Thoughts on TIOBE index and the future of Rust

I remember Rust making headlines when it broke into the TIOBE Index Top 20 list for the first time in June 2020.

I noticed today that Rust is back on the Top 20 for Oct. 2022, however there are no headlines this time. I remember checking the index a few months ago, and Rust was not in the top 20.

I would like to have your thoughts on how this affects Rust in general, i.e. its future adoption, language development, etc.

I don't think anybody serious in the industry cares about the TIOBE index. To me at least it looks like noise.

(they give themselves a serious sounding name "The TIOBE Index", it sounds authoritative but it's really just nothing)

1 Like

It's as precise as a horoscope. They show popularity with two decimal places, but don't have even a single digit of precision.

  • According to TIOBE, C has lost 60% of its popularity between 2015 and 2017, and then doubled next year. This just can't be true for a half-a-century old stagnated language.

  • TIOBE thinks C# had a 6-fold popularity increase in Apr 2022.

  • In TIOBE TypeScript is 40th, behind Prolog, Visual Fox Pro (a niche proprietary language that has been dead for 14 years), Dart, and Rust.

It's just wrong data. It's a poor unstable proxy (subject to proprietary query interpretation and anti-spam algorithms that vary over time, query terms don't reflect colloquial names, and name mentions aren't usage), and it's a misrepresentation of the data (searches include historical documents, which is not a reflection of current usage). It's all noise, no signal.

All models are flawed, but TIOBE is beyond useless. If you need something, RedMonk is somewhat sensible.

15 Likes

Interesting. Maybe I have to change the title to better reflect what I would like your thoughts on.
I understand that you (and many other) consider the TIOBE index bad at estimating a language's popularity. I believe you are right.

However, my question goes beyond that. Certainly, the fact that Rust was in the headlines due to the TIOBE index, may have had an impact in the Rust ecosystem (more awareness among younger programmers and decision makers, etc.).

As @erelde said, in his experience anyone "serious in the industry" does not care about the index. But there may be many decision makers that are not as "serious", for which the TIOBE index "sounds authoritative" and thus influence their decision of adopting a new language or not (just an hypothesis of mine, without any data). Anyone has experienced this or something similar?

I'm tracking data from crates-io index, and I'm not seeing spikes in Rust's popularity (beyond false positives from spam). It does have a steady growth:

openhub shows it pretty steady too (with caveats: their last month data is noise, and they're heavily biased towards open-source).

So I think at this point Rust is popular enough that no single headline is enough to affect its popularity measurably in a short term.

8 Likes

I would think those people (financiers, investors, etc) are so far removed from the distinction between "programming languages" and the software they use or invest in, it doesn't make a difference.

They read an article somewhere telling them "Python is popular" they get some understanding that something that makes their computer run is a popular tool for some engineers somewhere.

:point_up_2: this.

Also, I use Rust because I like Rust. The only time I care about its popularity is when it comes to finding a new job, but my current job is mostly non-Rust anyway and I like it, so it's not even that much of a factor. I also know that Java, C#, PHP and JavaScript are very popular today, yet I don't want to use them. So from both directions, the particular value of the TIOBE index has approximately zero effect on my decisions if I'm allowed to choose the language. I imagine I'm not the only one.

I wouldn't be worried about the future adoption of Rust. It's already a major player, its benefits can't be ignored anymore, and several big companies are investing in it. Even if Rust ultimately doesn't make it into the list of the 20 most popular languages in the long run, it'll stay with us for a long time.

7 Likes

That page is fascinating, thank you for creating such an excellent overview!

I also love the description for crypto magic beans linked to from the "Categories" tree map :joy:

3 Likes

Around 10 years ago, not knowing better, I thought the index was a real/accurate snapshot of the programming languages landscape. Still, it was nothing more than an interesting talking point among (also clueless) colleagues. I guess it is just nice to see a language you like being "popular" (at that time I was rooting for Python to become more popular).

Around 2019 I was leading a small team of embedded developers, and I was seriously considering switching the team development from C to Rust. In the end, I decided against it. The TIOBE index had nothing to do with my decision. :rofl: So yeah, I tend to agree with you guys.

2 Likes