A lot of people have ideas they would love to be able to build on blockchains, but a lack of money or development experience can be a deterrent. Fuelled by another article I wrote recently about the Hive Application Framework (HAF), an idea arose. As we advance, the vision for Hive appears to be focused on microservices (many independent features that come together to form a collection of functionality).
Hive Platform As a Service (HPAAS)
Platforms that make it easier for developers and non-developers alike to build applications already exist. You have Firebase by Google, which provides a platform of functionality and easy-to-use SDKs (Auth, Storage, Database). I've been an avid user of Firebase for years until Supabase came along.
Supabase is an open-source competitor to Firebase. It's a layer sitting on top of a PostgreSQL database (like Hive Application Framework), except it provides authentication and many other features that Firebase offers without the cost. You can host it and be responsible for managing the infrastructure, or Supabase offers hosting you can use (with a free tier).
Now, I see something similar for Hive. It's clear that @blocktrades sees Hive Application Framework as something more-developer-oriented, but what I am proposing caters to developers, tinkerers and others who might not be comfortable with server management and other more DevOps/system administration tasks. A Hive Platform As a Service offering could leverage HAF and extend it. It does not need to be entirely new.
What would HPAAS look like?
- Self-hosted or hosted subscription (perhaps a free tier)
— Revenue share - in situations where the user can't afford to pay for a subscription, a cut of revenue generated on a dApp could be taken to discount or bypass the fees - Authentication — easily allow people to log in to your dApps
- Database — custom tables, ability to manage tables via a UI
- Functions — allow actions to be performed based on blockchain actions
- A well-documented SDK that allows interaction with the features of the platform
And I am not writing all of this because I am too lazy to do it myself. I would love to build something like this in collaboration with others in the community. An initial prototype built using HAF would be quite easy. The other parts would require collective effort and thinking.
Thinking about it logically, you could build this platform using existing libraries. Hive Application Framework for the main database layer, Dhive for the SDK layer (extended to support platform specific features and operations). Between these two libraries, you would have the basis for the platform in its raw form.
Is there a demand for this?
I can get caught up in my little world sometimes when it comes to ideas. So, before anything is built, would anyone see the value in this and use it? Are there counter arguments against a Hive platform service offering?
The whole point of this would be allowing people to quickly prototype and build Hive dApps with as little code as possible. There would still be a code element to all of this, but for basic features like authentication very little code.