The problem could be anywhere in between the internet and your server.

Ofc. it could be your routet. But I think the following is more likely:

It might also be your internet service provider that doesn’t allow those ports for inbound connections.

Or you’re behind a CGNAT so your real external ip is different from the one you think it is. (look up online how to test this)

Sounds good.

Hmm next you probably should confirm ports 80 and 443 are actually reachable from the internet.

Use an online port checker like https://canyouseeme.org/

After that you should check your apache config like somebody else already suggested. I haven’t used apache in a while but if I remember correctly:

Ensure it says: Listen 80 NOT: Listen 127.0.0.1:80

(and same with 443)

Also check your VirtualHost — it should look something like:

<VirtualHost *:80>
    ServerName yourdomain.com
    DocumentRoot /var/www/wordpress
    # ... other settings
</VirtualHost>

(and same with 443)

Njalla’s default TTL for DNS records is 3600 seconds (1 hour). If you just created or modified the A record, it can take up to that full hour for the change to propagate across the internet, which would perfectly explain why Certbot is connecting to the right IP but failing to fetch the file (the request might be hitting an old IP or a cached null response).

Before changing any more configurations, you should verify what the rest of the internet is actually seeing for your domain right now.

Check the current DNS record

You can usedig to see exactly what IP your domain is resolving to, and importantly, the remaining TTL on that record.

From your local machine (or any computer), run:

dig yourdomain.com +noall +answer

This will output something like:

yourdomain.com.    3412    IN      A       203.0.113.45

The second column (3412) is the remaining TTL in seconds. If that number is counting down from 3600, the record is still propagating. If the IP address shown there doesn’t match your server’s current public IP, the change hasn’t taken effect yet for that DNS server.

Check from a different perspective

To ensure it’s not just your local ISP or router cache serving an old record, query an external public DNS server directly:

dig yourdomain.com @1.1.1.1 +noall +answer
dig yourdomain.com @8.8.8.8 +noall +answer

If these external servers show the correct IP but Certbot still fails, the DNS is fine, and the problem is somewhere in your network routing or web server config. If they show a wrong IP or no record at all, you simply need to wait for the TTL to expire.

https://strudel.cc/

nothing to do with guitar, but this is my favourite thing to toy around with.

I discovered it through this video https://www.youtube.com/watch?v=iu5rnQkfO6M

Are you talking about the “dumang ergo v2” or the “keyboardio model 100”?

In either case: The home keys are right under my fingertips when they are in their natural position.

The concept reminds me of my Dumang DK6 ergo v2 that I’ve stashed away and never use. Same reason as you, it’s just less ergonomic than my keyboard.io even though it’s more customizable (you can place every individual key anywhere on the baseplate in any orientation)

There is no “stealing” when we talk about digital goods, as copying files is virtually free of cost (just a littld bit of electricity).

So if the file gets duplicated and you get the copy you stole nothing.

Here you go:

Everything should be documented there

https://git.sr.ht/~deckweiss/x11_activity_session

It probably can be adopted to wayland (or even native kde tiling) with a bit of tinkering.

KDE activities don’t get much dev love, so be warned.

The activity script hooks are (or were for a long time, idk how it is now) an undocumented feature.

I’ll dig all the stuff out for you later today.

I tried for fun and noticed that half the websites I visit don’t have an ipv6 DNS record

it’s like with ipv4

If you are on linux, and want ai assisted stuff like you mentioned there has been this for a while: https://github.com/qwersyk/Newelle

( or the weeb version if you prefer: https://wiki.nyarchlinux.moe/nyarchassistant/ )

and it can use locally run models. But have realistic expectations. If you want it to work well, you need a beefy GPU, a lot of RAM and swap. The “intelligence” is kind of limited if you run low spec models, to the point of it maybe being utterly useless.

I searched around a bit and found https://github.com/dreiekk/calcdav

not sure if it still works, last commit was 3 years ago (on the other hand, there was no AI vibecoding back then, so thats a plus), but it looks like it might do exactly what you’re looking for

there’s prep and glam I like to do or tend to set up but that I would prefer to explicitly set up instead of it being done automaticall

Back in the x11 days I had a script that would take a config file and open multiple programs in a specified arrangement across my displays.

I used KDE activities by task and had such a config for each task. KDE activities can run arbitrary scripts on being started. So when I opened the “work” activity for example, all my work apps would open up in my preffered arrangement. When I opened the gaming activity, steam would start on my side monitor and the main monitor had all of the other gaming related shortcuts on it etc.

Together with the preload daemon or a custom vm-touch (i switched from one to the other at some point) it was blazingly fast and very comfy. (Again, I overprovisioned my RAM so I used it by filling it post boot with a cache of pages that my apps load on startup)

Then wayland came and broke it and I didn’t bother to fix it yet.

But everybody has their own prefered workflows, I’m not saying one is better than the other. Just wanted to share.

This is great, but I like my keyboard.io

Sorry, I can’t help you with your problem.

But just in case you were serious about “We don’t shutdown.”:

In my case - clean boot takes 25s. Waking up from hibernation takes over 60 seconds, because of huge RAM. And sleep is broken due to some USB interface shenanigans. Soooo yeaaah, I fully shut down and power on every day.

Oh and btw. by default windows doesn’t do a full shutdown, but a sneaky hibernate. You can see that for example if you “shutdown” windows (not reboot.), then power on the pc and boot into linux - trying to access the windows drive, you will see an error that windows “didn’t shutdown properly” and is still claiming access to the windows drive. Because it’s hibernating and changing content on the drive might break the wakeup.

I think people are sleeping on GLM.

Tried it out recently and I like the results a lot so far.

GLM4.5 and 4.7 was good already, now they released 5 and 5.1 https://github.com/zai-org/GLM-5

It says it’s for vibecoding but I use it like I would use chatgpt and it gives useable ansers to all of my varied questions. (ofc. you always have to check for correctness, even if it’s correct most of the time, which I do cause I’m paranoid)

I guess the only downside is how frigging huge it is.

We’re not.

Huh? You linked me an article where server cost is the lions share of signal operation.

Most users doesn’t even donate 1€ when using free messengers.

How does signal operate then.

They don’t offer ANY “nice-to-have” features

If they don’t have high server costs, unlike the example from the article you brought up, they should hire cheaper software engineers from a different country or scale down development and have a longer runway.

Like I said - them having this problem is probably due to poor planing.

Well, when talking about server costs, Threema somehow has been running on a 5€ lifetime license and business customer subscribtions for over a decade.

While briar and simplex are peer to peer and have nearly no ops costs.

Sure, it can be made to be very expensive, but I’m arguing that doing so is a business/design decision.

Servers can help improve the UX, but are expensive. Threema for example, only stores media on their servers temporarely, so they have way lower storage cost with a small tradeoff in userfriendlyness (of having to migratethe old media files you want to keep when you get a new phone). And so on.

If your nonprofit only has 65k, don’t hire multiple devs and provide nice-to-have features that lead to high ops expenses in servers and storage. It’s called minimal viable prpduct for a reason.

Sounds like bad planning. There are like 3 other e2ee messengers that are open source and have enough funding to operate for years without doing appeal-to-emotion dona-… extortion campaigns