In my mind they are different:
API tests run the entire app locally (with some dependencies mocked out) and call the API over HTTP to test the entire flow.
The synthetic monitoring calls APIs in prod.

But maybe what I call API testing isn't the same as what you call API testing 😄

My Factorio experience is a bit like that: huge spikes and then nothing for months/years. At this point I'm afraid to open the game because I know my free time will just disappear 😄

I have no intentions of building this, but I find it fun to explore some of these ideas to see how they might (or not) work 🤔

The widespread use of AI in coding also makes this tricky, since your code nowadays will likely include a little (or a lot) of code generated by AI, also defeating the point of people "choosing" you because they like your code.

I don't have a good answer for this one. AI to filter out AI code? 😆

You could write code directly into the app, but that somewhat defeats the purpose of it being your "true self", since you would likely polish the code a lot more than you would if you were writing it for an actual project.

By pulling random code from your account I'd hope to show the real you.

One challenge would be making sure it's actually your code, and not either something someone else contributed to your repo, or a random repo you just copied and pushed.

Maybe using code from PRs would mitigate this a little?

Or a more complex system to detect and exclude copied code?

I hope your Saturday improves!

But I suppose it's a good reminder to set up uptime monitoring/alerts on important services, specially self-managed ones. I know I've been putting it off for a long time 😬

The Github issue mentions turning off Cloudflare auto minify, but I believe that has been deprecated 🤔

In my case it was my own fault - I added a build step to minify the HTML back when I was using 11ty and I kept it after the rewrite to Svelte. I never thought it would be an issue 🤷

Ok, I might give it a try one of these days!

As for the script, the core of it is:
HETZNER_API_KEY="xxx" lego --email [email protected] --dns hetzner --domains "*.domain.com" renew

The rest of the script just runs this for a few different domains and moves the certs to the right folder 😀

Yup, just regular crontab to run a 10 line bash script 😀

The lego CLI does all the hard stuff to get the actual certificate. I just need to run the right command and provide credentials (since I'm using a DNS-based challenge to verify domain ownership)

My certificate setup may not be as simple, but I'm happy with it

I have a simple cronjob once a week to renew wildcard certs for my domains using Lego (github.com/go-acme/lego). Then I run a Traefik proxy for all my containers that uses those certs. Haven't thought about it much since I set it up!

I've been using Portainer for years now and I'm happy with it. Coolify looks tempting, but I don't have a real reason to switch at the moment.

First time I've heard of Dokploy though 🤔

That looks nice! Thanks for sharing, I haven't come across that before. Might give it a try!

I think it depends, but one big one will be a lot easier to manage and takes a way a lot of complexity. So I would generally go for a single one until I really *need* the benefits that multiple small ones could bring (uptime, scaling, etc)

I prefer your version to be honest.
Personally I thought it was pronounced more like "veet" 😄

Maybe I've just got used to it, but personally I prefer Gitlab's UI over Github 🤷

I agree with the main point though, nothing wrong with an open source project copying a familiar layout!