Improving the DX for Newbie and Casual Array Developers

Array programming languages offer unmatched expressiveness for array operations. Despite their relevance in machine learning and scientific computing, languages like APL, J, and k remain niche. We attribute this limited adoption to "Array Phobia," characterized by two key barriers: "A...

Full description

Saved in:
Bibliographic Details
Published in:2025 IEEE/ACM Second IDE Workshop (IDE) pp. 16 - 20
Main Authors: Thomas, David, Samadi, Will
Format: Conference Proceeding
Language:English
Published: IEEE 03.05.2025
Subjects:
Array programming languages
cognitive barriers
Computer architecture
Computer languages
Conferences
developer experience
Dyslexia
IDE affordances
language design
Machine learning
Programming
Scientific computing
Semantics
Syntactics
Visualization
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Array programming languages offer unmatched expressiveness for array operations. Despite their relevance in machine learning and scientific computing, languages like APL, J, and k remain niche. We attribute this limited adoption to "Array Phobia," characterized by two key barriers: "Array Dyslexia," which impedes code comprehension, and "Operator Overload," which overwhelms users with dense operational semantics. To address these challenges, we present Alice, an IDE that enhances array programming through innovative features including literate expressions and contextual visualization. We demonstrate Alice's capabilities through Poky, our experimental array dialect that maintains the power of traditional array languages while introducing carefully selected syntactic improvements. While we use Poky to illustrate the IDE, Alice's features are language-agnostic and applicable to other array languages. Our results suggest that thoughtful IDE design can lower the barriers to array programming without compromising the features of traditional REPLs.
DOI:10.1109/IDE66625.2025.00008