This project serves as an end-to-end working example for testing, building, linting, and deploying an AWS Lambda Node.js function to multiple environments using AWS CloudFormation, Azure Pipelines, and Azure DevOps. The complete source code is located in this GitHub repository, and the build output is publicly available via Azure DevOps.
For a weekend project, I wanted to build a nontrivial Mendix widget that required additional NPM packages, interacted with entity attributes, and had internal React state not managed by Mendix entities. Having just used the Ace editor to build a querying environment for my fantasy football league’s databases and views, I thought building similar functionality in a Mendix application would be an excellent fit.
Visual Studio Live in San Diego was an amazing opportunity to learn about new Microsoft technology like Azure’s artificial intelligence offerings and .NET Core 3.0. I attended some talks about leadership and Agile as well. I’ve put together several proofs of concept and documented what I learned about:
SonarQube is a very powerful source code continuous inspection tool. It’s incredibly valuable to have integrated into a project’s CI/CD pipeline to help maintain code quality and prevent issues. In the process of adding this to some existing open-source C# projects, I discovered that there’s not quality, up-to-date documentation on integrating C# inspection with Travis CI. This GitHub project serves as a functioning example.
While my development choices for the frontend change as rapidly as new JavaScript frameworks come and go, I’ve remained pretty consistent with using the .NET ecosystem for my server-side needs. I still use Node.js, Spring Boot, and Flask for several side projects, but I’ve grown to love the power of .NET and C# over the past several years while it has been the default technology at my job. My two biggest complaints were the monolithic scale of .NET Framework and the fact that it required Windows (or Mono which hasn’t always been backed by Xamarin/Microsoft). Both of these changed with the advent of .NET Core. This alternative is cross-platform, modular, and much more performant. While both .NET Framework and .NET Core are implementations of .NET Standard, .NET Core very much seems to be the preferred way of the future.
It’s very well documented that I’m a huge college football fan. We’re presently in the College Football Playoff era of Division I football, which involves a selection committee choosing four playoff teams to compete to be the national champion. The previous era, known as the Bowl Championship Series era, involved a combined poll of human experts and computer algorithms choosing the two best teams to play in the national championship game. One such algorithm is known as the Colley Matrix. Though not a factor in the post-season selection process anymore, it’s still referred to at times, particularly when debating the selection committee’s decisions. Based on the whitepaper written by Colley himself and [an existing JavaScript implementation](an existing JavaScript implementation), I developed this NuGet package for simulating head-to-head matchups and applying the Colley algorithm. This algorithm can be applied to any sport or competitions without tie games.
This NuGet package provides a C# interface to easily query and fetch kanji data from the Kanji Alive public API. This package is designed to simplify development of Japanese learning desktop and web applications on the C#/.NET platform.
It’s safe to say that I’m biased because of my alma mater, but I personally think that burnt orange is one of the most versatile colors. It’s somewhat polarizing due to being the primary color of the University of Texas, but I use it as one of the core colors here on my personal website to show my Longhorn pride. Inspired by this color palette, I’ve created a theme for Atom, my favorite text editor, and released version 1.0.0 just in time for the start of college football season.
I have long battled my internet service providers, and they have waged war on their customers for many years as well. Poor customer service, data caps with obscene overage charges, and many other issues have plagued me for years like many other Americans. Like many of these people, I’m stuck with only one viable option for a provider due to their monopolistic practices. Recently, I was forced to upgrade to a more expensive monthly plan after going over my data cap three times non-consecutively over a year and a half period. Rather than charging me the usual absurd per-gigabyte overage charge, I’m now stuck paying more per month for the remainder of my time with my ISP.
This software development paradigm has become prevalent in many modern frameworks and tools. It can be described as streamlining and standardizing aspects of development similar to all or most projects while specifying configurations for those that deviate from the established convention. For example, most Node.js developers call their main file server.js. Let’s say that I want to write a generic deployment script for Node.js applications. It’s much easier to include a simple command like node server.js than to try to determine the developer’s intended start file. This simple agreement saves a great deal of time and configuration details. The emphasis of this paradigm is that the knowledge will hold for any team or company, and so organizations full of old tribal knowledge become a thing of the past.