- Joined
- Jul 12, 2011
- Messages
- 1,229
- Reaction score
- 475
Why not just work on an actual Stormcaster front-end? I don't know what features you already have working with your back end but isn't this a duplication of work? Either way, cool to see Laravel being used to update a still-popular project.
Great question! The tldr; of the answer is, I've learned a lot in the last couple years, and Laravel has gotten significantly better alongside me
For a bit more context - I think stormcaster had a lot of good ideas, and was generally well executed for where I was at as a developer four years ago. It's stateless, utilized JWT, implicit model binding and a bunch of other Laravel Things to make the code succinct and generally easy to understand. My main issues with stormcaster today includes:
- JWT is not great or ideal for session authentication for a variety of reasons. With L7.x, we now have the
You must be registered to see links
which makes token based authentication a first class citizen within Laravel. Upgrading stormcaster to 7.x just to ditch JWT is not trivial or worth it.- stormcaster is not well tested. There's a lot of holes in the tests, and I wasn't comfortable picking up an old codebase without complete test coverage. I'm attempting to maintain 100% (or near 100%) coverage for controllers and model logic to fix this issue, should I drop this project again for a bit
- lack of documentation. with v2, swagger specs are self-documenting and makes it easier for me (or others) to understand available endpoints and develop frontends.
In general though, Laravel is Laravel, and at the view layer there hasn't been a ton of breaking changes in the last few years. This is great because I can just copy and paste (as I have in some spots) the methods and bits of code I still like today, so there isn't a ton of duplication in the "business" logic - I'm mainly just rewriting the core of the app.
Lastly, my goals for v2 are a bit different than stormcaster. The main distinction is that I want to render a frontend within the same Laravel app (instead of just being an API), making it just as easy to setup and get running as it is with MapleBit v1. If I had to guess, people use MapleBit purely because it's incredibly easy to download and run anywhere, and ideally v2 should be the same. hope this answers your question!
PS: I'm just kind of developing this stuff for fun, I don't really care if it's never completed. Having a well paying job as a developer means that I just write code in my free time when I'm bored and want to explore new ideas - this is my outlet for some of that energy (I don't really care to make any money out of MS now or in the future). I've also considered just writing this in my lang of choice rather than PHP, but to be honest, Laravel is such a delightful experience to write web apps with!
-----------------------
added some react views today:
You must be registered to see links
Last edited: