New week, new Rust! What are you folks up to?
Slowly starting to decouple Tetra from any particular windowing/graphics API, with an eye towards getting stuff running on WASM or in headless mode (the latter would be nice for unit testing).
So far I’ve managed to pull out all of the platform-level stuff (i.e. everything I currently use SDL for) into a trait, which is a decent start!
Last week was pretty busy with the 4th. This week I’m mostly working on
uom (type-safe zero-cost dimensional analysis) documentation and PR reviews including one to add 6 new thermodynamic quantities!
I’m working on bifröst which aims to provide “virtualization via containers.” My friend and I were trying to use valgrind for some school projects, but there wasn’t support for macOS Mojave. We originally worked on a solution in bash, but choose to write the project in Rust (because, fun).
Instead of writing a solution for one app in particular we are working on making the application general enough to virtualize more than just valgrind usage!
I’m writing a set of (hopefully) safe bindings for scanning documents with SANE: https://gitlab.com/teozkr/scankiosk/tree/master/sane-scan (rustdoc)
It should now be feature-complete in that you can set up a scan and get an image, but there are still a few more features I’d like to get around to before moving on to my primary goal. Notably, async scanning is pretty much a requirement for hooking up a nice GUI to it.
My longer-term goal is to cobble together an all-in-one “scanner kiosk” system that is suitable for libraries and other public locations.
Currently a lot of libraries just seem to hook up a scanner to a Windows computer and let people have a field day with it, which has a lot of security issues (often questionable and fragile policies for cleaning up after users (if there are any at all), encouraging people to log into their email on public computers, etc).
Given that the people who have to go to the libraries to scan their documents tend to be in a vulnerable situation already, and that the documents that people are still scanning tend to be quite sensitive (employment contracts, doctor’s notes, and so on), that seems like a travesty.
- Reorganizing logic to better fit the ECS paradigm.
After succeeding to implement & display Heapsort with colours for learning purposes, I am trying to implement the Rubik’s Cube in Rust, which I know will be a challenge.
I’ll have a question about it in the help section, by the way.
I am back with my work on as idiomatic as I can bindings to Tensorflow in Rust. It is very challanging and very fun for me mostly, because Tensorflow C API is not the best documented library I have ever seen.