Older development projects

Some older projects, now either discontinued or long ignored.


Programming sensor networks is a very different experience from programming modern systems: low memory, restricted computational and communications capabilities, poor development and debugging environments, and highly restricted battery power. It’s the latter that’s a real killer, especially when deploying into the wild: how do we keep the power consumption down? I did a summer project to explore these ideas, and SleepySketch was one of the main results.

The Arduino is a hobbyist hardware platform that’s peculiarly attractive for sensor systems, as it’s so cheap and surrounded by an ecosystem of extension modules (“shields”) that simplify development — especially for those of us whose electronics skills leave something to be desired. As a hobby device, though, the Arduino software platform (“sketches”) is designed for simplicity, not for longevity in sensing applications. So I wrote a harness to help keep the power down.

SleepySketch is a framework within which to deploy Arduino applications. Its main goal is to keep the microcontroller asleep for as much of the time as possible, waking up to run tasks and then going into low-power mode. This lets the programmer supply the things that should happen, and leave SleepySketch to manage the (hopefully extensive) down time.

Repository on Github


Attila is an abstract, re-targetable threaded interpreter, developed as a personal project and an experiment in language design and cross-compilation.

Threaded interpreted languages (TILs) provide a different view of language implementation to that traditionally used. I’ve written in the past about using such systems (typically known as Forth systems because Forth is the most widespread TIL) for use with sensor networks and as a possible platform for extensible virtual machines. Exploring either of these ideas requires access to the TIL that’s easily modified and easily targeted at a range of platforms. Existing open-source systems like GForth, while extremely powerful, are also rather too large and complex for the task.

Attila is an ab initio implementation of a TIL that’s intended form the start to be simple to port and cross-compile. It includes all the usual Forth features and words, the ability to define new primitives in C, and a cross-compiler able to build the system from a bootstrap and cross-compile it onto other target platforms.

Repository on Github


The Vanilla project was an experiment in language design, intended to test two hypotheses:

  1. To what extent are programming languages composed of independent sub-features that can themselves be independently described and composed?, and,
  2. If this is so to a large extent, is it possible to build effective language tools (specifically interpreters) from such independently-compiled features?

The answer to the first question is a qualified “yes”: a programming language is, to a large extent, composed of features that are either independent of other features or interact with them in modular ways. An if statement, for example, depends on there being a Boolean type in the language, but is independent of the types and expressions that appear in its two arms. This suggests that we should be able to define the typing and semantic rules of if statements independently of almost all other language features.

In developing Vanilla we developed a modular, composable type-checking and interpretation framework within which to define language features and their associated rules. We then developed a large range of language features within the framework, including most of the machinery needed to build imperative, object-oriented and (strict) functional languages.

The main disadvantage of Vanilla is that it uses Java, both as a run-time host language and for expressing the rules and structures. The code that designers have to write is extremely stylised and repetitive. It would be better to have a more powerful language for describing languages, and also to allow rule-based expressions.