Hacker News Reader

Comment by Mathnerd314 | Hacker News Reader

PostgreSQL database with 25k Java/Spring codebase locked into Stripe APIs.

I guess it will fit some people's use cases, but my impression of Java is that refactoring anything, like changing from Stripe to some other payment provider, will take months.

View on HN

It supports PayPal: https://github.com/alfio-event/alf.io/blob/master/src/main/j...

And Revolut's business API: https://github.com/alfio-event/alf.io/blob/master/src/main/j...

So adding new payment methods can't be that difficult

Paypal took from May 12 to July 28 (https://github.com/alfio-event/alf.io/issues/77), and Revolut took from March 18 to April 8 (https://github.com/alfio-event/alf.io/pull/622). Maybe "a few weeks" is a better estimate for the difficulty, but it's certainly not a weekend project.

Hi, one of the devs here. I would like to clarify that this project has been developed as a side-project, it's not (at least, during that timeframe) our main work.

So don't take the time to develop a feature as a 100% full time work :)

Why do you think this is the case? They have a PaymentProvider interface and multiple implementations for PayPal and Stripe. Should not be that much work to add another provider.

Less about Java and more about the code pattern around it.

Yeah, I've written a lot of Java, and I'd be happy to inherit a well-tested, well-factored Java code base. The language is more of a pain, but automated tools like IntelliJ's refactoring tools (which depend on Java's explicitness) make up for a fair bit of that. But as you say, the patterns encouraged in the "enterprise" space are often terrible, so I'm automatically suspicious of anything like that.

Reply on news.ycombinator.com