← back Home 🏠

Software development

Fri Mar 23 2018

(Disclaimer: this page is a bit dated - you know, technology changes...)

What is my favourite platform you may ask. Open Web Platform I'd reply. Web is an amazing medium that is free to use to everyone. No vendor lock-in, no proprietary tools. On the other hand, it suffers from the fact that it was constructed as a network for exchange of documents. It was just simple hypertext back then. Times have changed. Today's internet users expect smooth, interactive yet reliable tools. It's challenging to deliver those, but pleasing when done properly.

What am I up to?

Professionally I've spent last five years or so working as a software developer for several companies. To name a few - FNZ, Adacta or most recently Snapshot where I also tried to implement some basic user-centered design methods. That was quite fun and eye opening. I'd like to do this more!

During those years spent as a software developer, I had a chance to get in touch with some incredibly skilled and open-minded people. It was also excellent opportunity to get my hands on useful technologies in web development space. To name a few:

• JavasSript - Web's lingua franca these days. Hard to get stuff done without JavaScript. Recent ESx flavors make it more and more versatile and accessible tool in your toolbox. Slightly suffering from backwards compatibility though. - React - React is awesome, right? Yes, Vue is steadily growing, Angular has a huge community, but counting its pros and cons, I believe React still wins in frontend world. Feel free to prove me wrong.
  • Redux - It is hard to build apps without them having state. React's heavy inspiration in Elm made me interested in Elm lately. Other than that, Dan Abramov has made a great yet simple to understand tool for React ecosystem. Kudos for that!
  • Typescript - Type safety is nice. Typescript can help you prevent dumb errors, but also slow you down. Depends on optimally defined rules. At first, I was not a big fan, recently we have a good relationship.
  • Also got in touch with other tools like Webpack, Reselect, Storybook, Jest for testing, postcss, npm and others.
• .NET land - in all companies I've worked so far, backend was written in .NET. Last two years I've spent working with .NET Core. I aim at writing SOLID backend code, trying to keep minimum of logic in controllers when using MVC. If I could choose between frontend or backend, I'm all for frontend work though.
• Responsive webdesign - feels like a proper mindset when building web these days. It is crucial to deliver inclusive designs in the world of growing device fragmentation.
• Progressive web apps - Again more of a mindset focusing on app interactivity and responsiveness. Every second wasted rendering, data fetching or parsing app content is wasted time for users. Or worse - we might lose them. Strong emphasis is put on user experience and is aligned with my core values in software development. Progressive web apps help delivering content even when users are suffering from lie-fi or do not have any connection at all. I see it as a way of progressive enhancement to a standard online/offline duality.
• Elm - I've been digging into it lately. Having no null value is so comfortable. Also twisting your head into functional paradigm is an interesting endeavour. Elm architecture (variation on model-view-update) seems like an evolution of well-known MVC pattern. Smart Elm compiler is also really helpful tool with great developer experience.

How about the process?

Is waterfall dead? Is agile dead? I hope not. Both have their place and both can be successful. Hopefully we can take the good out of each methodology and use it in our advantage.

Generally speaking - cross-department collaborations in combination with self-motivated individuals seem to make a difference when building well tailored products. Diverse feedback that you get from people with different backgrounds is priceless when iterating on some concept. It might take time to process it and learn from it. If your aim is to build superior products, it's inevitable. Inclusion of end users in creative process is preferred. This can give us instant feedback. The speed of feedback loop is important factor of being able to pivot product well.

I did mention iterating previously. That's because it is an important part of incremental way of making software. Fast (not rushed) prototyping helps to find a proper way to move forward. If we find out we have made a bad decision previously, taking a step back is not that expensive. Learning to throw away is important skill to learn in order to succeed. Platonically falling in love with imperfect prototype can easily lock peoples thoughts to some specific concept.