Nadia Eghbal has done extensive research on open-source and has written many articles detailing her findings. One specific analysis that I found interesting was where she analyzed the health of Web 2 open-source projects using objective metrics.
So, we asked ourselves…“What if we could apply those metrics to Web 3 open-source projects?”
And so we did.
The rest of the post will share our findings and conclusions.
PS: Her research is extensive. If you’re interested, I recommend reading her book.
Before we dive into the data, we want to clarify that rather than looking at ALL Web 3 projects, we analyzed the data for the top Web 3 projects. “Top Web 3 projects” is defined as the top 5 - 10 projects from the top 5 blockchain ecosystems: Bitcoin, Ethereum, Solana, Cosmos, and Polkadot.
These projects include:
Now, onto the data!
The first metric we explored is “Magnet vs. Sticky.”
We classify projects as attractive (highly magnetic and highly sticky), stagnant (highly sticky, weakly magnetic), fluctuating (highly magnetic, weakly sticky), or terminal (weakly magnetic and weakly sticky)
In other words, a project is high on the Magnet and Sticky scale (i.e. “attractive”), it attracts a lot of new contributors, many of whom keep contributing over time. This is a positive indicator of health.
On the flip side, if a project is low on the Magnet and Sticky scale (i.e. “terminal”), it doesn’t attract many new contributors and has a high churn rate. This is a sign of poor health.
Before we dive in to the larger data set, let’s look at a subset of data to make sure you understand how to read the Magnet vs. Sticky chart.
Here’s a Magnet vs. Sticky chart for development frameworks on Ethereum:
We can learn a few things from this chart.
If we look at the stickiness scale, we see that Hardhat is slightly more sticky than Truffle. Ape, on the other hand, is VERY sticky, but it doesn’t attract nearly as many new developers as Hardhat. What explains this?
Here, we see that Ape has a much higher ratio of existing vs. new contributors. This explains why it’s on the far right of the sticky scale: a few developers consistently commit to the project.
Now, let’s expand the scope and look at the Magnet vs. Sticky chart for the top projects in each of the 5 ecosystems defined above.
Note: We based the charts on contributor activity in the last 6 months.
Looking at the chart, we can glean some interesting insights:
What other insights do you see in this chart? We’d love to hear in the comments!
Next up, we classified each Web 3 OSS project as Federation vs. Clubs vs. Toys vs. Stadiums.
Let’s first define our terms:
Federations are projects with high contributor growth and high user growth. These tend to be highly impactful projects in the ecosystem. For example, federations in Web 2 include Rust, Node.js, and Linux. Given the high contributor growth and user growth, federations are more complex to manage from a governance standpoint.
Clubs are projects with high contributor growth and low user growth. Examples of clubs in Web 2 include Clojure, Haskell, and Erlang. While not a lot of people use these languages compared to popular ones like Python, JavaScript, and Java, they are useful in certain niches. Clubs may not have a wide reach, but they’re loved and built by a small group of enthusiasts. They also tend to be more selective in the new contributors they accept. If a federation is a city, a club is like a small town.
Toys are projects with low contributor growth and low user growth. They’re like side projects — they have the potential to be widely used, but as of now, they aren’t. Examples of toy projects are open-source GitHub projects with less than 10 stars.
Stadiums are projects with low contributor growth and high user growth. They tend to be powered by one or a few developers. Examples of popular stadium Web 2 open-source projects include webpack, Babel, Bundler, and RSpec. Unlike federations, stadiums have just one or a few maintainers to make decisions on behalf of a broader user base. That means the governance of stadiums is more centralized than a federation.
Source: Working in Public
Projects can certainly move from one quadrant to another – clubs can mature into federations, and toys can grow into stadiums.
Now for the fun part! Let’s take a look at which quadrant various Web 3 projects fall into.
Above, you can see the distribution of the top 150 crypto projects across the 4 quadrants and find that they’re more or less projects in every quadrant.
Now, let’s look at the same chart but with just the top projects in the top 5 ecosystems (see table at the beginning of the article).
Here are a few insights from this chart:
See any other insights in the chart? We’d love to hear about it in the comments!
We hope this data provided some new and interesting insights about the health and nature of Web 3 open-source projects. If you have any questions, reach out to us at preethi@dappcamp.xyz!