Oops – I have not been blogging since declaring “a new beginning” for this blog. Having figured out “serverless” as a way to make blogging more manageable, it has also made me wonder whether the effort I put in was worth it anyway.

Let’s face it. Original writing is difficult and time-consuming. Would I rather be the one who reports on stuff quickly? Then it would be more worth it to curate a pretty awesome Twitter account. My favourite posts this year have revolved around original writing. Frankly, I would rather be programming than writing. Having tracked nearly 1000 visits in 2020, it would appear that the audience agrees too.

If I were to look back at my 2019 post, it feels like I focused a lot on the first two resolutions and still had a difficult time with resolution 3. I’m not a resolutions kind of guy, but let’s go for it anyway.

1. Write something about the accessibility of legal materials online in Singapore.

While writing pdpc-decisions, I thought it would be great if a single, humble website would serve legal materials as data. After spending several months trying to convert PDFs into recognisable text, this goal looks much more complicated than expected. While I am doing this as a side concern and not a job, I will not have the time even to make a humble API.

The next best thing I could do is to write about it. I guess a single post regularly updated to reveal sources of legal materials and how they are scattered around the internet will be helpful. Someone can pick the pieces up, and we might just make that humble API.

Achievability rating: * * * * / 5. Not the website, just the write-up.

2. Read better books on data analytics/programming/statistics for a guy who reads fast but has no time.

I am having a lot of trouble finding a book I like these days. I have had my eye on “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow”. I think it’s the best book I have browsed so far that balances theory and practical pointers. 2020 was a year of hands-on programming, but my data science journey has been limited by my imagination and lack of theoretical foundation. Short of attending a university course, I think locating and reading good books is the only way forward for me.

Achievability rating: * * * / 5. Limited by capacity restrictions at all major bookstores around me and my patience with Amazon’s web store.

3. Release a POC version of my secret project, fairest

It might be called proof of concept, but in my mind, it’s mucky. I have not talked much about it here because I was busy writing the code. Fairest is a glorified spell-checker for legal and legal type documents. The general structure of the code seems to work. I do get reports claiming that my 80-word sentences are way too long.

Still, fairest needs to be pretty. Getting the API up and running with FastAPI was fun, but now I have to brush up my HTML skills for the frontend. To be honest, I am not looking forward to writing Angular again. To be more open, I am dreading CSS.

Achievability rating: * * / 5. Doomed by my self-deprecating sense of humour.

So that’s it for 2020. Now let’s get going to 2021. I can’t wait.