The application phase of the Google Summer of Code is now closed. We will be reviewing submissions over the coming weeks, and final announcement will take place on May 8th.

In the meantime, here’s a first look at this year’s batch:

A sankey diagram showing the distribution of applicants. 74 Males, 11 females. 55 applicants from India, the rest coming from 15 other countries around the world. More than half did not meet the mandatory requirement of submitting at least one pull request. Of those that did, distribution is pretty even across projects, with 15% submitting their own idea.

 

  • We got twice as many proposals this year compared to last year
  • Once again a majority (51/90) did not include a list of contributions, even though we made it very clear that this was a mandatory requirement ಠ_ಠ
  • A few people submitted not one but two proposals (hence the jump from 34 candidates selected to 39 eligible proposals).
  • Overall quality of code submitted was pretty good!
  • Very little AI-generated BS overall (some used it to improve their structure or wording, but that’s fair game). Despite what you can read from “thought leaders”, AI code is not that great.
  • We had six « official » ideas on our project page, but made it clear that we were open to other proposals. It was nice to see that we did get some out-of-the-box thinking (about 15%, slightly more than last year).

One interesting / surprising feature of the above chart is how much Indian applicants make up more than the rest of the world combined. This may be a research question by itself, but early feedback tells us that there is a big culture of competitive programming in the country, along with a whole ecosystem of Youtubers teaching code to their followers and explaining the value of open-source as a way to practice and learn.