Proofs are… programs?
Type checking is the same as verifying a proof? WHAT?!
Software Engineering
One of the most practical classes I took.
Internet, Web & Cloud Systems
Also one of the most practical classes I took.
Functional Programming
One of my favorite classes. The Haskell language… so concise and clean.
Pure vs effectful code, yes. Pattern matching, duh! Lazily evaluated hunks, why not. Property based testing via QuickCheck (opens in a new tab), woah. And monads, sure.
Taught by Dr. Yao Li (opens in a new tab) who did an excellent job of keeping it fun and light while shepherding us through mind bending concepts.
For the final project… I… wrote… this story… about… Santa and… monad transformers…
Develop electronics simulation web application (SQL/C#/.NET/TypeScript/Vue). Take lead on client communication, meeting organization, feature definition & implementation. Assist with scoping additional work & project management.
A version control authoring model playing on the idea of "stacks" of changes but adds "types".
The unit of change is a trail which is simply a segment of commit history (or, a stack of changes).
A trail must match a "pattern". In other words, a trail is "typed". Types are configurable.
The model also includes a multi-boundary history. One boundary preserves an immutable trunk, the other end allows dumping commits onto the tip, and in between is a malleable middle zone.
The unit of change is not an individual commit, but instead the formation or amendment of a trail.
A collection of farms on undeveloped lots in the city.
Fun time working with a ever flowing mix of people on whatever needed to be done that day… Weeding, hard scaping, cultivating, harvesting, distributing, hauling, etc.
Part of a team which built a location-based city resource aggregator for individuals living on the streets.
My role was "validation": walking around downtown engaging with people. It seems I am naturally approachable and a strong listener.
We came in the morning of day two and a hotshot on our team whipped it up on Firebase overnight.
We learned most people knew about a lot of the resources, but trying to get into a shelter to get a bed was often times impractical. You might hear about an opening but show up to find out it's full.
From there, we refined the idea to focus on bed reservation. However, we then learned from shelters that this is not so easy. People actually need to show up in person first for a shelter to do their own dilligence to take them in.
In the final demo, people were awed by the software that was built. But I think what won us the competition is when, after being silent, I snapped forward in response to a question and began talking about everything we learned from people on the street.
As a naive engineer and trying to be frugal, I'd shop around… but my time usually never outweighed whatever discount I found or negotiated as a first time customer. It was essentially cost prohibitive to spend time comparing prices for a single prototype single quote.
