Haskell

The most recent release of GHC includes a technology preview of a WebAssembly backend. This allows Haskell code to run in many new contexts.

GHC's WebAssembly backend is based on an earlier research compiler called Asterius. In this talk, Cheng Shao, the main author of both projects, presents an introduction to WebAssembly and why compiling Haskell to it is different than other platforms. The talk includes many details about why the backend is the way it is, along with instructions to work on it.

This talk was presented at the 2023 GHC

Contributor's Workshop in Rapperswil, Switzerland, colocated with Zurihac. The workshop was arranged by the Haskell Foundation, OST Eastern Switzerland University of Applied Sciences, and the GHC development team.

Cheng Shao is a full-time software engineer at Tweag, where he has been working since 2018. His main contribution to GHC is the WebAssembly backend, which grew out of his early research project, a Haskell-to-WebAssembly compiler codenamed Asterius. He focuses on maintaining the GHC WebAssembly backend, adding new functionality, as well as other GHC work that involves code generation and the runtime system.

cross-posted from https://slippy.xyz/notes/9vujpeyssp

cross-posted from: https://infosec.pub/post/12126221

h/t @bgamari@mastodon.social

Answering the question raised at the end of Part 1, we take a look at how a hypothetical Strict Haskell would tie the compilers hands despite pervasive purity. We also examine how laziness permits optimizations that come with no intrinsic cost and compare its benefits to a strict language with opt-in laziness.

Part 1:

• Laziness in Haskell — Part 1: Prologue
Series Playlist:

• Laziness in Haskell

— Contact: • Tweag Website: https://www.tweag.io/ • Tweag Twitter: https://twitter.com/tweagio • Alexis King's Twitter: https://twitter.com/lexi_lambda

We teach you Haskell

Here's the conclusion of the paper Wadler is referring to in this interview:

Proposition as Types informs our view of the universality of certain programming languages. The Pioneer spaceship contains a plaque designed to communicate with aliens, if any should ever intercept it. They may find some parts of it easier to interpret than others. A radial diagram shows the distance of fourteen pulsars and the centre of the galaxy from Sol. Aliens are likely to determine that the length of each line is proportional to the distances to each body. Another diagram shows humans in front of a silhouette of Pioneer. If Star Trek gives an accurate conception of alien species, they may respond “They look just like us, except they lack pubic hair.” However, if the aliens’s perceptual system differs greatly from our own, they may be unable to decipher these squiggles. What would happen if we tried to communicate with aliens by transmitting a computer program? In the movie Independence Day, the heroes destroy the invading alien mother ship by infecting it with a computer virus. Close inspection of the transmitted program shows it contains curly braces—it is written in a dialect of C! It is unlikely that alien species would program in C, and unclear that aliens could decipher a program written in C if presented with one. What about lambda calculus? Propositions as Types tell us that lambda calculus is isomorphic to natural deduction. It seems difficult to conceive of alien beings that do not know the fundamentals of logic, and we might expect the problem of deciphering a program written in lambda calculus to be closer to the problem of understanding the radial diagram of pulsars than that of understanding the image of a man and a woman on the Pioneer plaque. We might be tempted to conclude that lambda calculus is universal, but first let’s ponder the suitability of the word ‘universal’. These days the multiple worlds interpretation of quantum physics is widely accepted. Scientists imagine that in different universes one might encounter different fundamental constants, such as the strength of gravity or the Planck constant. But easy as it may be to imagine a universe where gravity differs, it is difficult to conceive of a universe where fundamental rules of logic fail to apply. Natural deduction, and hence lambda calculus, should not only be known by aliens throughout our universe, but also throughout others. So we may conclude it would be a mistake to characterise lambda calculus as a universal language, because calling it universal would be too limiting.