Changelog Master Feed

Researchers test open source language models and their success in achieving parity with chat GPT on certain tasks. SQLite's JSON functions are rewritten. Ratatui, a rust crate for TUIs, is introduced. A complex app built in vanilla JS is presented as a case study. Headscale is an open source, self-hosted implementation of the Tailscale control server.

In this podcast, the hosts discuss API design, including what makes a good API and why we often do REST wrong. They also explore the concept of HATEOAS and whether GraphQL is a solid choice. Other topics include the benefits of Neon, a serverless Postgres service, the advantages of hypermedia APIs, and the challenges of rewriting a music streaming app from a flash-based player to HTML5. They wrap up with plans for future episodes and sponsorship mentions.

Guest: Gregg Tavares, author of WebGL/WebGPU Fundamentals. Topics: Low-level technologies pushing the web into video games & machine learning, web development with WebGL and WebGPU, comparison and use cases of HTML rendering engines and WebGL/WebGPU, user consent for heavy compute on devices, getting started with WebGL and WebGPU, appreciation for Greg's work on gpufundamentals.org, podcasts as a medium for technology discussion.

Drew DeVault, creator of the Hare programming language, discusses the language's aim to be a 100-year language, the importance of open source, fostering cultural stability, and crafting effective interfaces in systems programming. They also touch on topics such as modern developer tools, Postgres, creating a stable programming language, and the challenges of attracting users who prioritize stability over new features.

In this podcast, the hosts discuss the problems arising from the use of machine learning algorithms in welfare systems across Europe. They explore 'suspicion machines' that generate suspicion of certain groups in need of welfare benefits. The podcast also delves into a case study on a model deployed by Rotterdam to detect fraud, including the limitations of the dataset and the impact of being flagged for investigation. The speakers discuss the process of building a predictive model for Rotterdam and the challenges and questions surrounding systems with probabilistic assessment.

Robin Berjon tries to formalize a name for those who can’t be named, Amy Lai tells the tale of the weirdest bug she’s ever seen & Facundo Olano trumps the “code is read more than written” cliche with his own: “code is run more than read.”

Gergely Orosz joins the hosts to discuss the tech hiring market, the impact of AI on software development, Neon's serverless Postgres for developers, OpenAI's success and leadership, Socket's solution for security concerns with open source dependencies, hiring trends in the tech market, the rise of social media influencers, and a conference with fake female speakers teased for future discussion.

In this podcast interview, Solomon Hykes talks about Dagger, Tammer Saleh and James McShane discuss going beyond cloud native with SuperOrbital, and Steve Francis and Spencer Smith share insights into the state of Talos Linux and their work at Sidero Labs. The topics covered include the technology behind a copy and write branching feature, experiences at CubeCon, frustrations with Wi-Fi and tethering issues, hopes for Dagger, the value of conferences, the usability and fatal flaw of Nix OS, customizing Kubernetes lab environments, the importance of continuous improvement, challenges and solutions for running Talos in Kubernetes, enhancing documentation search functionality, and improving marketing efforts.

Guest Laura Kalbeg discusses the growth of XState project, introduces Stately.ai for seamless state management solutions compatible with various tools like XState, Redux & zustand.

Sam Altman, CEO of OpenAI, joins Daniel & Chris to discuss the recent OpenAI debacle, including his sacking and subsequent return. They explore the events, people involved, and potential impact on the AI industry. The discussion touches on the history of OpenAI, the importance of diversity in model selection, and the introduction of a new model called Q star. They also reflect on an AI debate during Thanksgiving and the implications for AI risk management and regulators.