Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/csev/wc-experiment

Chuck's Web Components Experiment
https://github.com/csev/wc-experiment

components web

Last synced: 6 days ago
JSON representation

Chuck's Web Components Experiment

Awesome Lists containing this project

README 

        
Dr. Chuck's Web Components Experiment

=====================================



This is a place where I will develop code to figure out how I want to use

web components in my own development.

I see the future of web applications as increasingly web components based.

I see web components and JSON APIs as a great coding pattern regardless of your

front-end approach and back-end approach.



If web components are truly ready

for the "big time" they should work in static HTML and you should be able

to view source and understand what is going on.



You can see my playground at https://www.dr-chuck.com/wc-experiment/



It is simple, static, and crude.  I will add more sample code over time as I learn more

about the way *I* want to use web components.

I don't want to adopt a framework - I want to use browser functionality as it comes from the browser.

In a way all I see all of the front end and back end frameworks as doing research in how web components

need to function and informed the design and development of web component

implentation in modern browsers.  So that is a good thing.



But I do not want to develop code to any framework.   I want to develop a pattern that can work with

PHP, JQuery, and Bootstrap 3 or Java and Velocity.   I support far too much code that cannot afford

to rewrite itelf completely every time a framework development team sneezes.



How I want to go about this

---------------------------



* I want it all in the markup



* It is OK to download a static asset from a CDN - like a polyfill for older browsers - this needs

to be pretty static - if it did not change for a year things won't break



* I am OK with extending *one* thing like LitElement or HTMLElement - as long as that code is

so old and boring that it is in maintenance mode and will only issue new releases if something

in browsers breaks it.  (kind of like JQuery).



* I could tolerate a simple templating library that is old and rusty and not

changing much anymore like Handlbars.   Cool, hip, recent and getting better rapidly are

all negative aspects of any framework



JQuery is mature enough for me - but I want to move to vanilla JavaScript because JQuery did its job

and shamed vanilla JavaScript into fixing itself.  #ThanksJQuery



No Thanks

---------



Many interested in web components are advocates of one or more "crutch" technologies that to me

are fine ways to "prototype" web applications but not a good way to build a web application that

needs to last 5-10 years without major rewrite.  If you are a VC-funded startup - you are trashing and

rewriting your code constantly - I can't afford to waste that kind of money on my software.  So I avoid

quick "feel-good" solutions.



* Please don't suggest that I use React or any other client side framework. Backbone and Angular 1

are probably mature enough for me - but that only makes my point.



* Please don't suggest that I use Polymer, Node, or any other slick way to "manufacture" web components.

If we are still at the point where it is *impossible* to do web components without these

"helper" technolgoes - I will wait.



* No npm, bower, or other component manager - eventually I might be willing to so some kind of

server side packaging / path patching to make a large number of web components easier to maintain

and more efficient to download.  But for my experiments, I want to see simple solutions that

you see and fully understand when you "view source" in a browser.



* Please don't suggest I use your webcomponent library like PatternFly. Patternfly 3 was

markup or React based based - PatternFly 4 is React based - because they got lazy.  I don't

want to depend on a framework builder remaining aligned with my philosophy over time.

I don't want a framework I adopt to force me to do a complete rewrite on their schedule.



* Please don't suggest I use a CSS library like Material Design - the approach should

work with any markup themeing.   It won't surprise you that I like Bootstrap 3 and think

Bootstrap 4 is a complete waste of time.



* Please don't suggest I rewrite everything in Ruby on Rails and use HAML.  I am

similarly disinclined to use Scala or Rust.  The server side should not matter.

Using an "all-the-rage-at-the-palo-alto-coffee-shops" technology is a negative

in my book.



* Don't mention GraphQL or Apollo - the markup should be data store agnostic.  These appeal to me

but no one can tell me how they make subscriptions / notifications scale with the back-end is a

relational database.  Just saying "magic" is not a sufficient technical description for me.



* Even though I like PHP - don't tell me to use Laravel and Blade - they are impressive on paper but they

need write access to the htdocs folder to work because they compile their templates for performance.

Giving the Apache process write access to the htdocs folder - what could go wrong?



* I am not interested in the *amazing* web components you already have. I am sure you have the worlds

most impressive button.

I already have buttons.  I use the button tag and Bootstrap 3 to style them.

I don't need to run Doom on my buttons.

I want to make my higher-order own extendible web components with a focus on those adopting my approach

to be able to extend my web components to add their own flavor.  I am guessing that your

component library is a bunch of brittle spaghetti code

that you update furiously and in non-upwards compatible ways every time a new browser is released.



If it is not possible to build web components without resorting to these approaches - I will just

keep waiting.  But I am actually quite confident that web components are mature enough for

me to adopt and use in long-lived applications.