refactor(app): redux-simple-router + universal-redux npm package

[bdefore]

Some of you might have seen my universal-redux npm package: https://github.com/bdefore/universal-redux .. there's another PR out for it here: #626. It extracts away a lot of Webpack, Express, and universal rendering from this project, and leaves behind just the parts from the example project itself, plus the API server. But that PR stays with redux-router which the community is starting to move away from in lieu of redux-simple-router.

I've just released a 1.0.0-beta1 version which switches to using redux-simple-router, following along much of the steps of @jlongster 's PR fork: reactjs/react-router-redux#13

What's nice about this is that those who have used this project as a boilerplate who'd like to stay with redux-router can pin to ^0.16 of universal-redux while further development can proceed with using redux-simple-router by requiring ^1.0.0-beta1. It also extracts ApiClient into a more formal redux middleware, which would make exploring redux-fetcher easier to do.

I've deployed it here: http://universal-redux.herokuapp.com

Far as I can see everything seems to work great (Widget, Chat, routing).

[bdefore]

Related code of universal-redux version 1.0.0-beta1 lives in redux-simple-router branch: https://github.com/bdefore/universal-redux/tree/redux-simple-router

[Ghostavio]

wasn't this supposed to also fix the redirect loop on IE9?

[bdefore]

@Ghostavio If you're referring to #416, I believe redux-router 1.0.0-beta5 fixes the issue, but that shouldn't apply with this PR since redux-router has been removed. I noticed that I had a stale push to Heroku and have updated it. Can you try again at https://universal-redux.herokuapp.com ? I tested it and it seemed to function normally on Browserstack running Windows 7 IE9.

[Ghostavio]

Yes, sorry, it's working just fine, thanks :)

[quicksnap]

Just a note for posterity: this is a huge change, so it's under heavy scrutiny. I have a project to bootstrap and will give it a spin on there. I'd love to hear any feedback from other people on their experience with the npm module.

I think it's a great idea, but it can open up a new class of problems, sot it'd be great to get as much early feedback as possible. If it works out well, we could maybe cut a different beta version that depends on the module, or some other path.

Also, I wouldn't feel comfortable merging this without discussing with @erikras. I believe he's involved in other matters at the moment.

[bdefore]

@quicksnap all very sensible. also, i would see this issue as a blocker for now: bdefore/universal-redux#7 i'm currently investigating.

[bdefore]

@quicksnap all of my 0.x changes are now merged into the 1.x code (on master) ... you should be good to go aside from the issue mentioned above.

[korczis]

Are here any plans to merge this PR? @erikras ?

[catamphetamine]

Since I didn't get which PR shold be commented upon, I'll just copy & paste my comment here too:

Is redux-router really done?
I saw some of the old issues being closed recently.

As for the movement towards an npm package - that's a must.
This project started loosing popularity because it became too complex.
Not a single newcomer will be able to wrap one's mind over all these piles of code.

Where the demarkation line should be drawn is another question.
Obviously noone wan't to know about Webpack weirdness and all the moving parts, that's for sure.
Ideally an average user would like to have a sinlge "DO IT" button to push to start his own project based on this boilerplate.
In this sense this universal-redux thing may be the way to go for the project owners.
It needs to be Rails-alike kind of thing.

[andresgutgon]

@bdefore how are those changes related with this PR:
#833

Is compatible this code with using async-props?

Thanks for all the hard work. This is great :)

[bdefore]

@andresgutgon universal-redux takes a very similar approach to that PR. it also supports async-props as of version 3.0.0-rc5.

[andresgutgon]

Great, sorry I saw it after comment here. It's great your work on universal-redux

[sars]

@andresgutgon It's just very fat PR, my is much smaller and i believe it's in the same direction.

[bdefore]

I'm going to close this PR as it's superseded by #759 Please direct any further comments over there.