Andrew Healey

Great question! Andrew Healey wrote a great blog post about sandboxing, particularly related to us: https://healeycodes.com/sandboxing-javascript-code And we recently launched more secure semantics: https://blog.val.town/blog/restricted-library-mode But in the general case, it's a hard, never-ending cat-and-mouse game, particularly with free and unauthenticated use. We will eventually probably need to restrict use severely unless you have somehow proved you're a real person (ie connect to your github account) or are a paying customer.

Andrew Healey just took a pass at this and did a great write-up about it here: https://healeycodes.com/doom-rendered-via-checkboxes It looks decent but dithering would be nice so I created an issue for it here: https://github.com/bryanbraun/checkboxland/issues/20

This is pretty standard ( or at least used to be 20 years ago ) in high performance chess programming, see https://www.chessprogramming.org/Bitboards https://healeycodes.com/visualizing-chess-bitboards

If you really wanna get your blood boiling how about some Doom with HTML checkboxes - https://healeycodes.com/doom-rendered-via-checkboxes

Yeah, that's what I was thinking: just throw the whole lot inside a Docker container and call it a day. Unless you're dealing with potentially malicious code that could break out of a container, that should isolate the rest of your machine sufficiently. Alternatively, PyPy is actually fully sandboxable. On Linux, you can also use `seccomp.` See, for instance, https://healeycodes.com/running-untrusted-python-code

I think it's awesome that one of the first comments suggested that he ported DOOM to the checkbox renderer, and then 3 days later someone else actually implements it: https://healeycodes.com/doom-rendered-via-checkboxes

Very related: Doom rendered via checkboxes https://healeycodes.com/doom-rendered-via-checkboxes

Two examples although both in the 'fun' category: - https://healeycodes.com/boids-flocking-simulation/ - https://dstoiko.github.io/posts/go-pong-wasm/

I wrote about building the boids flocking simulation with Ebiten and deploying it with WebAssembly[0]. I found Ebiten very easy to build a simple project with. The project maintainer is helpful, kind, and highly responsive too. There is a #ebiten channel in the Gophers slack that is quite active. - [0] https://healeycodes.com/boids-flocking-simulation/

Obviously general purpose programming language(s) excluded[0]. -- per 'minimal' Does usable / minimal mean no ecosystem / integrated development environment? No ecosystem/IDE would seem to exclude command line shell & unix 'one program does one thing well'chaining/sourcing. 'Excel as a Turing-complete Functional Programming Environment' would be excluded too![8]. Do at least need the ability to abstract/store & forward to get away from 'endless tape streams'. (traditional grep vs. grep with 'state awareness') Would seem to some what depend on the domain of discourse. 1) Esoteric Programming Languages: "Let's Take Esoteric Programming Languages Seriously"[1][2]. Is a metaphor programming language an esoteric IDE?[3] (vs. apl/mumps) 2) Domain Specific Languages[4]: Usable/minimal with a philosophical background : 'mov is turing complete'[5] symbolic macro application of electrical engineering concepts of NAND[6] APL works for some enginering backrounds. One can assemble a checklist of doom[7] ------------------------------------------------------------------- [0] : https://en.wikipedia.org/wiki/General-purpose_programming_language [1] "Let's Take Esoteric Programming Languages Seriously" : https://news.ycombinator.com/edit?id=48334594 [2] "The Hacker Folk Art of Esoeric Coding" : https://news.ycombinator.com/item?id=45371188 [3] "Esolang, the esoteric progrmaming languages wiki": https://news.ycombinator.com/item?id=16894676 …

Relevant: https://discuss.python.org/t/extending-subinterpreters-with-sandboxing… https://stackoverflow.com/questions/3068139 https://wiki.python.org/moin/SandboxedPython https://github.com/jailctf/pyjailbreaker https://healeycodes.com/running-untrusted-python-code https://lwn.net/Articles/574215/