Free ones are less common now (no-ip went paid.) Afraiddns is still free but requires regular account logins.

If anyone goes with finamp, sign up for the test version as it’s UI is significantly better than stable.

You need the web site to use a certificate from the same root authority as your client certificate. Otherwise browsers won’t present the certificate to the server. That means either warnings on connect or adding the root cert.

I do think if you are doing it with them in person it is doable to add it.

I recently did exactly this. Only works with the web UI, no apps support it, but working well and those without the cert just get a 400 error. Not sure if non technical tbh, since you will get warnings when adding your root certificates to any device, and that might scare some who don’t understand what it does.

Also set it up through wireguard, so can punch out of double NAT.

I’ve not used that software before, but a quick look at the guide makes me wonder if you have a https redirect in the nginx config. If that is hard coded to domain1 it will always redirect to it. Update it on the new config file to point to the right uri if that is the case.

Usually I setup with a single domain for the server, and use that for web mail, IMAP and SMTP certs. That way you don’t need to worry about the extra certificates needed for each domain (set your Mx to use the common name of your cert.)

Yea I didn’t think the post was that professional. Also the “unminified” version is just the minified with more white space. It still has poor names and no explanation of the binary blob.

Forked from gitea. The owners of the project implemented a change to the pr system where by you had the sign an agreement that the code belonged to them. This was seen as an intent to relisense at some point. Devs that wanted it to stay open moved to forgejo.

Until recently you could swap forgejo in to your gitea data, but now they are incompatible due to divergence.

Recommendation would be that you want to set up your ssh so that it only accepts publickey authentication. You also want to make sure you are not using a proxied DNS value, as CF only proxies http requests.

Personally I didn’t bother to setup ssh access as https typically works fine.

If you are talking about the home page, I blocked the container for “featured content” using ublock, that kills short and news in the feed. If you get linked to a short, you can use a extension to redirect the URL to the normal watch player.

Even if you have a proper clean IP, running a mail server is a hassle imo. By far having a single relay to send is fine if you get things set right, but also dealing with incoming spam is just way more work than paying to have it hosted.

I much prefer paying for email hosting and just dealing with outgoing emails if needed.

The number of people that are anti wires, but also want WiFi signals everywhere is too high.

An easy option is to add the data folders for the container you are using as a volume mapped to a local folder. Then the container will just put the files there and you can backup the folder. Restore is just put the files back there, then make sure you set the same volume mapping so the container already sees them.

You can also use the same method to access the db directory for the migration. Typically for databases you want to make sure the container is stopped before doing anything with those files.

Pretty much, add the new domain to certbot, then create a new server directive with the new name and certificate paths. I would probably just copy and paste the existing one for your current domain, then just change the domain. SNI will then select the right certificate for you.

You can set certbot to create a single certificate with both domains, but tbh the above is just simpler to setup and maintain.

The first half sounds so true to me. Like it was an intern that really wanted a replica set, but instead of using the same platform the company was using, hacked together something running on Linux. Ofc they didn’t tell anyone how it worked, and everyone else knew windows server so no one poked it.

It used to be running on a spare pentium 4, but was virtualized as no one knew why things stopped working when it was turned off

It’s less of a main, and more of a “don’t do this if being imported.” You can just throw code without that block and it will run.

Perhaps she only used an extra core, GB or so on her friends’ servers?

Which is fine when people do not reject the answers that are different from what they were expecting. Learning that the problem you have is a reason that noone does this, is a valid thing to learn.

It’s usually when I see people moving the goal posts on replies, or complaining that they didn’t answer the exact question that i see as frustrating. Or “I don’t want to do that” with no more info.

But if you are aware of other solutions, you should state that in the question and give your reasons. It’s a waste of time if you know someone might suggest what you have dismissed already.

The html question is a classic for this, they want to find non self closed tags. Why? Why can’t they use a parser? What are they doing with this info? All questions that would give you a good idea on how the problem can be solved. Playing with regex would be a valid answer to that, but is not stated. Unfortunately I find so’s format discourages extra interrogation.

The answer is not an attack on the person, but a frustration at the people before that ignored previous answers to use a parser.

Except in 99% of cases the person is asking an xy problem, and if they ever explained the why, they would get a proper answer.

Often the reason no one does the hyper-specific thing, is that there are better non code solutions, it’s massively insecure, or is just stupid micromanaging.

If you don’t mind having email go through Gmail etc, then you might not want to full host, but just run a local IMAP server. There are some pop to SMTP solutions you can use to pull your emails (fetchmail.) you can then use your account as an outbound relay. Keep in mind you’ll only be able to set this up for a single account if you use something like Gmail.

If you buy a business product like workspace or m365, you should be able to setup relay/hybrid connectors for multiple accounts.

Since you added a question mark, commands is the correct general term. However there are two types that can be a command. Functions: which are written in pure powershell and cmdlets: which are commands provided by dotnet classes. (Also exes and a bunch of other stuff common to other shells can be a command, but that’s not important.)

The reason they have different names is early on functions didn’t support some of the features available to cmdlets, such as pipeline input. There was later a way to add this support to functions.

In practice call them any of the 3 and people will know that you mean.