I believe Polymer is actually out ahead in terms of it's technology and philosophy in many ways. With native Shadow DOM and Custom Elements support coming to a majority of browsers this year, I think the skepticism will subside, more people will realize the benefits and Polymer will get more and more press.
React solves the same kind of encapsulation into modular component problems but a lot more besides.
Why, there seems to be more and more uptake of web components .
> React solves the same kind of encapsulation
What exactly react adds? (Never used react so I have no idea)
I compared TODOMVC examples and polymer looks really awesome.
Besides we can reuse webcomponents with other frameworks out there, so that seems like a good idea to me.
Maybe I'm wrong but React doesn't provide that I think?
Doesn't the question that started the thread imply the exact opposite? Where's this "uptake" -- outside of Google, who can use whatever they want, their teams use stuff that few outside use too, like Dart.
The shadow-dom (and it's speed, predictability benefits), being able to run on the server, uni-directional data flow, being able to run on native platforms, all are things that webcomponents don't provide.
See here for some more discussion: http://programmers.stackexchange.com/questions/225400/pros-a...
>* Besides we can reuse webcomponents with other frameworks out there, so that seems like a good idea to me. Maybe I'm wrong but React doesn't provide that I think?*
Well, the core React lib (e.g. without flux helpers, router, etc) can be used as the view/template layer in many frameworks that allow you to replace their own view layer.
I'm asking about demographics here avoiding the discussions. People are leveraging webcomponents all the time, recently i saw job opening at slack for Polymer dev.
> Where's this "uptake" -- outside of Google, who can use whatever they want
I saw questions about webcomponents/polymer pop up all the time for aurelia and other frameworks (how to integrate nicely), github uses components too, salesforce. It seems that here is it treated like something that doesn't exist.
> The shadow-dom (and it's speed, predictability benefits)
React doesn't provide shadow DOM I think? Polymer does, i think it is the opposite?
> Being able to run on the server
This is cool and a nice selling point, I've been interested to do that multiple times but it appears that this is only possible if you build your app in node.js?
> React native
This is also awesome selling point, I still have reservations about PATENTS clause in all FB-related open source projects.
I don't think it's only here (on HN) though. Check Reddit reactjs, polymer and webcomponents subreddits for example. The first has constant traffic everyday, the other have 1-2 recent posts, and get to month old posts on the first page.
Same for tutorials and blogs. Polymer/WebComponents related ones are few and far between, React ones are everywhere.
GitHub: polymer 14000 stars, react 37000 stars (and 4 times as many forks).
Regarding the "shadown-dom", yeah, you're right, meant to write "virtual-dom" (a much lightweight dom copy, with faster access and the ability to diff to avoid doing costly work to the real DOM).
>This is cool and a nice selling point, I've been interested to do that multiple times but it appears that this is only possible if you build your app in node.js?
You can run it at any app server that can execute JS. So not just Node but also Meteor, and Java servers to (Java as of late comes with JS interpreter built in).
And you can also fork out from any other framework and do the rendering part in a JS interpreter, e.g. like this:
https://github.com/tildedave/universal-react-without-node
I'd say the most practical versions would be Node and Java though.