Smalltalk, Haskell and Lisp(storytotell.org)

103 points by todsacerdoti 1 year ago | 48 comments

A History of IDEs at Google(laurent.le-brun.eu)

441 points by laurentlb 5 months ago | 286 comments

Blog post by Laurent Le Brun recounting the history of IDEs and code editors used internally at Google for its main codebase.

Transcoding Latin 1 strings to UTF-8 strings at 12 GB/s using AVX-512(lemire.me)

110 points by mariuz 3 years ago | 87 comments

An Overview of the Starlark Language(laurent.le-brun.eu)

100 points by thunderbong 2 years ago | 43 comments

Beating the Compiler(mattkeeter.com)

172 points by mkeeter 2 years ago | 60 comments

For processing strings, streams in C++ can be slow(lemire.me)

81 points by greghn 2 years ago | 73 comments

Using the railway network as a flatbed scanner(philo.gay)

419 points by otherayden 2 months ago | 69 comments

Personal blog post describing using an industrial linear scanning camera to capture very wide photos from trains and ferries, with example images.

Security Disclosure(blog.railway.com)

571 points by passenger 4 days ago | 334 comments

Railway blog post disclosing that an unauthorized party may have accessed and decrypted customer environment variables, urging users to rotate all secrets.

SIMD-accelerated computer vision on a $2 microcontroller(shraiwi.com)

53 points by zdw 18 hours ago | 9 comments

A blog post describing a SIMD-accelerated FAST corner detector written in assembly for the ESP32-S3, roughly doubling the reference implementation's performance.

AI has access to a vastly larger working memory than the human brain(davidepiffer.com)

524 points by rzk 2 months ago | 452 comments

A Substack essay by Davide Piffer arguing that AI's edge over mathematicians comes from a vast symbolic working memory rather than superior reasoning.

Deep Learning Is Not So Mysterious or Different(arxiv.org)

445 points by wuubuu 1 year ago | 112 comments

arXiv paper by Andrew Gordon Wilson arguing deep learning generalization phenomena like benign overfitting and double descent are explainable by soft inductive biases.

Trains were designed to break down after third-party repairs, hackers find(arstechnica.com)

608 points by Stratoscope 2 years ago | 211 comments

Ars Technica article reporting that hackers found a train manufacturer's firmware made trains break down after third-party repairs; the manufacturer called it slander.