#Python #ProgrammingLanguage #Coding #DataAnalysis #MachineLearning

Python Knowledge Base
https://python-code.pro/

Share of programming languages used by #GTK3 / #GTK4 applications :

30% #Python
23% #Vala
21% #Rust
16% #C
06% #gjs #Javascript
06% #C++ #Cplusplus
02% Other (including #Csharp, #Go, #Haskell, #Lua, #Crystal, #Swift, #d )

62% use GTK4, while still 38% use GTK3.

Method: Source [1] lists 453 awesome #gtk (3/4) #opensource applications and their #programminglanguage. Date: 2024-11-11.

[1] https://github.com/valpackett/awesome-gtk

#GTK #FLOSS #Linux #Programming #Development #Gnome

@GTK @gnome

“Python becomes the most used language on GitHub, overtaking JavaScript after a 10-year run as the most used language. This is the first large-scale change we’ve seen in the top two languages since 2019…”

https://github.blog/news-insights/octoverse/octoverse-2024/#the-most-popular-programming-languages

In the past couple days we successfully got our new language to host a hello world webserver and open a hello world desktop gui, and pushed past 2ksloc of some nontrivial bootstrapping, prelude, and tests being maintained. There's still a lot to do, like implement all the missing features to make non-hello-world projects, and the typechecker is held together with duct tape and bailing wire and it really shows sometimes. Volunteers welcome if you want to suffer with me on a really cool project to push the boundaries on what's possible in a practical programming language and proof assistant.

One of the neat consequences of the design of the language I'm working on with operatives instead of macros and a modular categorical semantics, is that everything that would be a keyword or special operator in other languages can be an operative in ours, which means they can be replaced and changed and modified in user code. This means that in this language the problem of two different dependencies of the same project wanting to use different versions of the language reduces to the problem of different dependencies wanting to use different libraries, and no system of complicated features pragmas or standard command line arguments that grow and grow and grow over time is necessary. Instead we can just have different versions of the syntax as ordinary libraries, versioned, polyfilled, dependency injected. Similarly, the formally meaningful modular semantics system means that we can add and remove features without that being a global decision; an old library that uses an old semantics can be transported to a new semantics by a functor, and we can prove that the functor implements all the axioms of the original semantics in terms of the new semantics, and then the old library will work just like a native library on the new semantics, with all its types, code, proofs and properties available for use and to the optimizer. A library written next year and proven correct should still run correctly in 300 years with no maintenance specific to that library. The compiler will of course need maintenance to keep it running on new hardware generations and adopt new technology.

But this isn't just a speculative future thing. Right now, modular semantics are useful in writing a project that works on multiple targets. For example, it might be useful to write a library that does some math which can work when compiled to JavaScript, wasm, or native. These targets have huge differences between them, but with modular semantics it's possible to just write using the minimal set of primitives needed, working from abstract high level collection data structures and math operations, and then any project that works in a specific system can just request a version of your package transported to their system, and all the high level data structures and properties will be filled in with whatever their platform uses to interop at full native speed. This also works in reverse; code that runs in a webpage using webgpu can have different primitives available based on which device they will run on, and not only share common libraries and types between them but also use target specific features like garbage collection or warp level parallel operators, even if the code is mixed in a single file to collect both facets of the implementation of a specific feature, and the compiler will give a nice type error if one of them gets used in the wrong place.

Alphabet Soup: Haskell’s Single-Letter Naming Quirks - When you used punch cards or tape to write a computer program, brief variable name... - https://hackaday.com/2024/10/14/alphabet-soup-haskells-single-letter-naming-quirks/ #softwaredevelopment #programminglanguage #softwarehacks #conventions #dictionary #mischacks #variables #variable #haskell #coding #news

Python is a protean and important programming language that caters to a wide range of disciplines and operations.

#Python #Programming #Coding #DataScience #MachineLearning #WebDevelopment #Automation #AI #SoftwareDevelopment #Tech #LearningToCode #OpenSource #DevCommunity #ProgrammingLanguage #DigitalSkills

https://www.hituponviews.com/python-a-comprehensive-guide-for-beginners/

I love how small the #ProgrammingLanguage / #compiler world is sometimes.

I was checking something on the grammar for #Smalltalk and its cascade construct and saw on the (original) C2 wiki that “Budd’s Little Smalltalk” ignored one of the odder corner cases of the cascade construct because it’s a pain to parse.

I go to Budd’s home page and it turns out it’s the same Budd who wrote the “An #APL Compiler” book that’s just a ton of fun.

Small world.

Guess it's time for an #introduction

Hi, I'm Jan, a software developer who is probably way too obsessed with #Rust. Besides Rust, I'm really interested in #TypeScript, #WebDev, #WebDesign, #ProgrammingLanguage's, #OpenSource and so much more.

I like to keep an open mind for other people's perspectives, because, IMO, only then you can truly learn something new and form connections with each other. However, I do have strong opinions on certain topics.

Come say hi and let's be friends!