Hacker News Reader

Comment by austincheney | Hacker News Reader

I am curious how you come to that conclusion. Framework lust, particularly in the JS ecosystem, is inflated and often not representative of technical authority, experience, or sometimes even competence.

If I were a hiring manager and the candidate couldn't write very basic vanilla JS I wouldn't hire them either. I don't care that they have a bunch of github stars. The inability to solve simple problems without a framework is a liability.

View on HN

Github stars mean very little. It is free to give one and many people just star every repo they see.

The #1 repo by stars is FreeCodeCamp. I have not used the service, but if https://news.ycombinator.com/item?id=11911647 is to be trusted then the count is inflated:

> GitHub part is so flawed... because FreeCodeCamp inflate their popularity by requiring during onboarding for the user to star their repository yet keeping them at the top. It's like if I told everyone to vote up my Hacker News posts... and gave them the link to do it and GitHub has no remedy for that inflation.

I'm also guilty of starring repos that seem cool but have not read the code or tried to use it. I've never used D3 in a project and can't speak to its quality, but I've seen some wicked cool demos and starred it as a bookmark of sorts.

Free Code Camp is a massive outlier. It received most stars of any repo at 177,914 stars in 2016, and the second most (the google-interview-university which has appeared a few times on HN) has only received a fraction of that at 28,727 stars in 2016, which likely did not use as much growth hacking.

Full list of the Top 1000 Repos, freshly queried from BigQuery: https://docs.google.com/spreadsheets/d/11bGpZq6ixlhrmQnzEUqb...

While it is an outlier it is also proof that it is a poor system of measurement.

It is? I think the problem is how we (mis)interpret the data to mean anything more than what it actually says.

So what's the takeaway, "Repo stars are an accurate measurement of how many stars a repo has"?

I tend to agree that GitHub stars are a very dangerous metric to go off of. I tend to star repos that I think look "cool", often without trying to use them or even looking at the code at all. I've never considered it a damaging act, but I also didn't realize that some folks are encouraging others to make hiring decisions based off of GitHub stars.

Reply on news.ycombinator.com