Programming languages and systems

Overview

There are very different requirements for these intended audiences. Whereas a computer does not care much about how its instructions are structured or what names are used to describe operations and data, human programmers do. To help the human readers (often ourselves a couple of weeks or months down the road) understand what a program does, we build abstractions that hide complexity and construct software through layers upon layers of increasing abstraction to bridge the gap between the ”1’s and 0’s” and the domain where some software is used. Thus, the act of programming is one of operationalising, and thereby understanding complex workflows, and explaining them to others and ourselves.

Different programs stress about different aspects of computation. When we write software for the bottom layers of the stack, each byte or instruction is on the critical path to performance for each layer above. This warrants focus on resource management, optimisation and efficient resource use. Higher up in the stack, we may be more concerned about the ease of maintenance, extensibility and closeness to the domain, or the time it takes from inception to deployment.

The research area of programming languages and systems is concerned with the fundamentals of software development. What should a language look like to be suitable for all of the above? How do we verify that bad things never happen (e.g. we get an invalid result) while guaranteeing that good things eventually happen (e.g. we get a result)? We study different approaches to programming: functional and declarative paradigms where programming is about explaining the meaning of operations, as well as imperative paradigms where programming is about step-by-step instructions. Specialist paradigms like constraint programming is useful for efficiently solving certain kinds of problems, etc. We also study aspects of programming in software development: how to guarantee that programs are working correctly (verification), how to infer operations from input-output examples (synthesis), how to efficiently execute programmers' instructions (compilers and runtime systems), etc.