[Plone-conference] Food for thought on AngularJS and Plone

Laurence Rowe lrowe at stanford.edu
Mon Nov 3 12:52:51 UTC 2014


Very interesting post Lennart. I agree with most of what you say (especially about modularity and library-ishness, but my thinking diverges somewhat on the less JS the better part as in my experience hiding complexity never really works out.

The patternslib/mockup approach is probably a good one for gradually introducing more JavaScript into Plone as it is now, but in the long run I expect we will need to take the more radical step of going all in with JavaScript. The problem with progressive enhancement is that it often means building every interaction twice and that has to be unsustainable in the long run. It also makes it hard to reason about how the various progressive enhancements applied to a page will interact, and testing it all can be tricky.

Eventually I'd like to see all of Plone's HTML rendering move to JS so that the existing mess can fall away. Producing JSON from Python is just so much more pleasant.

I'm not entirely convinced that Polymer/web components will end up being the future either, they introduce a huge amount of complexity. I find ReactJS feels quite declarative even though it is pure JS.

The important point though is that web projects are becoming much more reliant on JavaScript and people will need to build custom front-ends in a range of technologies. I think that makes the JSON API the best place to separate JS and Python code in the long run.

Laurence

----- Original Message -----
From: "Lennart Regebro" <regebro at gmail.com>
To: "JC Brand" <jc at opkode.com>
Cc: "Plone-conference at lists.plone.org" <plone-conference at lists.plone.org>
Sent: Monday, 3 November, 2014 11:28:01 AM
Subject: Re: [Plone-conference] Food for thought on AngularJS and Plone

Also: 

http://regebro.wordpress.com/2014/09/08/thoughts-on-javascript-frameworks/ 



On Mon, Nov 3, 2014 at 11:41 AM, JC Brand < jc at opkode.com > wrote: 


Hi everyone 

Firstly, thanks to Netsight and everyone who had a hand in organising and hosting such a wonderful conference. 

There were some earnest discussions around the usage of Javascript frameworks with Plone and about bundling certain frameworks by default. 

For those who still might believe that bundling AngularJS with Plone is a good idea, here is some food for thought. 

The planned 2.0 release of Angular contains big backwards-incompatible changes in syntax and the reaction online has been strident, 
including this Hacker News discussion: 
https://news.ycombinator.com/item?id=8507632 

and this article: 
http://blog.dantup.com/2014/10/have-the-angular-team-lost-their-marbles/ 

Additionally, Angular 1.3 will be supported for just 18-24 months after the 
release of 2.0. 

Cheerio 
JC 

(p.s. I got the links through the "Javascript weekly" newsletter) 



_______________________________________________ 
Plone-conference mailing list 
Plone-conference at lists.plone.org 
http://lists.plone.org/mailman/listinfo/plone-conference 


_______________________________________________
Plone-conference mailing list
Plone-conference at lists.plone.org
http://lists.plone.org/mailman/listinfo/plone-conference



More information about the Plone-conference mailing list