Coming online
Administrator - 11/01/2026
I built originally built KUUT (My initials as KWT) as my first .NET project yo basically learn as much as I could about .NET Core 2 and as a place to keep the code, electronics, models, experiments and occasional strange idea that do not fit neatly into a conventional CV. It is a kind of CV though I have yet to add it in. I felt seeing my work running was as good as a list of languages and my history of jobs. It took me a month end to end to exhaust/absorb most of Core's main feature functionality, some parts were unnecessary/not great approach wise but it was a much a blog as it was can I just build a CMS/Auth/Search/Typescript/DI/async/EF Core/singleton toting site and produce something functional as a learning exercise. coming back to it I revised several parts and moved to .NET 10.
Tags:
.NET C# TypeScript Three.js CMS portfolio electronics 3Dmodelling
Retrospective on a site first built in 2022. Project period: June 2022 to present.
I built this site because I wanted somewhere of my own on the web and to learn. I have never been particularly interested in social media, and a conventional CV is a poor container for most of the things I do. It can list C#, electronics or 3D modelling, but it cannot show the strange bit in the middle where those subjects collide and turn into an actual project.
KUUT started in June 2022 as my first serious attempt to learn .NET and C#. Building a blog would have been enough. I decided to build the content management system as well, then added users, roles, comments, search, dynamic routes, metadata and an animated logo because apparently I do not recognise the phrase “small learning project” If it's written in block caps impact font on a lorry outside my window. Even the animated to logo was pure math in 3js and a little more excitable than the current subdued one.
That decision made the site much harder to finish, but it also made it useful. I was not following a tutorial towards a known result. Every new feature forced me to work out how another part of a web application fitted together.
Learning the stack by building the stack
The back end began with ASP.NET Core, C# and Entity Framework Core. The content system uses a shared model for posts, projects and pages, with common routing and search behaviour around it. LINQ made it possible to keep much of the query code readable while still composing different filters and relationship depths. Dependency injection tied the services together without turning every controller into a cupboard full of global state.
I also built an attribute-based search experiment. Searchable properties are marked in the models, converted into a compact index surface and resolved back to their content routes. It is not how I would build a large search platform now, but it was a useful way to learn reflection, generic code, serialisation and the consequences of designing a data format before you fully understand the problem. There are lessons in the code as well as features.
On the front end I used TypeScript, Webpack and Three.js. The original logo animation constructs its geometry in code, loads the SVG mark into a Three.js scene and moves the parts with timed tweens. It was excessive for a header. It was also far more educational than replacing it with a GIF.
The site has its share of scars. Some early abstractions are more enthusiastic than elegant. A few comments record the exact moment I discovered that the clever solution had become the problem. I am leaving enough of that history visible to be useful, while improving the parts that affect security, reliability and maintainability. A portfolio that only shows immaculate final code says less than one that can explain what changed and why.
Coming back to it
After the initial 2022 build, the project sat quietly for a while. I returned to it in 2026 and brought the runtime and deployment work up to date. The current application targets .NET 10, builds its TypeScript with Webpack 5, uses SQLite for local development and can run as a Linux container for deployment.
The CMS has also grown into something closer to the site I wanted in the first place. It can handle structured article content, authenticated asset uploads and public delivery through stable asset identifiers. There is a local Three.js viewer for geometry-only OBJ and STL files, so it can display my 3d modelling work
Project is still on going. I've come back to it, understood the decisions made with no experience, preserved the parts that work/are well built and replaced the parts that I know now to utilise methods that aren't so good. It's been educational doing it.
What belongs here
My work tends to spread across disciplines. I write software in C#, TypeScript, JavaScript, C++, Java and SQL. I build and modify FDM and resin printers, computers, electronics and photographic systems. I spend an unreasonable amount of time thinking about and building realising solutions to power electronics, electromagnetic interference, AI, psychology, physics and how a mechanical detail changes the behaviour of the whole system.
Some of the things on this site are practical engineering/electronics. Some are software experiments. Some are models built because a friend asked for raptors with laser beam eyes and did not expect me to take the request seriously. They all show the same habit: I want to understand how a thing works, then I want to make a version of it. There's also just parts I wanted to publish like a journal.
This is therefore part blog, part workshop and part extended CV. The articles will include source code, measurements, circuit diagrams, interactive tools and 3D models. There are some failures, some successes. Most often they're just
anything, anywhere with interesting and complex engineering, and being a blog their are some which are just thoughtful.
The first version of KUUT was how I learned .NET. This version is where I document what happened next.