It is my pleasure to announce the new and improved Programming languages Zoo, a potpourri of miniature but fully functioning programming language implementations. The new zoo has a decent web site, it is now hosted on GitHub, and the source code was cleaned up. Many thanks to Matija Pretnar for all the work.
The purpose of the zoo is to demonstrate design and implementation techniques, from dirty practical details to lofty theoretical considerations:
- functional, declarative, object-oriented, and procedural languages
- source code parsing with a parser generator
- recording of source code positions
- pretty-printing of values
- interactive shell (REPL) and non-interactive file processing
- untyped, statically and dynamically typed languages
- type checking and type inference
- subtyping, parametric polymorphism, and other kinds of type systems
- eager and lazy evaluation strategies
- recursive definitions
- interpreters and compilers
- abstract machine
There is still a lot of room for improvement and new languages. Contributions are welcome!
My student Marko Koležnik is about to finish his Master’s degree in Mathematics at the University of Ljubljana. He implemented Agda Writer, a graphical user interface for the Agda proof assistant on the OS X platform. As he puts it, the main advantage of Agda Writer is no Emacs, but the list of cool features is a bit longer:
- bundled Agda: it comes with preinstalled Agda so there is zero installation effort (of course, you can use your own Agda as well).
- UTF-8 keyboard shortcuts: it is super-easy to enter UTF-8 characters by typing their LaTeX names, just like in Emacs. It trumps Emacs by converting ASCII arrows to their UTF8 equivalents on the fly. In the preferences you can customize the long list of shortcuts to your liking.
- the usual features expected on OS X are all there: auto-completion, clickable error messages and goals, etc.
Agda Writer is open source. Everybody is welcome to help out and participate on the Agda Writer repository.
Who is Agda Writer for? Obviously for students, mathematicians, and other potential users who were not born with Emacs hard-wired into their brains. It is great for teaching Agda as you do not have to spend two weeks explaining Emacs. The only drawback is that it is limited to OS X. Someone should write equivalent Windows and Linux applications. Then perhaps proof assistants will have a chance of being more widely adopted.
I am looking for a PhD student in mathematics. Full tuition & stipend will be provided for a period of three years, which is also the official length of the programme. The topic of research is somewhat flexible and varies from constructive models of homotopy type theory to development of a programming language for a proof assistant based on dependent type theory, see the short summary of the Effmath project for a more detailed description.
The candidate should have as many of the following desiderata as possible, and at the very least a master’s degree (or an equivalent one):
- a master’s degree in mathematics, with good knowledge of computer science
- a master’s degree in computer science, with good knowledge of mathematics
- experience with functional programming
- experience with proof assistants
- familiarity with homotopy type theory
The student will officially enrol in October 2015 at the University of Ljubljana. No knowledge of Slovene is required. However, it is possible, and even desirable, to start with the actual work (and stipend) earlier, as soon as in the spring of 2015. The candidates should contact me by email as soon as possible. Please include a short CV and a statement of interest.
Update 2015-03-28: the position has been taken.
I remember how hard it was to assimilate category theory when I was a student. A beginning student on math.stackexchange.com is asking for a solution to a basic lemma about pullbacks. It really is the sort of thing one should do by oneself. Nevertheless, here it is, in gory details.
Last year I participated in a project whose goal was to record at low cost my lectures on video and put them on-line. Since the most expensive parts of recording are having a camera man and manual post production, we set up a static camera and just uploaded raw video online at videolectures.net. As you can see for yourself, the sound is good (I wore a microphone) but the whiteboard is mostly illegible. In addition, it took about two weeks for the lectures to show up on-line because there were men-in-the-middle. So that got me thinking whether there was a better way.
Continue reading Video lectures as screencasts