Let’s try to simulate a non-trivial circuit in Racket using the techniques proposed in the previous post.

Vermicel is a 32-bit RISC processor core that supports most of the base instruction set of the RISC-V specification (RV32I). It was initially designed and implemented in VHDL under the name « Virgule » to serve as an illustration for the digital electronics course that I taught. The VHDL source code is not publicly available yet but you can get an overview of what Vermicel is by using its web-based simulator emulsiV and reading its documentation.

In the series My first domain-specific language with Racket, I have created a simple hardware description language (HDL) called Tiny-HDL. The intent of the series was to illustrate the techniques for creating a domain-specific language in Racket, from parsing to code generation, with a focus on name resolution and semantic checking. In this post, I will focus on the runtime aspects: how can we simulate a digital circuit description in Racket.

This is the final step of my Racket DSL implementation roadmap. It consists in promoting Tiny-HDL as a standalone language with two flavors: a Lisp-like syntax and a custom syntax.

Racket offers a sophisticated module system that allows to organize a program into multiple files. In this step, we will show a solution to take advantage of Racket modules in a DSL.

In this post, we add more semantic checks. A circuit description will be considered correct if it respects some common electronic design rules.

The semantic checker for Tiny-HDL will be written in two steps. In the name resolution step, the checker introduces scopes in the abstract syntax tree, and uses them to map each named reference to the corresponding declaration.

In step 1, I have written the full adder example in Racket, following a few mapping rules to express the concepts of the Tiny-HDL language. The next step consists in writing a code generator that implements these mapping rules so that we can generate Racket code automatically.

In the previous post, I have sketched an informal specification of a small hardware description language called Tiny-HDL. Our goal is to execute circuit descriptions, written in Tiny-HDL, on the Racket platform. Which means that we need to implement a compiler from Tiny-HDL to Racket.

As explained in the proposed language implementation roadmap, we will start in the execution step, with a hand-written Racket example program that implements the Tiny-HDL concepts.

For many years, I have felt an attraction to languages from the Lisp family, but I was held back by two concerns: my indecisiveness about which variant to learn, and the lack of a personal project to implement with it. Recently, I have started drafting the Hardware Description Language (HDL) of my dreams and I have decided to try Racket, a Scheme variant that is also advertised as a language-oriented programming language.

This post is the first of a series about « midi@3:14 », my home-made MIDI keyboard. In the following sections, you will read about the hardware design and assembly. The next posts will be dedicated to the firmware and companion software.

Elementary OS is a Linux distribution that comes with its own desktop environment called Pantheon. The developers do not officially support installing Pantheon separately from the rest of the distribution. In fact, while Elementary OS Freya is based on Ubuntu 14.04, it requires newer versions of several packages.

Fortunately, if your computer is already running Ubuntu 14.04, there is a reasonably easy way to transform your existing setup into a seemingly genuine Elementary OS setup.

Sozi is a zooming presentation tool that I have been developing for five years in my spare time. Sozi is free software. It has gathered some enthusiastic and faithful users, and I usually receive very positive feedback.

However, people do not always realize that Sozi is mainly the work of a single person with the help of several contributors. There is no company behind it, no structured development team, no professional-grade support, and no ambition to gain market shares.

In this post, I will try to clarify my motivations for developing Sozi.

On the Nexus 7, I have always been surprised by the poor reactivity of hand-writing and drawing applications. Generally, there is a noticeable delay between the gestures of the finger (or stylus) and the progess of the lines being drawn. In some cases, it is difficult to draw small details rapidly: the resolution of the lines decreases with the speed of the gestures.

I have created a simple demo application to test how the Flatfish tablet with Firefox OS behaves and to experiment ideas to achieve an acceptable result.

In January this year, Mozilla launched the Tablet Contribution Program to help test and improve Firefox OS on tablet devices. I volunteered as soon as the applications opened and on the 1st of April, I received a confirmation that I was one of the 500 selected participants. A few months passed. And two weeks ago, just before leaving for holyday, I finally received the tablet.

In this post, I will give an overview of my first impressions with the device. More posts will come with details about what I am doing with it.