Google Summer of Code reflection

9 min read
Celebrating 15 Years - Google Summer of Code

Google Summer of Code 2019 was a life changing experience for me. Thanks to Django, I learned a lot of things through the summer. Here are some things I'd advise if you're looking to participate this year.

1. Find your dream project/organization

This one may not be easy, depending on your interests. The organizations were announced a few weeks ago. It would be an advantage if you'd already been eyeing on some organizations. If not, don't waste your time!

Browse through the list and see what suits you. You can pick more than one organization, but I'd advise picking no more than two. It's fine if you submit three proposals, though.

For me, I had been looking at Django's JSONField idea since the previous February. I was lucky to stumble upon it.

> Some backstory

I took a Web Design and Programming course the previous semester. We used Django. It was the first time I used it thoroughly (I had finished the tutorial during the summer break but that was it). During the course, I encountered the need for a JSONField in my model.

Then I saw Django has it, only to find out it's only for PostgreSQL.

I deployed my course projects to Heroku and used the Heroku Postgres add-on. However, I only had SQLite locally. So, the postgres JSONField wasn't an option.

I wasn't interested in using a third-party package just for that, so I used a TextField and made use of json.loads and json.dumps instead. It worked, though not as simple as using a JSONField.

When I saw Cross-DB JSONField on Django's GSoC 2019 ideas list, I was hooked. I participated in Google Code-In back in high school and heard about GSoC then. I really wanted to participate in GSoC even before I attended university.

In fact, I also signed up in 2018, but I didn't submit a proposal since I couldn't find my dream project nor organization. As I was still in my second semester, I also felt my experience was lacking. Seeing the JSONField idea the next year felt like destiny.

2. Get yourself out there

Join the organization's communication channels. Mailing lists, IRC, Slack, Discord, whatever. Get to know the community. Find the people who are going to be mentors. Read the organization's contributing guide.

If you found an issue/ticket that you can fix, go ahead and fix it! Having contributed to the organization will obviously increase your chances of getting selected.

3. Start early

Start working on a draft proposal as soon as you can. The sooner you write one, the sooner your mentors can review it. Don't wait until the last minute. Feedback from your mentors will help you plan your project. They might give you ideas on how your project is best approached.

4. Have some confidence

You might think you're not capable. You might be having trouble writing your proposal as you don't really know how you're going to implement it. Don't worry. Do more research on your project. Ask your mentors and the community. Believe in yourself.

> Another backstory

Imagine a 2nd-year university student who's taking a Database Systems course. That person has never actually executed an SQL query directly into a client-server database terminal before that.

They've heard about


but they'd never done it themselves. Django's ORM helped them not to worry about SQL.

Now imagine that someone, aiming to implement a JSONField for Django that works on MariaDB, MySQL, PostgreSQL, SQLite, and Oracle Database. Yes, even Oracle!

To top that, they hadn't even tried the postgres-only JSONField, nor the other third party JSONField packages, nor the database systems other than PostgreSQL and SQLite. I probably would think it's unlikely to happen.

But hey! That person was me. As you can imagine, writing my proposal wasn't easy. I could imagine what the end product would look like. However, I barely had an idea of how I was going to implement it.

I kept reading the docs of those DBMS. I could understand what each system's JSON functions do on the database level. The thing is, I wasn't sure how to "connect" them with the Python code.

Given that the submission period was also around midterm exams, I didn't have a lot of time to do research. You're allowed to submit up to three proposals, but I only submitted one. It was all or nothing.

5. Don't just wait

If you've submitted your proposal, don't just wait. You're not the only one who submitted a proposal. Some may even submit more than one! Your mentors need time to do the review. Use it to show them you're the one.

Find more issues to fix. Even the documentation ones! You can also create new issues/tickets if you find something off. Just... don't just wait ;)

FYI: I submitted my first PR to Django during the student application period. I sent another one the next day. As you can see, they're both very small tickets, or "easy-pickings" as the Django folks call them.

However, I didn't stop there. Between the application period and the announcement, I also wrote a documentation ticket and submitted a PR to fix it. Again, it's an easy-picking, but consistency is the key.

Let's take a short intermission.
If you've made it here, thanks a lot for reading!
And, if you made it through the selection, congratulations!

Now, for the real deal: working on the project.

Patrick shivering

It was around 04:00 AM. I woke up to a chat from my friend. He said,


I immediately checked the GSoC website. I was very excited to look at the dashboard and saw that I got selected. I still couldn't believe it. I literally shivered. I also found out that two of my friends also got accepted. It was only the three of us from our university who submitted a proposal, so it was a big relief to see that we all got selected.

Once you've got through the selection, you'll code for the rest of the summer. However, there are some things I would advise if you want your journey to go smoothly.

6. Communicate

In GSoC, you work on a project, remotely. There will be obstacles as you go along the way. If you don't communicate with your mentors, they won't be able to help you. They don't even know what difficulties you are having.

So, please, communicate with your mentors. Use email, GitHub, Slack, whatever.

Don't disappear mid-project. You're a programmer, not a magician. (Unless...?)

7. Write a blog

Seriously, though, do it. Your blog posts will help you communicate with your mentors and the community about how you're doing with the project. Tweet a link to your blog. People do read them.

Don't worry too much about your English/writing. Despite some mistakes here and there, people still get your idea most of the time. Some of them may even provide ideas to help you.

If you have friends who are also participating in GSoC, that's great! You can share each other's progress and maybe ask them for help. If not, you can still write your progress on your blog and share it on social media.

Your blog will also be a part of your portfolio. Getting selected already takes a lot of effort, make sure you make the most of it for your career as well.

Also, the most important thing is, your blog will help you remember how you implemented things in your project.

8. Read the docs and code

Most likely, you'll be using a lot of tools to work on your project. Some of them you may not have used before. Don't be lazy to read. Read not only the docs but the code as well.

You're working on an open source project. If you're adding a new feature, it's likely that studying the existing source code will help you implement your project.

I saved myself from a lot of hassle by reading the implementation of other fields in Django's source code.

9. Have fun

It's your project. You're contributing to open source. People will use your work. You get paid when you succeed. Can it get even better?

Yes, if you make it a fun experience.

You'll learn a lot over the course of your project and that's fun! However, there may be times when it feels overwhelming.

It's okay to take a break for a while. Maybe play some video games. Watch some movies. Read books. Anything, really. It's summer, you can still have fun! Just make sure to relax responsibly.


All work code and no play makes Jack a dull boy.

Anyway, I think that's it for now. If you have any questions, feel free to ask me! If you want to see my journey, checkout my GSoC blog. You can also see my proposal and project page.

I hope you find this useful!

This post was previously written as a thread on my Twitter.