Having built an app with jasonette if you’re looking for other’s feedback on the roadmap these would be mine based on my experience as a developer building an app and based on feedback from users of the app.
- reliability - unit tests, catch regressions early, highlight errors between platforms etc. I know nobody wants to write unit tests, but it’s table stakes for viable frameworks these days.
- change the parser from recursive descent to something modern that can catch errors. Ideally catch, report and recover from errors.
- developer tooling, even something as simple as being able in the json to set a base url and then have all the app links be relative so I can switch between dev, staging and production really easily would help, but you only have to look at React to see good examples of this.
- performance - the same content rendered in safari is buttery 60fps smooth compared with jasonette which stutters and breaks the feeling of being native. Same content, known size images, the issue is that jasonette is not memoizing the calculations made as content pages into view. I can understand the cold start issue because you have to calculate once, but after that these item heights should be memoized. I realize that performance is hard, but there’s still some how language fruit in this area.
- full set of UI components - drag/drop, UICollectionView etc etc
I personally don’t care about programming-free app development because ever since the 4gl languages of the 90’s that’s been a misnomer - anything more than a trivial demo app needs programming, even if it’s declarative as jasonette is. Maybe this is useful to other people but I think simple apps are a waste of time, might as well just be a web app. For anything that clearly shouldn’t be a web app (or can’t be) programming is always required, which is why I have my roadmap wish list which includes lots of developer support. The demo of an Instagram-like app is just a view, there’s no actual interaction which is why it give s the appearance of programming-free but only because it’s missing all the parts that require programming.
The project would also benefit from “liquidity”, more apps built with jasonette would mean more (other) people finding and reporting bugs. The Ethan app doesn’t have enough downloads to even get a rating and it’s been over a year since it was last updated, I don’t know how many apps are actually built and released in jasonette but it would be incredibly useful if the community was larger.
But this is Ethan’s project so he can do whatever he wants with it, I just wanted to at least give my feedback after having spent the last year using it to build an app.
This is a lot for one person thought to take on and unfortunately lots of grunt work that people seldom want to do, frameworks like React of course have huge investments of people and money so it’s obvious why it’s so comprehensive. This is like the browsers of the 90’s - back then anyone could build one, but that quickly changed. The same is becoming true of these frameworks.
I’m curious what other others would like to see on the roadmap. Not that this is a democracy but it might be useful to Ethan to hear of his user’s experiences building actual apps.