David Abreu

Phantasy star ❤️

Best Pixel Art Nintendo Switch Games Our pick of impeccable pixels

Best Pixel Art Nintendo Switch Games. (Repost)

YouTube video by How to ADHD Is Internalized Ableism Holding You Back?

Seeing my TikToks / The Anti-Planner featured as a means of dismantling internalized ableism means so much.

Thanks so much for the shoutout, @howtoadhd.bsky.social! 😭💖

ECS Orchestration Part 2: Service to Service comunication The first post in the series was about choosing the correct network type for running containers on...

dev.to/dbanieles/ec...

The Elastic Container Service (ECS) for containerized workloads on AWS is a good choice. There is no control plane to pay for and the setup is simple. When your services running in ECS need to communicate with each other you have choices on how to make this work. (1️⃣/3️⃣)

🧵

I wrote a short post on learning the fundamentals of distributed systems, with a few suggested resources to read and a few suggested projects to try.

notes.eatonphil.com/2025-08-09-w...

Gg wp

Wow, nice astrophotography setup! What’s your equipment?

you gotta be kidding me 😅

The 'relief' of @anthropic.com engineers after seeing all that traffic from AI editors suddenly turning to GPT-5 😬

Assert with Grace: Custom Soft Assertions using AssertJ for Cleaner Code - Elias Nogueira Introduction If you don't know what a soft assertion is, please read Soft Asserts – Why should you use them for Unit and Integration tests? This article is a continuation of Assert with Grace: Custom…

Want cleaner, more expressive tests in Java?

✅ Avoid endless try-catch
✅ Group failures with elegance
✅ Use AssertJ the right way

Soft assertions can raise your testing game without clutter.

Learn how:
eliasnogueira.com/assert-with-...

#java #quality #test #qualityengineering

OpenAI’s new open weight (Apache 2) models are really good The long promised OpenAI open weight models are here, and they are very impressive. They’re available under proper open source licenses—Apache 2.0—and come in two sizes, 120B and 20B. OpenAI’s …

OpenAI released their long-promised open weight models today under clean Apache 2 licenses and with benchmarks that put them shockingly close to o3-mini and o4-mini

I've run the smaller (20B) model on my Mac and it's very impressive, despite only using ~15GB of RAM simonwillison.net/2025/Aug/5/g...

Cavado. Northern 🇵🇹

@maven.apache.org 4 (from 4.0.0-rc4 on) contains the Maven Upgrade Tool with which you can automatically upgrade your Maven project. I finally found time to write a small article about it's feature and how to use it. Please test the tool and give feedback :)

maven.apache.org/tools/mvnup....

You just described every politician

Architecting Multi-Agent AI Systems in Java with Quarkus and Langchain4j
How to orchestrate scalable AI workflows using message-driven agents and local LLMs with Kafka
buff.ly/qqRHZr8
#Java #Quarkus #AiAgents #LangChain4j #Kafka

Technology | 2025 Stack Overflow Developer Survey

once again, #rustlang is the most admired programming language on the stack overflow survey: survey.stackoverflow.co/2025/technol...

Osaka's Expo 2025 is shining a spotlight on Japan’s fusion cuisine Japan’s beloved fusion dishes like curry rice and tonkatsu may seem traditional today, but they’re rooted in surprising Western influences - a story Expo 2025 is bringing to the world’s table.

Japan’s beloved fusion dishes like curry rice and tonkatsu may seem traditional today, but they’re rooted in surprising Western influences - a story Expo 2025 is bringing to the world’s table.

YouTube video by Dan Vega Spring AI Advisors: Intercept & Enhance Every AI Call (AOP for AI!)

🚀 Spring AI Advisors are like AOP for your AI interactions! Intercept, modify, and enhance every AI call without touching your core business logic:

✨Add logging automatically
✨Inject context with RAG
✨Enable chat memory
✨Build custom advisors

youtu.be/1MGiDBI2Ci4

1982

Beautiful

Yep, it’s pretty much it 😅
I even started to get job offerings for “AI code reviewer”

A scatter plot comparing small language models based on Win rate (%) vs Model Size (Billion parameters). The chart evaluates models on 12 popular LLM benchmarks. ⸻ Axes: • X-axis (horizontal): Model size (in billions of parameters), ranging from ~1.7B to 4.5B • Y-axis (vertical): Win rate (%)—higher is better—ranging from 2% to 5% ⸻ Highlighted Insight Areas: • Top-left corner: Ideal zone for models that are better (higher win rate) and smaller (faster/cheaper) • Diagonal line/gray band: Represents the tradeoff baseline; models above it are more efficient per parameter ⸻ Models Plotted: Top-right quadrant (largest, highest win rate): • Qwen3 4B – Highest win rate, ~5% • Gemma3 4B – Slightly below Qwen3 4B Mid-left (smaller but strong performance): • SmolLM3 3B – Strong win rate (~4.4%), outperforming larger models • Qwen2.5 3B – Moderate win rate (~3%) • Llama3.2 3B – Slightly below Qwen2.5 3B Lower-left (least performant): • Qwen3 1.7B – Lowest win rate (~2%) ⸻ Conclusion: • SmolLM3 3B stands out as most efficient, achieving high win rate with a relatively small size. • Qwen3 4B and Gemma3 4B are top-performers overall but less efficient per parameter. • Models like Qwen3 1.7B lag significantly behind in both size and win rate.

SmolLM3: a highly detailed look into modern model training

this is amazing. They go into great detail on just about every aspect. The number of stages, algorithms, optimizer settings, datasets, blueprints, recipes, open source training scripts, ..

huggingface.co/blog/smollm3

"It's Heartbreaking" - The Pokémon Company Tech VP Joins Industry In Criticising Microsoft Layoffs "What a colossal waste of talent."

"It's Heartbreaking" - The Pokémon Company Tech VP Joins Industry In Criticising Microsoft Layoffs. (Repost)

I heard JRPG music? I’m all in!

The anime was amazing :)

Such a clean (embedded) Rust walkthru - and a cool use-case!

“Navigating Mars with Rust: Developing an Autonomous Pathfinding Rover” from @adacore.bsky.social

blog.adacore.com/navigating-m...

Tales from the jar side: A rant about Java LTS versions, Scary Awesome AI, and the usual silly social media posts What happens when a microscope crashes into a telescope? They kaleidoscope (rimshot)

Tales from the jar side: A rant about Java LTS versions, Scary Awesome AI, and the usual silly social media posts open.substack.com/pub/kenkouse...

How Do You Teach Computer Science in the A.I. Era? Universities across the country are scrambling to understand the implications of generative A.I.’s transformation of technology.

How Do You Teach #ComputerScience in the #AI Era?

www.nytimes.com/2025...

#GenerativeAI

Get the location of the ISS using DNS I love DNS esoterica. Weird little things that you can shove in the global directory to be distributed around the world instantly(ish). Domain names, like www.example.com usually resolve to servers.…

Get the location of the ISS using DNS – Terence Eden’s Blog buff.ly/V3ocDxl
#dns 🫶🛰️