GSOC 2025: Upgrading Tour from google sheets and local storage to database with authentication

[AQIB-NAWAB]

Project Title

Upgrading JSON Schema Tour Website with a Database

---

Brief Description

The JSON Schema Tour website (https://github.com/json-schema-org/tour/) is a cool tool. Right now, it uses Google Sheets and local storage to manage data, which works but has limitations. This project is about moving the data to a proper database to make the website faster, more scalable, and easier to maintain. It’s a simple but impactful upgrade that will make the site more reliable and open doors for new features in the future like admin dashboard , generating stats and so on.

---

Expected Outcomes

1. Replace Google Sheets and local storage with a database (like PostgreSQL or MongoDB).
2. Move all existing data to the new database.
3. Build a backend API to connect the database with the website.
4. Add authentication and authorization to secure the website and allow role-based access (e.g., admin vs. regular users).
5. Make sure everything works smoothly and is well-documented.
6. Admin dashboard

---

Skills Required

• Frontend Skills: TypeScript, Next.js, JavaScript.
• Backend Skills: Node.js, Express, JWT (JSON Web Tokens).
• Authentication: NextAuth, Firebase, or custom authentication.
• Databases: MongoDB, SQL, or PostgreSQL.
• APIs and Git: Familiarity with RESTful APIs and version control.

---

Mentors

TBD

---

Expected Difficulty

Easy
This is a straightforward project. It’s mostly about moving data to a database, connecting it to the website, and adding authentication. Perfect for someone who’s comfortable with basic web development.

---

Expected Time Commitment

175 Hours
It’s a small project, so 175 hours should be enough to get everything done and tested.

[jdesrosiers]

Something to keep in mind is that JSON Schema is supported entirely by donations and that doesn't amount to much. The tour site is a static site and adding and a backend, database, and auth will mean recurring hosting costs. So, you'll want to include in your proposal the projected hosting costs for making this change so we can decide if it's worth spending our limited resources on this project.

The proposal is pretty vague on the details of what benefits we get out of this investment. For example, what specifically would be included in an admin dashboard that would justify cost of setting up authentication?

[AQIB-NAWAB]

Thank you for bringing up these important points! . We can keep costs low by using MongoDB Atlas Free Tier for the database and custom JWT authentication (no recurring fees). Hosting can stay free with Vercel/Netlify for the frontend and backend.

The upgrade brings:

Better data management (no more Google Sheets/local storage).

Admin dashboard to manage users, track progress, and update content.

Scalability for future growth.

Estimated costs: $0 (free-tier services). This is a small investment for a faster, more maintainable, and feature-rich site.

With the database in place, we can generate valuable insights and analytics, such as:

• User Engagement
• Progress Tracking
• Trends Over Time

These stats can be displayed in the admin dashboard and used to improve the tour experience and make data-driven decisions.

Why This Matters

Improve Content: Identify weak points in the tour and update them.

Engage Users: Use insights to make the tour more interactive and user-friendly.

Show Impact: Demonstrate the value of the JSON Schema Tour to the community and potential donors.

[jdesrosiers]

I'm a little skeptical we'll be able to use free-tiers. Free-tiers are generally for small hobby sites. I'm not sure how much traffic the tour site gets, but are we sure we'll fall under free-tier usage?

I'm still not seeing a solid argument for why we need this. It sounds like a solution without a problem.

Admin dashboard to manage users, track progress, and update content.

• track progress -- I'm pretty sure we're already tracking people's progress using something like google analytics. Which is going to provide a dashboard way better than anything we'd custom build.
• update content -- I don't see any reason why we need the site to have features to update content. It's maintained by developers who are plenty comfortable working in code and making PRs to update content. CMS style content management is usually included for non-programmers. That's not us.
• manage users -- I don't think we need logged in users and I don't think people are going to want to create an account to login to a site to go through a tutorial. I wouldn't. I'd just leave when I was asked to create an account.

Scalability for future growth.

I'm not sure the tour site needs to grow. It will need some maintenance and some content updates from time to time, but otherwise it does what it needs to do. Making it ready for some hypothetical changes in the future that might never come is not a good enough reason to make changes.

My suggestion is to redo your proposal from a different perspective. Start with a real pain point for users or maintainers and make that the focus. Transitioning to a database shouldn't be the goal, it should be part of the solution to achieving the goal.

For example, let's say that maintainers are finding it onerous to maintain lessons on the site and after discussion, it was decided that it would be better to have an admin portal where they can manage lessons. (I doubt that decision would be made, but for the sake of the example, let's say it happened.) Then you could make a proposal for an admin portal that includes transitioning to using a database and registered users.

The point is that your proposal needs to be addressing a real need, not just something that could be useful in the future. I think the only real need that you're presenting is analytics and I believe that's already handled. If it's not, I think it would still be better to use something like google analytics than to require users login and try to build analytics ourselves.

[JeelRajodiya]

Hello everyone, I agree with @jdesrosiers. I don't think the tour needs to be scaled further. in my opinion currently the tour is in pretty good condition (compare the tour with other similar projects). Moreover, The content is written in Markdown JSX, so if someone knows how to open a pull request on github, it is quite straightforward for them to request changes in the content. Adding a Admin dashboard just to update the content would add increase the work of the developers as well, it would be more continent to just update the Markdown files.

[AQIB-NAWAB]

Got it, that makes sense. I’ll rethink the proposal with a real pain point in mind. Thanks for the feedback!