The post discusses the Model Context Protocol (MCP), which allows for the extension of skills in language models by associating local functions with natural-language descriptions. It explains how MCP servers can enhance coding ass...
About:
thisContext is a website focused on adventures in livecoding.
Website:
Subscribe to RSS:
The blog post discusses the development of phase four of the Catalyst Smalltalk virtual machine, focusing on creating a WASM GC version from a Smalltalk implementation. It highlights the use of simulation to ensure type correctnes...
The post discusses the development of the Catalyst Smalltalk virtual machine using Smalltalk and WebAssembly (WASM), particularly focusing on the use of WASM GC for garbage collection. The author introduces the Epigram compilation...
The author has successfully bootstrapped a Smalltalk virtual machine and object memory as a WebAssembly (WASM) module, focusing on its speed and interoperability with other WASM modules. The initial tests involved evaluating basic...
The blog post discusses the author's project, NeuroScope, which aims to enhance mechanistic interpretability (MI) of large language models (LLMs) by identifying and composing cognitive structures within them. The author highlights...
The post discusses the application of AI in optimizing just-in-time compilation for the WASM GC Catalyst Smalltalk virtual machine. It traces the history of dynamic method translation in various programming systems and proposes us...
The author reflects on the profound impact of Vanessa Freudenberg in their life, sharing memories of their collaboration at ESUG in Prague, where they worked on a virtual machine for Pharo and Cuis. The author expresses gratitude ...
The text discusses the use of server-side Caffeine to display in web browsers, using the Tether remote messaging protocol. It explains how the Deno JavaScript runtime can run Caffeine server-side in a Web Worker thread, and how th...
0a voice-controlled Ableton Live
2025-01-05 •
The author discusses their use of AI programming to integrate generative AI text and speech into Caffeine, with the goal of writing voice control for Ableton Live. They explain how the AI translates spoken commands into code that ...
0Context-Aware AI Conversations in Smalltalk
2024-12-03 •
The author discusses using AI language models for Smalltalk conversations and tools for integrating conversations into the workflow. They have been fine-tuning the model with English system prompts and have been impressed with the...
0dynamic translation of Smalltalk to WebAssembly
2023-07-26 •
The text discusses the dynamic translation of Smalltalk to WebAssembly in the context of Catalyst, a WebAssembly implementation of the OpenSmalltalk virtual machine. It explains the linguistic levels in play, the process of transl...
The text discusses the process of creating WebAssembly (WASM) from JavaScript for the Squeak Smalltalk virtual machine. It covers the automation of WASM generation from JS, type hinting, and further JIT optimization. The author al...
0Catalyst, a WebAssembly-enabled version of SqueakJS
2023-05-08 •
Catalyst is a WebAssembly-enabled version of SqueakJS, which uses a synchronized linear representation of object memory to coordinate JavaScript and WebAssembly. The author has modified SqueakJS to maintain a shared WASM memory bu...
0a WebAssembly Squeak virtual machine is running
2023-04-14 •
The author has replaced the inner instruction-dispatch loop of a running SqueakJS virtual machine with a handwritten WebAssembly (WASM) function, and run several thousand instructions of the Caffeine object memory. The performance...
0tactical Squeak speedups with WebAssembly
2023-04-07 •
The text discusses the use of WebAssembly (WASM) to improve the performance of SqueakJS virtual machine. It explains the process of identifying performance-critical code, replacing it with WASM code, and the benefits of doing so. ...
0Caffeine Web services through Deno
2022-07-09 •
The text discusses using Caffeine in a Deno worker to provide Web APIs to Smalltalk in a native app, bridging native apps and the Web. It explains the update from using NodeJS to using the Deno JavaScript runtime, and the extended...
0HTML User Interfaces with Livecoded Web Components
2022-07-09 •
The text discusses the use of SqueakJS to run Squeak in web browsers and create web-native versions of the Smalltalk IDE. It explains how to create user interfaces with livecoding web components and renovate web browser UIs using ...
The text discusses the use of production rules in Smalltalk grammar to improve parsing, compiling, and searching. It explains how to express grammar production rules as live objects with behavior, and how to use them to parse sour...