Flutterby™! : Booted 2014-05-19 13:21:18.83526+00

Booted

2014-05-19 13:21:18.83526+00 by meuon 7 comments

Am updating and re-writing a project/product, which for various reasons beyond my control, requires using the latest version of: Bootstrap. As I look at the 3k lines of actual program, and compare, I realize the CSS and jQuery required is much larger than the actual functional code, which is not some trivial fluffy stuff.

The web has taken an evil turn somewhere..

Since the new as-yet-not-completely-adopted standards for javascript, HTML, and CSS provide a lot of the features currently provided by various libraries and shims, the Polyfill route seems the right direction to me. That way your code will be compatible with the new standards when they are supported by browsers and you can code to those standards NOW instead of waiting.

The trick is making a clean break from the current crutch-du-jour and coding without its help; instead coding to the 'bare metal' of the standards. I think a lot of people would have a hard time breaking their Underscore or JQuery habits. Much less Less.

#Comment Re: made: 2014-05-19 16:26:31.266303+00 by: meuon [edit history]

While I like the concept of Bower, which is used to install a lot of things, including Polymer, and Bower is Git Based, but it requires both Node and NPM.. I find my self looking at a world of crutches, built on crutches, built on crutches, all the way down. The issue is the stack of complex dependencies and crap that's got to be installed and maintained keep increasing.

My goal is/was to make a simple 5 page/view web application to present data and take payments, look like the clients other web assets that use Bootstrap 3.1

The ecosystems being built on top of ecosystems on top of.... to do simple things is just driving me nuts.

And honestly, Polymer doesn't seem to fit a need I have - right now. But then it was difficult to understand what it brought to the table unless I need to do things with polygons.

My point is, using Polyfills means using a crutch that will eventually go away, once browsers catch up to the standards. So if you base your code ENTIRELY on those standards and eschew all the complicated frameworks and libraries, you eventually can drop the crutch and (theoretically) your code doesn't change.

Polymer Project does add an opinionated layer on top of the Polyfills, providing a clean way to implement data-binding and the like. However you don't need that layer to take advantage of the Polyfills. For now, though, you will have to manage the Polifill libraries; thus tools like Bower or simply maintaining your own by downloading the source and building it with Grunt or something.

That said, if I can figure out how to use my own text replacement library instead of the built-in Mustache one, I think I like the Polymer Project way of doing things. However, they baked their Mustache code in very tightly and did not provide any simple way to do text replacement functionality outside of Mustache- style. If I could just find a single entry-point for that processing other than the Mustache-aware Binding code . . .

#Comment Re: made: 2014-05-19 18:28:14.527117+00 by: meuon [edit history]

And somehow, during all this digging, I trip across the "Shadow Dom" and Web Components. Which is where Polymer comes in and: "which automagically turns templates into shadow DOM and registers custom elements for us" What a Monday.

Also X-Tag, which is the Mozilla Project's answer to Polymer. It starts with the basic Polymer Polyfills, adds a few of its own, and then does it's own thing on top. Seems to be a bit less opinionated, but I'll know more when done researching it.

I'm gonna have to dig in and read through this stuff, I know for a while jQuery was pitched as the DOM access method that browser builders optimized for, so I'm torn between "oh, good, one less abstraction layer", and "Again? That's what we were told last time..."