Wow - long time since I’ve written anything. Mostly because, with the world ending and all, I did not feel like anything I had to say was fully formed yet. Everything needed a little time to simmer.

I’ve also grown disatistfied with all the hot-takes flying around. So I decided to just shut up for a while (not that I have been a particularly prolific blogger in the before times).

However, I recently had someone try to bait me into the whole tidyverse vs. base R discussion, and I realized my current thinking was worth writing about.

So, which is better: tidyverse or base R?

Honestly, I don’t care anymore. Use whatever you find useful.

What do I use?

I use base R. I am proficient in it, and I can do everything I need to in it. I am generally more efficient than most other R programmers I know. The people I know who are better R programmers than me also use base R. Does that mean base R is best? Again, don’t care.

In particular, I use base R in Emacs. My workflow is very minimal. It works for me.

But what about <insert function/feature/bug>?

Again, what I am doing works for me. If you do something different, and it works for you, great. Our needs might be different. Or they might be the same. Either way, as long as we both are able to do what we need to, it’s all good.

When I find that my workflow does not handle what I need to do efficiently, I will change it.

What if you can do X,Y,Z better than me?

I am happy to show you what I know.

So, you are just set in your ways and resistant to change?

Not at all. I devote a large chunk of my time to learning new things. I am constantly adding to my data science toolbox.

That does not mean I am constantly chasing the hot new thing.

I am getting annoyed of being told how I should be doing things by people who, frankly, don’t know much. I cannot count how many times I have been told “You should do X because <insert famous person> said so.” Or because some random blog/Medium post said it was the best. Often when pressed, people cannot tell me the principle behind their recommendations. Typically, there is very little understanding.

It gets old.

Why the “new” position?

I build furniture in my spare time. I work 100% using hand tools. This system works for me.

People who use power tools have never-ending opinions on this. But, it turns out that unless they dip their toes into handtool work, their opinions are pretty much meaningless. They don’t understand what they are critisizing. Entering into a debate with them is pointless. An exercise in frustration.

But how the hand tool community deals with this, on the whole, was inspiring. The criticism is met with a big collective shrug before getting back to work. Why? Because what matters is the end product: Is it well-built? Will it stand the test of time? If yes, does it really matter how it was built?

The same is true with this silly coding dispute. I tried that other way - it did not work for me. I found it fragile and frustrating to use. It works for other people. That’s fine. It doesn’t work for me.

I would rather focus on what we are building.