Rainer Gerhards

2026-01-222026-01-22

AI Code Generation in a 200k LOC C Codebase: What Actually Worked in rsyslog

If you want the CS summary: it is at the end.

I keep seeing the same take pop up: “AI is overhyped. Mostly money burning.” Sure. There is hype. There is also a whole lot of low-effort “vibe coding” that produces low-quality output at impressive speed.

Illustration of a robotic hand carefully adjusting gears in a large rsyslog C codebase machine, symbolizing AI-assisted maintenance of mature software.

But there is also something else: if you treat AI as a serious engineering tool and you are willing to do the unglamorous work, it can make measurable difference and boost productivity and quality.

So here is a concrete case study: agentic code work in rsyslog.

2025-12-08

How We Run Open Source and Closed Source at Adiscon

I know many people know I am with Adiscon. They also know we do both open source and closed source software. That combination often raises eyebrows, and I occasionally get the same question: how do we manage this without open-core games, dual-licensing traps, or hidden agendas?

Blue and orange streams converge into a gear, symbolizing concept transfer between open source and commercial development. — Blue and orange streams meet in a gear: open source and commercial work, shared concepts, clear boundaries. (Image: Rainer Gerhards via AI)

So I decided to write it down, plainly.

2025-12-072025-12-07

Alhambra: Beauty, and Engineering That Quietly Holds It Together

I finally visited the Alhambra in Spain this year.

I had been to that part of Spain in my youth, in my twenties, but I did not manage to visit the Alhambra back then. And somehow I also never made it back in between. So it was fantastic for me to finally be there.

Panoramic view of the Alhambra on a wooded hill above Granada, with cypress trees in the foreground and mountains under a sky of scattered clouds. — The Alhambra, Granada. (Photo: Rainer Gerhards.)

2025-12-042025-12-04

The Context Wall in AI Code Generation

The ultimate promise of AI agents in software development is autonomy. We want to be able to hand off a task and have the agent execute it reliably. However, in my recent experiments with AI-driven code generation, I’ve hit a recurring roadblock: The Context Wall.

Developer bypassing a digital 'Context Wall' by manually injecting a policy file to correct an AI's code generation.

2025-12-022025-12-02

Security Theater at the Download Page

(See CS summary at the end.)

I needed to download VMware Workstation Pro for a new test environment. That now requires a Broadcom account. Registering that account did something that, in practice, pushes users toward weaker security – and it is also counter-productive for a “free download” funnel.

A laptop screen showing a password field where pasting is blocked. A broken padlock icon signifies this leads to weaker security. — Symbolic image: Blocking paste in password fields counterintuitively leads to weaker security. (Image: Rainer Gerhards via AI)

2025-11-282025-11-28

CyberSicherheitsForum Baden-Württemberg 2025: Notes from an rsyslog Perspective

Yesterday I attended the CyberSicherheitsForum Baden-Württemberg in Stuttgart. Program link: https://cybersicherheitsforum-bw.de/

For non-German readers: Baden-Württemberg is one of Germany’s federal states (Bundesländer), and an important IT region in its own right. Companies like SAP SE and Schwarz Digits are based here, and the Open Source ecosystem is active as well. When the state talks about digital sovereignty or security strategies, it is usually backed by real capabilities in industry and administration.

Conference hall with many attendees seated, large screens on stage showing the opening session of the CyberSicherheitsForum Baden-Württemberg. — Photo taken during the opening session of the CyberSicherheitsForum Baden-Württemberg 2025 in Stuttgart. (Rainer Gerhards)

2025-11-172025-11-17

The Real Scope Behind the rsyslog Documentation Overhaul

For a concise Computer Science summary of this effort, see the section at the end of this article.

When I began the current documentation overhaul, the objective was never limited to cleaning up a few pages. From the beginning, the plan was to prepare rsyslog for the AI era. And the truth is simple: without modern AI tooling, this work would not have been feasible at this depth or speed.

Symbolic illustration showing documentation, an AI head, and a graph structure representing RAG.

2025-11-072025-11-07

YAML for rsyslog — a new option, not a revolution

rsyslog’s configuration language has grown into a very capable scripting environment — RainerScript — that can express almost anything a log pipeline might need. For complex systems, that won’t change.
But many modern environments — especially containerized and cloud-native ones — expect configuration in YAML. So the idea is simple: rsyslog should understand that world directly.

Symbol image for “Engineering Rational” type of postings. (Image: Rainer Gerhards via AI)

2025-10-292025-10-29

Having a good Time with DO Gradient AI

I thought I provide a quick “happy camper” status update. Even though this was only a side-activity, I had an exceptionally good day working with Digital Ocean Gradient and open source models. I am working on support Agents for the rsyslog ecosystem.

rsyslog site with DO Gradient-based AI Agent. Right now, we do not auto-expand it, but this might change.

2025-10-242025-10-17

Schema transformation and the road to YAML in mmjsontransform

The recent GitHub issue #6251 marks another step in our ongoing modernization work. The topic sounds narrow — enhancements for mmjsontransform — but it connects to a much larger theme: how rsyslog handles structured data and how we can make that power easier to use.