Why We Built Skygear

Developed by Oursky, Skygear is an open-source severless multi-cloud platform for secure enterprise apps.

In this blog post, we would like to share our vision for Skygear, why you should consider using it in your next web or mobile project, and why you might want to join us in developing it, either as an open-source contributor, or join Oursky — we have office in Hong Kong and Taipei, and open for remote position!

What’s our vision for Skygear?

We believe Skygear can help the world deliver better apps because:

• We need a modern “Rails for serverless” framework.
• We need a cloud vendor-neutral serverless solution.
• We need more security and privacy.

1. We need a modern Ruby on Rails.

Although Oursky uses mostly Go, Python, and Typescript for back end nowadays, we used to develop in Ruby on Rails (RoR) a lot 10 years ago. It is very opinionated, and you can do typical functionalities swiftly and easily, such as user signup, login, and forgot password. Other examples are using Paperclip to upload images to S3 and resizing it, and using a simple config on models for a full text search.

That’s exactly what we need in the serverless age. Even for developers who don’t embrace cloud functions and microservices, most of us can’t go back to RoR as the mainstream is doing single-page application (SPA) or mobile apps. The world is re-developing all the typical functionality we need in Node.js (or your favorite frameworks). We don’t have something like Rails for SPAs and mobile apps that use serverless frameworks. We envision Skygear to be the “Rails for Serverless.” It is a long road to go but we will persist.

2. We need a cloud vendor-neutral serverless solution.

While the next “Rails for serverless” is not coming anywhere near mainstream adoption, all giant cloud providers like Amazon Web Services (AWS), Google Cloud, and Azure are doing it within their own ecosystem. Why? Because they want customer lock-in. If you use AWS EC2 only, it is fairly easy to port to Azure VM. If you use AWS Lambda, Cognito, API Gateway, and DynamoDB at the same time, it would be painful to rewrite everything if you want to switch to other providers. Lock-in is what each cloud vendor desires.

Software should have freedom. Years of developers’ work should not get locked in with vendors, and closed-source solution is something against our ideology. We believe that developers should be free to choose vendor services that best fit their projects, and be able to switch whenever necessary. That’s why we open-sourced Skygear from day one.

3. We need more security and privacy

We believe security and privacy are one of the most important issues today, and we need different sectors in society collaborating to address them. From a technical perspective, we wish to contribute by making it very easy for developers to build secure software and respect user privacy whenever they need. 

“Security by design” is hard. It is very difficult to implement authentication system properly, and it’s no wonder we’re hearing about many incidents of passwords being exposed in plain text or weak hashing algorithms. It’s even harder if you want to respect privacy by having an end-to-end encryption storage system.

We take this as the ultimate mission of Skygear. We’ve started with a good auth system, and in the long run, Skygear will be capable of:

• Integrating with DevSecOps and make DevOps and CI much easier.
• Coming with “gears” (services of Skygear platform) and client-side software development kit (SDK) for end-to-end encryption storage or message. 
• Detecting abnormal behaviors, such as abnormal login attempts and traffic detection

Skygear is admittedly far from what we described above, but we have to start somewhere and one step as a time as a small team. As a company, we wish to become a sustainable business with an open-source product. Finally, if you want to solve these problems together, you are welcome to join us or contribute to the Skygear project!