Pragma

• There were many bugs in v0.1.0, we fixed them in this version and we added
• You can now install Pragma using Homebrew on macOS
• You no longer need to install Java to use Pragma
• Wildly improved automated schema migrations

Todo

Further improve automated schema migrations by allowing users to specify type transformers (JS functions) when they change the type of a field that has data in the database.

After 14 months and a half of designing the solution, and writing code we finally released the first functioning version.

It took us ~14 months because when we started we didn't have much experience in writing complex software other than web and mobile apps.

We started writing the language in Rust, and after 3 months we started rewriting the whole thing in Scala because while Rust is performant, it didn't allow us to move/iterate fast enough.

We spent a lot of time on making sure that the code is clean, and changes to it in the future are quick and easy. Because of that now we can implement complex features and fix bugs very fast!.

We've learned during these 14 months more about software engineering than we could've learned in 10 in a normal software engineering job, because everyday there is a new set of challenges.

I'm really glad that we made it this far, now we need to get as many developers as we can to build the future with this language.

We're starting to build a community on Discord and Reddit.

Reddit: https://www.reddit.com/r/pragmalang/
Discord: https://discord.gg/gbhDnfC

I worked on multiple SaaS projects for other clients and myself. I noticed that I'm doing the same thing on the backend every time I add a new feature:

• Create a migration to the database
• Create the necessary routes
• Validate user input
• Check if user is authorized to access this route (sometimes)
• Make some calls to the database
• Make some calls to other external APIs (sometimes)

So I started to think about this problem, but I couldn't solve it by myself. But I created some prototypes for the solution.

Then I met my co-founder Muhammad Tabaza on a local developer Facebook group, and after 2 weeks I met him in-person. When we met in-person, I shared the problem with him and my ideas on what the solution should look like.

We started implementing the solution a week after.