blog.worldmaker.net VS hyperapp

The biggest flaw that I think most other flaws derive from is making the huge dependency on RxJS, but then not trusting it and building in escape hatches that end up making "Angular Best Practices" wind up becoming "RxJS Worst Practices" and performance suffers accordingly. I noticed that even the Roadmap now mentions that they are finally re-evaluating the play between RxJS, Angular Change Detection and Zone.js. Rx.JS is "push" oriented so why do you need two complex "pull" oriented systems around it?

(In my opinion: Zone.js and Angular Change Detection are these extremely baroque, over-complicated systems that aren't necessary at all if Angular actually trusted RxJS to "push" changes. Or they could skip the huge RxJS dependency and go all in on "pull" like Vue/Svelte. What they have today is a worst of all worlds compromise that benefits no one.)

Blog post is here: https://github.com/WorldMaker/blog.worldmaker.net/blob/gh-pa...

Or you can read it on my actual blog site but it scrolls horizontally when the browser is wider than tall and that got a lot of hate the last time my blog was on HN that distracted from the contents: http://blog.worldmaker.net/2021/06/26/angular/

While we are at it, here's the Component Framework I built out of what felt like necessity to try to make following RxJS best practices easier in building Angular Components (and make it easier to switch components to ChangeDetectionStrategy.OnPush and apps towards being able to set Zone.js to "noop"), which I built out of some struggles doing performance optimization work in Angular and some patterns I was already following in Component building: https://worldmaker.net/angular-pharkas/

Please check out https://github.com/jorgebucaran/hyperapp

The likely reason it never caught on, is that it has similar pitfalls as Backbone:

- manually attaching DOM elements to view controllers

- manually attaching child views

- models which have to be wired individually via .listenTo

- possibility of infinite loops if the events accidentally recurse

A better tiny alternative would be hyperapp[1] or even Preact, that has a similar bundle size.

[1] https://github.com/jorgebucaran/hyperapp

If they're focused on performance and bundle size, it's your chance to try some minimalistic exotic stuff like hyperapp (https://github.com/jorgebucaran/hyperapp) or mithril (https://mithril.js.org/) Just for fun

With a bit of a deadline (due to a mixture of procrastination and confidence that Vue would work) I needed something quick. I have also used Hyperapp in the past but that looks like a dead project right now (although arguably it has all the functionality you need so why keep developing it?).

I’m glad it left that impression! My thoughts have clarified a bit since I read that post, and I think what I describe is more declarative, like React. But the best places to read about it (for web devs) are in Elm!

There is also this new thing I found that seems to really lean into the core of what being functional means here: https://github.com/jorgebucaran/hyperapp

After a while, you see that basically all systems can be modeled as event-driven, functional systems. It’s a flexible model, and fits beautiful into web dev where the semantics are very clear: the system is the web app and events are clicks, keyboard events, asynchronous calls...

Hyperapp is 1kb.

- Hyperapp (https://github.com/jorgebucaran/hyperapp) - Preact - Svelte - React / Vue

I'll tell you what I've learnt from struggling with a bug that made me lose a couple of weeks. The application framework used in this post is Hyperapp, but I guess the same problem can be found in frameworks based on transforming the state of "Single Source of Truth" with pure functions (such as Elm, Redux, so on) if we use them in a wrong way.

Personally, I try my best to avoid bringing in dependencies as much as possible, and try to limit my exposure to only dependencies with low/shallow transitive dependency counts. Unfortunately, this is pretty hard, especially in corporate settings. What we need more of are the opposite of what we've been collectively praising: we need more monolithic packages. Case in point: lodash.template is currently vulnerable with no mitigation, even though lodash itself is not. That's just sloppy publishing practices. Esbuild is a great start over the webpack/babel maze of dependencies. There's a stdlib effort along those lines that hopefully would also help. There's a bunch of micro-frameworks that are used in production just fine and have little to no dependencies.