Sipp - A Web Framework for Indie Hackers

Hey @jsjoeio, thanks for signing up! 😀

I'm still considering how tightly to integrate Mocha vs Jest. Likely, it will probably be something where you could easily swap test-runners.

Here's why I prefer Mocha at this point:

• Jest is slower at scale (>1000 tests), this is largely because Jest does a lot of mocking and setting up a sandbox for each test suite
• Jest disallows global objects (even when doing --runInBand). The reasons for this is the sandbox stuff plus parallelization/worker strategy. If your tests have some heavy, stateless objects that are nice to setup once and use everywhere ... you can't do this in Jest. I've had heavy memory leaks that are impossible to debug/fix because of this in complex apps.
• Jest does some silly things like if before throws an exception, it still runs all the tests.

Cypress and Mocha/Jest could co-exist where Cypress is used for end-to-end and Mocha or Jest for unit tests. But, since Cypress uses Mocha under the hood, I would probably pick Mocha as the default so it's all the same.

@jsjoeio, I see you in the list but looks like you didn't get the double opt-in... might be in spam?

@polluterofminds, ok, I clarified that the waitlist is just early access/updates/influence over tech choices and not for a paid product.

@polluterofminds, thanks for checking it out!

It's going to be an open source framework, so it won't be paid. The waitlist is just for early access before it's publicly available. (I believe that there are ways to monetize open source, but charging for a for a framework like this isn't something I think I could pull off.) Remix is trying to monetize with licenses—but the authors have a huge audience and built a good reputation in the React community, so they might have a shot.

I'll clarify that on the page.

@yonajune, reading some old stuff from 2016 from the Next team. This GitHub issues thread has this from Guillermo Rauch:

The biggest feature of Next.js is that it's not a backend. It's closer to what @getify calls a middle-end. A universal rendering frontend.

Based on everything I know about Next@10 today, this is still true. Next.js solves the server-rendered SPA pain point really well and is a category leader; it's not framework like Rails or Laravel. You've still got to roll your own data tier.

Thanks so much for reading @yonajune!

Agree, I definitely don't want to directly compete with Next.js (or really, a number of other great projects out there). I'm thinking about sipp as something categorically different—so if it sounds like a head-to-head competition, I'd definitely want to reframe.

It does seem like stuff like Remix and Redwood are more going head-to-head with next. (Blitz wraps Next, I think.) Next's sweet-spot seems to be somewhere between SSG and a completely dynamic, entirely user-generated content.

Based on your experience, would you build a full-stack app on Next.js?