Hans-Peter Grahsl

There's a big difference between launches and landings. A launch is when you merge all the code, close that final Jira ticket, and cut a new product release. A landing is when your customers pay for it.

#TIL ... 🧐 I wasn't aware about the following #academic #titles until today ... what could possibly go wrong? 🙃

Fantastic. Thx for doing this and sharing it.

Say goodbye to cloud-only code completion (and hello to so much more).

We’ve just released Mellum-all — a general-purpose code completion model that supports every major programming language and runs entirely on your machine.

Learn more: jb.gg/dmuavo

Release v0.10.0 · ollama/ollama Ollama's new app Ollama's new app is available for macOS and Windows: Download Ollama What's Changed ollama ps will now show the context length of loaded models Improved performance in gemma3n mo...

Ollama v0.10.0 is here!

- New native app for macOS & Windows
- 2-3x performance boost for Gemma3
- 10-30% faster multi-GPU
- Fixed tool calling issues with #Granite3.3 & Mistral-Nemo
- `ollama ps` now shows context length
And more!

buff.ly/gvDiSAc

#Ollama #AI #LocalLLM #OpenSource

There is still hope... #ClaudeCode

> Ha! Well, that's quite the plot twist! 😄 So I've been analyzing your project and giving you an assessment of... yourself. Must have been amusing to read my technical analysis and then my web search to figure out you're trustworthy enough for enterprise use!

Apache Iceberg Meetup: New Trends in Streaming and Data Analytics 2025 · Zoom · Luma Join us for an incredible evening of insights, innovation, and community as we dive deep into the world of Apache Iceberg.Whether you’re an engineer,…

📣 #Apache #Iceberg #Meetup: New Trends in #Streaming and #DataAnalytics 2025 with my fantastic @decodable.co colleague Sharon Xie and Xinyu Zhou from #AutoMQ

There's still time to register
👇👇👇
lu.ma/akwa6uzh

Könnte ein Workshop setting z.b. 2 x 1/2 Tag mit einer überschaubaren TN Anzahl nicht besser funktionieren? Vor allem vlt. ein klein wenig besser skalieren, als das von dir erwähnte 1:1 Setting? Mir ist durchaus bewusst, dass 1:1 den Mehrwert maximieren würde, allerdings eben schwer realisierbar.

🤩 Lots of great memories about @devoxx.uk which is a really fantastic #developer #conference. Also, I consider it a real privilege when getting the opportunity to talk about your own #opensource work which in this case is related to #securing #datastreaming workloads running on top of #ApacheKafka.

Debezium 3.2.0.Final Released debezium.io/blog/2025/07...

Over the past few months, I've been building a new Irontools extension called Iron Vector. It's a native, columnar, vectorized runtime for Apache Flink SQL and Table API. It's built on top of Rust, Arrow and DataFusion.

Rapid Prototyping a Safe, Logless Reconfiguration Protocol for MongoDB with TLA+ | MongoDB Blog Learn how MongoDB designed a safe, logless reconfiguration protocol using TLA+ to accelerate development and ensure correctness.

My colleagues Will Schultz and Siyuan Zhou wrote about a Raft enhancement they prototyped in TLA+, it's now how MongoDB actually adds and removes servers: www.mongodb.com/blog/post/te...

Quarkus in Action | Red Hat Developer Download Quarkus in Action, a practical guide to building resilient and scalable, cloud-native, enterprise Java applications using the Quarkus framework

I'm super happy that I can share that the FREE FULL VERSION of Quarkus in Action continues to be available until January 16, 2026!!!

Download and share!

developers.redhat.com/e-books/quar...

Thanks to Red Hat developer!!!

Apache Kafka 3.9.1 is now available for download: buff.ly/5ONjJgB

This bug-fix release has several fixes, and most importantly, it adds Java 23 support. #opensource

One such #LLM moment... priceless 😬

Legs? Where we are going, we don't need ... 🤪 #BTTF

Let's have some fun. In this NYT story, we see a picture of an OpenAI billboard on a college campus, suggesting the following prompt: "Give me a guide for mastering this Calc 101 syllabus. In two weeks. Small steps please."

What happens when we actually use ChatGPT to do this? THREAD!

But hey... also http sse and more recently streamable http 😜

3 hands 12 fingers... is that what you mean? ;)

4. 🗣 Deliver 2️⃣✌️ talks which is a real honor and privilege I'm truly thankful for 🙇‍♂️. Thx for joining me in my sessions which hopefully have been useful and gave you some new 💡 and inspiration. 🙏

After #Current25 London is before #Current25 New Orleans where I'd ❤️ to be part of again! 🤞

3. Experience lots of great moments during several side activities e.g. the 5k #MorningRun 🏃‍♂️ on Day 1, the #CommunityCatalyst 🏆 meet & greet during lunchtime on Day 2, and nice dinners 🍽 shared with like-minded and brilliant folks. 👇

2. Take the chance to briefly catch up with a bunch of amazing people in the #hallways 🤩, many of them I've been knowing for years, several who I'm lucky to call my 🫂 #friends. 👇

Had a fantastic time at #Current25 in London 🇬🇧 this year where I was able to:

1. Meet with #Decoders and help out at our @decodable.co booth as much as possible. 👋 Always a great way to listen, exchange ideas, learn about 🔥 questions & challenges from others in the #datastreaming #community 👇

Release 1.2.0 · quarkiverse/quarkus-mcp-server What's Changed devui: add input field for bearerToken header by @mkouba in #220 Add secure-http-server demo by @sberyozkin in #199 docs: add sberyozkin as a contributor for example by @allcontribu...

We've just released #quarkus MCP server 1.2.0 with initial support for "streamable HTTP" transport, default values for ToolArg/PrompArg annotations, feature filters, and other bugfixes and minor improvements... ☕ github.com/quarkiverse/...

Got to admit: Microsoft absolutely nailed its developer agent positioning.

Other startups: "we built an AI Engineer" - suggesting to leadership that it will replace developers (and thus worth paying $$$ for)

Microsoft: "a peer programmer" - suggesting it needs devs (it does!)

BTW have you seen that there is also A2A spec (by Google) for agent communication? google.github.io/A2A/specific...

yeah... SSE has been superseded with streamable HTTP in the spec I think it was in March. I personally consider the SSE way deprecated since then.

This stuff went viral out there - I've seen a bunch of crazy ... out there 🙈 The hype is real and it's scary man. Combine this with first class vibe coding and you get a pretty toxic cocktail 😬

My ⚡ talk choice at #Current25 is @hpgrahsl.bsky.social and vector stores in Flink

I think we can arrange a call maybe some time next week. let's see. we'll figure it out.