Today I learned

Congratulations on writing your first HTML document.

A friend of mine reached out today and asked for some help getting started learning how to write web apps - and while this is definitely not what I sought out to be an expert in, I’m flattered in a way that I can be a resource for this kind of question. And while this series is really designed for one person, feel free to follow along if you find it interesting.

In this post I am going to give some examples for parallelizing software, using Rust to explain it, but there is nothing special about Rust except that it forbids a few kinds of mistakes. In fact, the best approaches I plan to show in this post was popularized in Go (and to some extent Smalltalk many years ago, maybe others too). It helps that this is a very general problem because it may actually help you understand teams, companies, governments, and other organizations. Most of these strategies could work without code.

A friend of mine has been learning Python over the past few weeks, since he wants to improve his career opportunities in the future. He made the mistake of telling me about it. It’s fun to work through some of the basics again and we took about an hour to start from “Hello World” and end up estimating similarity of phrases at the end. The beginning is easy to understand, but the last step is a bit of a jump I added to make it more interesting.

If you’re moving from Numpy to NDArray, or even if you use Rust at all, you are probably mindful of performance and efficiency. So at the risk of preaching to the choir, I would like to emphasize that high efficiency pays for it’s development in several ways that are not always clearly visible on the outset.

• Pay less for servers, HVAC, electricity, or cloud VMs, quotas and the like
• Scale up to larger problems without any additional orchestration overhead
• Scale down to smaller hardware, or do something on edge devices once done on servers
• Increase reliability of clusters by lowering resource utilization
• Improve responsiveness of live applications

Being a data scientist by trade, one of the first things I want to do with Rust is accelerate my numerical code. You would be surprised how well optimized numpy is, especially for large arrays, and I’ll be honest that moving to ndarray in Rust alone is not going to make your code faster; it will probably be the same.

Most Rust applications will sooner or later need to take some command line arguments, even if they are GUI apps, and possibly even web servers or microservices. There are plenty of ways to do this, and while clap is not necessarily the simplist or applicable to everyone, it is quite flexible and appropriate for most circumstances. It’s also the most common, from my small and definitely unscientific survey of applications.

Hi friends! There are a number of tutorials that can take you through a basic Rust application, and in particular I would like to highlight the always-awesome Rust Book which can take you through tutorials on this and many other, much deeper and more thought provoking problems!

Let’s take a look at how to set up a USB TV Tuner with Linux using w_scan2, dvbv5-zap, and ffmpeg.

I got a new Raspberry Pi 4! I wanted to install Arch on it, and here’s how I set it up.

Today I saw a hourly weather radar forecast, showing a storm front passing through. But there was an interesting artefact you’ve probably seen before and not paid much attention to. The rain bands teleport across the map, fade out, and teleport again. Looking at the results, it’s pretty obvious how this works. Every point on the map had its values interpolated between discrete forecast steps independently of all the others.

Quick reference for finding types for your DESCRIBE.

Thanks to a brew package by FiloSuttile, and a post by Timryan, it’s actually pretty easy.

Build straightforward ML web apps using websockets that can still move to production