In 2027, at least 7 out of the top Tiobe languages will be Tree Languages | Hacker News Reader

In 2027, at least 7 out of the top Tiobe languages will be Tree Languages

longbets.org

2 pointsbreck8 comments

Unlikely. Languages younger than 5 or 10 years barely make the list, and the top languages are in the 20 years old range on average.

Language adoption takes a really long time, and all new languages are at a disadvantage because older language usage continues due to inertia.

That's the safe bet for sure. I'm predicting a paradigm shift. An interesting thing is that 6 of the current top 10 (Java, Python, C#, VB.net, Javascript, PHP) came out between 1991 - 2001, during the web paradigm shift. The 2004 list had 5 of those. So on rare occasions you do see some seismic change that catapults new languages to the top unusually quickly.

I agree that the web fueled the rise of those languages, but I disagree about the reason.

The web was a paradigm shift, but it also attracted a lot of beginners. You can see the effect particularly in JavaScript and PHP.

The less complicated it is to write and run a language, the more beginners will use it. Many times, your first language will become your only language (i.e. your Blub), and the inertia of that language's users will increase.

We've seen the same thing again with statistics and ML: Python is rising in popularity because a lot of beginners are taking boot camps and getting quantitative degrees, and they're learning Python.

The question I'd ask you is: what sea change are you expecting that will bring in a large number of beginners? And why will they learn (or be taught) a tree language?

This is a great perspective, thanks.

> what sea change are you expecting that will bring in a large number of beginners?

I think a number of kill apps will arise that exploit Tree Notation. Let me mention 2 teams I spend a lot of time with.

One is working on a paper for a new system for EMR where healthcare providers/clinicians/researchers use the same syntax for the whole life cycle of medical data. So nurses in Bolivia might write intake notes on pen and paper in an EMR Tree Language; doctors in India may dictate notes into an EMR Tree Language; patients may copy/paste their entire medical records from one provider to the next during an acute care visit; and researchers at the end of the pipeline might get their data in a clean strongly typed Tree Language before loading into a deep learning training model. https://github.com/treenotation/pau

The second is a new language for data science that will make inroads amongst the Python/R data science crowd as well as pull Excel users over. It's a language and runtime that runs in your browser on your computer or your smart phone, and does everything locally. Nothing to install, no funny syntax to learn, the world's clean data will be at your fingertips, easy to share EDA results. https://github.com/treenotation/ohayo

The WHO estimates there are about 40M professionals in the healthcare industry. I've seen estimates that there are 750M Excel users. None of the TIOBE languages exceeds 20M. So I think either of those 2 alone could be bigger in 2027 in terms of user base than any language on the TIOBE top 10.

But I expect dozens of giant killer apps that will be built by startups/teams that I have no involvement with. People are exploring it in the legal space, crypto space, web form space, educational programming space, gaming space, and on and on. The ideas are there for the taking. It's going to be like the 1990's web boom all over again.

> And why will they learn (or be taught) a tree language?

Simplicity is timeless. Binary Notation is the dominant notation of our day. It's 330 years old. All of the TIOBE top 10 are under 50 years old. Binary Notation can't get simpler without sacrificing a lot of functionality. Likewise, AFAIK Tree Notation can't get simpler without sacrificing a lot of functionality. It's more akin to some universal pattern in nature than it is to a man made invention. Now, I'm not saying it's perfect and odds are high that I have screwed something up royally (you can see the earlier versions starting around 2011 called Noted and then Space to see previous critical flaws I had), and that's what I'm hoping to be the outcome of this Long Bet: someone will point out some critical flaw that I haven't seen, or some people will try, fail, and then see the potential and jump on board the train.

Reference for "tree languages" https://arxiv.org/abs/1703.01192

Don't you mean https://en.wikipedia.org/wiki/Entish ?

This is an interesting idea.

Isn't Python a tree language? Whitespace is significant...

How about YAML?

Good question. As I define it no, a tree language uses white space and only white space for syntax. While both python and yaml use white space for a portion of their syntax, the bulk of their syntax is traditional BNF.

Reply on news.ycombinator.com