” as the article claims. Elon Musk is a fucking moron with money he did nothing to earn but threaten people. He threatened and fired an entire team, and LO, they were actually doing their jobs and he was mad because he continues to over promise on shit other people are doing.

More specific to this case: self-driving on cameras alone without depth perception is not possible now-if at all. That’s why Waymo was first to multiple markets with lidar systems, and their cars are not killing people, but just stupid and inefficient.

I really like the spirit of this, but the price and features are just okay considering there are other companies designing similar and better products with more flexibility and around the same price. I may pick up a board to work on it, still, but I’ll buy the Inet package with the same hardware and more Ethernet ports for $99 if given a choice.

deleted by creator

Guy who posted wants a fix. That’s the downvote.

Almost every ad platform is moving to have their ad DNS server names into the same mix as content servers. Without packet sniffing they are practically indeterminate.

Current list off the top of my head: YouTube, Netflix, Peacock, Disney, Paramount+.

It’s more costly for them, but 🤷

You have an outdated app that isn’t aware of that. Keep it until they force you to upgrade.

Also has a lot of backend options and local encryption at the client.