The number that I'm really interested in, are last commits more than 28 days ago. With this number, you can guesstimate how popular Next.js is with non core contributors and in the case of Next.js, it really is staggering.
The basic idea is, if somebody last committed more than 28 days ago, it is likely that they are not a "core contributor" since you would expect "core contributors" to commit more frequently than every 28 days. Hope that makes sense.
For Next.js, the percentage of people that last committed more than 28 days ago accounts for 80% of all unique contributors, which is what makes their project quite staggering from a "popularity" standpoint. Assuming everybody that last committed more than 28 days ago are not "core contributors" (which may not be accurate of course) we can see that Next.js had 169 contributors that basically decided to create a pull request that was good enough to be merged. For Redwood they had 23, which is respectable but not on the same level as Next.js
Based on my analysis of very popular open source projects, most projects typically only have 50% of their contributors in the +28 days range, while Next.js has 80%. Looking at last commits more than 28 days ago is a very good indicator of community health/interest.
I've been living and breathing this project for over two years. From an "is the project healthy" perspective, your explanation of what you interpret from these analytics makes no sense to me. I suspect there's some pre-existing bias as well. (I.E. yes, Next is a very popular front-end React framework and it's been around for a few more years.)
Note: I love analytics. I just think this is, well, weird reasoning from the face of it.
I'll counter with this — here are the metrics I focus on as a measure of "is our open source project healthy"?
- How many unique contributors are there per release? && Is the number growing, steady, or decreasing?
- How many new contributors are there on a monthly basis && Is the number growing, steady, or decreasing?
I agree and I wrote "the project is quite healthy", so I don't quite know why others would interpret it in any other way. I also agree with the metrics that you laid out and those metrics will be surfaced in the future along with others, since numbers need to be put into context. For example, how are people contributing. Are they making simple spelling changes? Are they making changes to core languages (i.e. Typescript vs Markdown, etc.)? Are they adding a lot of new code and so forth.
Something worth noting is, the analysis is based on Pull Requests that were created in the last 4 months for both Next and Redwood and as it currently stands Redwood had 55 unique contributors and Next had 210. By all metrics, Redwood is quite healthy (as I stated in my comment), it is just that Next has a lot more contributors (4 times more and it something that can't be ignored), especially with the number of contributors in the +28 days range.
Thank you for the reply. This is an interesting topic. I think at the core a comparison to Next is still apples to oranges imho and unnecessary here. But that's probably a distraction from the concepts at hand.
> Are they making simple spelling changes? Are they making changes to core languages (i.e. Typescript vs Markdown, etc.)? Are they adding a lot of new code and so forth.
I think what you're trying to get at here is answering the question, "Is PR A of greater quality or value than PR B?" I've decided that doesn't matter as much as:
- How easy is it for anyone to open and merge a PR of any kind?
- Once someone opens a PR, do they open another?
- What's the preexisting skill level of someone opening a PR relative to the difficulty of the PR they're attempting?
Trends. Over time. I can't tell you how many contributors start with simple doc updates (which, by the way, are as important as code for the Redwood community) and then follow up with a monster PR. Momentum builds momomentum. Collaboration leads to more collaboration. For the most part, "is A better than B?" is a distraction.
Again, all from my experience and not demonstrable otherwise.
> 4 months for both Next and Redwood and as it currently stands Redwood had 55 unique contributors and Next had 210
One of the reasons this likely feels low to me (in the case of # of Redwood contribs) is that we separate concerns across repos. Is this for the org or the repo?
It would be impossible to find an accurate number, but I'd be much more interested in the ratio of (unique contributors) / (project installations or downloads) and how that ratio trends over time.
For a metric like that, how might the Redwood vs. Next comparison play out?
You are welcome and your response has been extremely helpful as well.
> I think at the core a comparison to Next is still apples to oranges imho and unnecessary here.
I agree it is probably an apples and oranges thing from a tech perspective, but clearly this isn't the case from a marketing one. I responded because somebody had concerns about the health of Redwood compared to Next.
> "Is PR A of greater quality or value than PR B?"
This isn't really want I want to do. Understanding how a PR affects a project isn't so much about quantifying value, but rather, it is about quantifying effort. A single line change can take takes days or months to get right and I want to better capture the effort required to make a change.
The questions/metrics that you raised are quite good and it is something that I will be working with Open Sauced (which you know I believe) to surface. The pattern that you've described is something that I wish to confirm as the goal is to find a way for maintainers to easily identify contributors worth nurturing and for contributors to find projects worth their time to contribute to.
> Is this for the org or the repo?
Below is a combination of all recently active repos from Redwood
Redwood had 79 unique contributors with one or more merged pull requests and Next had 214 unique contributors with one or more merged pull requests. If we ignore the merged criteria, Redwood had 96 unique contributors and Next had 350.
Looking at merged pull requests, the project is quite healthy as the following shows:
https://oss.gitsense.com/insights/github?q=merged%3Atrue%2Bp...
Next.js is on a whole new level though
https://oss.gitsense.com/insights/github?q=merged%3Atrue%2Bp...
The number that I'm really interested in, are last commits more than 28 days ago. With this number, you can guesstimate how popular Next.js is with non core contributors and in the case of Next.js, it really is staggering.