Recommended talk: Digital Sovereignty Is the New Influencer Status

This week I came across a post recommending the talk from Molly White (@molly0xff) at SXSW, talking about “Digital Sovereignty Is the New Influencer Status“.

The abstract of the talk is: “The real power move for creators is ownership and control of their work and livelihoods. This freedom is actually closer for more people than ever before! Learn from two fediverse futurists how decentralized social media facilitates better publishing, community, and business models. These speakers will cover the theoretical and the practical, including how the landscape is evolving in 2025, where creators should focus their energies, and how they might thrive with or without “walled garden” social platforms.”

It’s an interesting talk, where she’s talking about how (and WHY) she keeps her content mainly on her website and federates it to some social networks for discovery.

How to set up Navidrome for MP3 Streaming on a RaspberryPi

Recently I noticed that I hardly ever listen my MP3 music collection – mainly because I don’t like my music players (both on mobile and on the desktop). So, let’s look for a nicer player!

I had only two constraints: 1) it should be able to use the files from OneDrive and 2) it should be able to cache them (in case of no connectivity). And as I went through the candidates, I stumbled across Navidrome:

Navidrome allows you to enjoy your music collection from anywhere, by making it available through a modern Web UI and through a wide range of third-party compatible mobile apps, for both iOS and Android devices.

Navidrome also supports Playlists and Internet-Radio! Nice. So why not let it run on a RaspberryPi at home and make it available through my VPN? But would I have enough RAM left? I just tried!

Continue reading How to set up Navidrome for MP3 Streaming on a RaspberryPi

Casino Data Jackpot – For Hackers: Merkur’s API Disaster

A couple of days ago, I saw a Mastodon post from Lilith Wittmann in my timeline. She linked to an article on her Medium page detailing a catastrophic security failure at Merkur AG. You can find the original Mastodon post here.

The casino company Merkur AG and its service providers have made almost all the data available in their casino systems publicly accessible. This includes payment data, gaming sessions, and copies of the ID cards of over one million players.

Lilith Wittmann’s Medium Post (German)
Continue reading Casino Data Jackpot – For Hackers: Merkur’s API Disaster

How MS Edge’s Immersive Reader Helps Me Slow Down

We all probably know the drill of a typical workday: back-to-back meetings, side conversations in team chats about some other topics, drafting & scanning emails, creating Jira issues, and juggling multiple project threads. The sheer volume of information coming in such a short time can be challenging.

Continue reading How MS Edge’s Immersive Reader Helps Me Slow Down

Trust No Statistic? Why Context Matters More Than Numbers

“Don’t trust any statistic you didn’t fake / manipulate yourself.” I guess most of us have heard statements like that. And it annoys me more and more. It’s often used just asa joke when it’s obvious that a statistics isn’t too easy to interpret – but I see this phrase more and more being used deliberately to produce fake news and to manipulate.

Continue reading Trust No Statistic? Why Context Matters More Than Numbers

How to selfhost Peertube

I’ve been thinking about selfhosting my videos for quite a while now. Sure, Youtube is the de-facto-standard. But honestly, I don’t perform on ANY platform that is steered by an algorithm, and I simply dislike the way those monopolies can “dictate” what is seen and can be seen. Elena Rossini recently wrote a nice post about it, that nails it quite well. I know: I don’t pay for it, so what should I expect …

Anyways! I hesitated for quite a while to rent a VPS, install Peertube, maybe according databases, redirect a subdomain, keep it maintained … ah well … or maybe just not.

But – also thank you to Elena Rossini (@_elena@mastodon.social) – I got aware of YunoHost:

YunoHost is an operating system aiming to simplify server administration and therefore democratize self-hosting while making sure it stays reliable, secure, ethical and lightweight. It is a copylefted libre software project maintained exclusively by volunteers. Technically, it can be seen as a distribution based on Debian GNU/Linux and can be installed on many kinds of hardware.

And I must admit that I really liked the setup procedure! It’s well described in the Yunohost Documentation and worked like a charm.

If you are afraid of self-hosting any application listed in the YunoHost App-StoreDON’T be afraid. VPCs/VPSs are really cheap to get as well … I chose a tiny VPC from Strato for example.

See the result on https://video.franzgraf.de/

Give it a try! Self-hosting might be easier than you think.

“Ask for forgiveness, not permission” – The Real Cost of Moving Too Fast

In Germany, there’s a saying: “Besser um Vergebung bitten als um Erlaubnis fragen” (“Better to ask for forgiveness than permission”) or “Gleich mal Fakten schaffen” (“Make decisions fast and set facts”). These phrases often glorify quick action, suggesting that speed leads to progress and success.

Well, I can tell you right now — I really hate that mindset. It may look like a shortcut to success, but in reality, it often creates a mess that no one talks about. The consequences are rarely considered in the rush for fast decisions, and I’ve seen more harm than good come from it. What starts as quick success ends up piling up technical debt, inefficiencies, and unseen costs that will have to be dealt with – sooner or later. And like financial debt, technical debt piles up quietly, and the longer you delay paying it back, the more difficult it can get.

Continue reading “Ask for forgiveness, not permission” – The Real Cost of Moving Too Fast

LLM-Search is a bit more than just “AI” – Podcast recommendation

Last week I heard the podcast SoftwareArchitekTOUR – Episode 102: Zuverlässige KI-Architektur from heise online. (german only, sorry).

I really liked the part where they discussed the technical part for a sematic search. Especially when it struck me, that the actual use of LLMs is just once per document and not in the search directly. Also, it suddenly became clear to me why you want/need a vector database for such an AI supported search.

Check it out if you can understand german or try to translate it.

8 Years in my Current Company

At the beginning of the month, a colleague reminded me that I had been with the company for 8 years now! A little anniversary! I reflected about the past years when at the same time a recruiter message reached me, what my motivation would be to “leave my comfort zone“.

Before being here, I changed jobs about every 2 years. The team was always great – super lovely people that I still miss, but I felt like I couldn’t improve anymore, I felt stuck and I felt like I wasn’t adding value to the company any more – and I then did the necessary steps.

It was the first time I joined a larger corporation. And obviously something was different here – otherwise I wouldn’t have stayed so long, right? So I started reflecting what I did all those years long. Was it worth staying? Am I, perhaps, settling into a comfort zone?

Continue reading 8 Years in my Current Company

How to add SSH public key authentication in Linux

It’s pretty easy, but every time I have to look up the right permissions for .ssh and the authorized_keys file. The solution is described on StackOverflow and the OpenSSH FAQ:

mkdir ~/.ssh
touch ~/.ssh/authorized_keys
chmod 700 ~/.ssh
chmod 600 ~/.ssh/authorized_keys

# now paste the user's public key here:
cat > ~/.ssh/authorized_keys

done.