Closed HenriPodolski closed 5 years ago
I work with the Polymer team now and there are many patterns still being worked out when it comes to large scale application development. I see some of the patterns Aura helped developers with in the past being reimagined as Polymer elements that make sense in a Web Component world.
For example, Polymer doesn't provide you with a global mediator but it does have an event system. What would a good Polymer mediator look like? Is the current Polymer event system robust enough? I tried putting this together to kick around a few ideas: https://github.com/addyosmani/polymer-eventemitter for this. We could have an <aura-events>
element or something similar to help.
On the encapsulation front, we now have Shadow DOM but Polymer doesn't give you a completely isolated view http://aerotwist.com/blog/the-web-needs-containment/. What if we had an <aura-view>
that developers used that did offer containment?
Polymer doesn't have the concept of a widget/element endpoint source, so you can't just say.. "import" all of the elements in a directory or URL just by using a HTML import. Aura could have a <aura-element-source>
for this. Polymer also doesn't yet have a solid routing component.
Concepts like the facade pattern are now covered by the attribute API model and I don't think it's worth implementing a 'permission' model at least not yet, but there are many things we could help with.
Since I'm here...some more thoughts :)
I would love to see us come up with an initial set of core Polymer elements with perhaps an owner for prototyping and maintaining each one. If the element is small enough, it could offer value whilst also not closing the doors to unit tests and documentation being written up for them.
If anyone is interested in getting involved, feel free to hit me up.
Thanks for sharing your thoughts. It is good to know that there is a vision. I was guessing that you or another developers from the aura developer team have some ideas or rough plans on that.
I was wondering if mission to refactor aura into a system of web components was progressing anywhere or if anyone was discussing it or even had a reason why aura doesnt belong in the web component world. Web components were not ready for production last yeah so why has aura stalled when it could still be useful in projects that are pre webcomponents
I wonder what is the current landscape looking like.
I see that there is not much activity around aura, it looks really nice, but I can't figure the status of it.
Is it still relevant? Should we use something new? How does it play with polymer now?
My primary case is building large scale web app.
attn: @addyosmani
THANKS!
I'm very curious on how aurajs could work together with polymerjs in a future version. Polymer loads components based on tags and bind functionalities to them, which is imho one of the important core functionalities of aurajs too.
I red in #243 that aura differs to polymer, when it comes to structure and architecture. What makes sense, when I think about how messaging and decoupling is handled via the sandbox.
But what are you guessing about aura's other roles in this stack and how could this be implemented? Do you have any ideas on that?