Интересное дело, базы данных — особенно реляционные — разрабатывали и разрабатывают не просто для хранения данных, но и для работы с ними, в частности — анализа. OLAP значит OnLine Analytic Processing, и поддерживается всеми основными РСУБД, как коммерческими, так и открытыми. Некоторые поддерживают очень продвинутую аналитику.

И тем не менее, когда Data Analysis стал "Sexy Job" оказалось, что данные выгружают в Excel и "развлекаются" с ними уже там...

https://handmade.network/jam

Another Handmade Network's Wheel Reinvention Jam. August 15 - 21.

https://nitor.com/en/articles/pitfalls-and-bumps-clojures-extensible-data-notation-edn

Why you need an up-to-date complete specification of a serialization format with several independent implementations. 😏

https://types22.inria.fr/files/2022/06/TYPES_2022_paper_24.pdf
For all Property-Based Testing fans out there: test properties of your properties! Exhaustively! 😂
(An upcoming TYPES-2022 workshop talk.)

Recent Strumenta's webinar about transpilers: https://vimeo.com/724885911/42b82a24fe
Slides separately: https://www.dropbox.com/s/134x10dkj4w7z8q/Let's%20talk%20about%20transpilers.pdf

And another one about parsers: https://vimeo.com/671065225/12498eb5d3

Советские переводчики были настолько крутые, что даже верлибры переводили в рифму! 😂

Решил реанимировать и подновить один JS frontend проект. npm outdated ни в какую не показывает devDependencies. На SO висит старый вопрос почему npm outdated не показывает ничего кроме devDependencies. 🤦‍♀️

Оказалось, что npm outdated почему-то не может найти devDependencies без npm install. npm install конечно не работал из-за протухших зависимостей.

Вдохновлено IKEA: Web frontend -- у тебя есть инструкция, но собрать по ней ничего не возможно.

https://github.com/cucapra/dahlia

A language + a Compiler written in Scala for High-Level Synthesis (FPGA programming).

Refocusing Programming Languages Theory from programs to programming: http://tomasp.net/histogram/

Tomas Petricek presents and discusses (at length!) a prototype of a new programming language (together with two interactive online editors!) built around the process of constructing a program, not the end artifact. The language named Histogram encapsulates a formal interaction calculus representing several user interactions with a system that result in a (partial) program.

As Tomas points out this approach is akin to interactive theorem proving with tactics, where the user interacts with the system writing down tactics which implicitly build a program (most often a formal proof is a (kind of) program).

But explicit focus on interactions allows going further and capturing not only usual notions of programs-as-texts or programs-as-ASTs, but also such atypical (in academic research) forms as Jupyter Notebooks with heavily non-linear (in reality) evaluation order and implicit global state of computation. Plus good old autocompletion. 😊

The fun part: we can develop a type system for interaction calculus with non-trivial metatheoretic and practical properties. Rephrasing Robin Milner's "Well-typed programs cannot go wrong" — "Well-typed interactions cannot produce ill-typed programs". 😁

https://unsound-workshop.org/

A workshop at SPLASH 2022 about unsoundness in specifications, systems or expectations. CFP is open till 2022-09-01 (but no formal publication).

Expecting lots of Rust folks. 😁

Do you like EDM? No, not that EDM, this EDM! 😂

Образцово-показательный проект на "современном JavaScript": https://github.com/lukeed/preact-progress
Один (sic!) файл исходников и два (sic!!!) файла конфигурации системы сборки. Это они ещё package-lock.json не закоммитили и нет отдельного конфига для Babel.
Если вы думаете, что это единственный такой проект — посмотрите по Гитхабу повнимательнее.

Кто катит в кубер по утрам
Тот поступает мудро!
Известно всем, парам-парам,
На то оно и утро!

Скучна вечерняя пора,
Девопсеры зевают.
Но если катим мы с утра
Такого не бывает!

"time-series of image data" — а я всегда думал, что это называется видео... 🤔

https://www.youtube.com/channel/UCzipicZiZ1843jAqmSGgYwg

Video recordings of talks from "50 years of Smalltalk". It has a talk on theorem proving... 🤔

https://www.youtube.com/watch?v=Iz70bD63RWY

Looks neat... If they really manage to turn it into a TV-show... 😏

https://news.ycombinator.com/item?id=11941656

Under the linked message advertising Luna language (now Enso) there's a thread between Alan Kay and Rich Hickey concerning "Data vs. Objects".

The most immediately striking thing is that Rich is much more thoughtful and deep than he seems! 😅

Adding to the discussion, I guess the fact (a datum?! 🤔) that we can decipher ancient languages for which we have no interpreters around for millennia to large extend might hint that (structured) data is indeed kinda more fundamental than objects...

http://aitp-conference.org/2022/

Apparently, there's an annual conference on all things AI and Mechanised Theorem Proving (automatic, interactive and their integration). Topics cover everything from mining archives of formal proofs to proving correctness of ML algorithms to AGI applications.

This year's speaker lineup is insane from Ben Goertzel and Stephen Wolfram to Kevin Buzzard and Talia Ringer.

For those who still doesn't follow closely Lawrence Paulson's blog (shame on you!) he reviews Turing Machines, some other equivalent formalisms and the history of their development: https://lawrencecpaulson.github.io/2022/07/06/Turing_Machines.html