Vision

Elvenware Logo

TOC

Overview

NOTE: This file was last updated in 2015. Much has changed since then. I keep it just because it might form a template for an updated version of this information.

For the IDE, I think we are using WebStorm. If we find it helpful to use Android Studio instead when working with Cordova, then it is worth mentioning that they are both made by JetBrains, and by and large work the same way.

I prefer using mocha, but it seems that yeoman has pulled us back toward jasmine due to the code they generate for our projects. Also, there has been a long marriage between Jasmine and Angular.

Fortunately, the syntax for mocha and jasmine is not just similar, but identical, at least in most cases. I want us to use chai for assertions in our tests because chai is flexible and works with both jasmine and mocha. It provides a single way to write assertions on both testing platforms. Karma works with either Jasmine or Mocha, and complements them rather than representing an alternative technology.

I keep wanting to play with Selenium, and came really close to doing so in Prog272, and then pulled back at the last minute and ended up creating integration tests using the same technique that jasmine uses here:

If you look at the bottom of that page you can see that they whole page is actually unit tested live on their site. Actually not a unit test, but an integration test.

Finally, lets move away from the tests and talk about the architecture for our apps. I want us to create an app that has:

Except for Cordova, that is what we see in both the AngularChartOrganic and AngularChartCensus assignments. If you look at the second screen shot in AngularChartCensus, you will see that it combines all the elements described above except for Cordova. But you can combine Angular and Cordova, as demonstrated in this Yeoman generator.

To sum up, we are working from two different ends:

This is the platform that we want to use in this class. As always, I could have done a much better job of presenting this during my lectures, but I feel the solution ultimately is a good one. Not perfect, but good enough.


Let's step away from the particulars and talk theory for a bit. The Microsoft (and my old Borland) worlds had a big advantage:

They also had disadvantages:

The open source community, had no such problem. They could:

Their problem, however, was that:

Now, hopefully, you can see one of the reasons I have focused on Angular: it provides, to some degree, a single solution by encapsulating the following technologies:

I'm probably leaving some important technologies out, but the point is that Angular helps resolve the problem that you are pointing out. We still need Google charts, but it is not that big a jump.

By adding in Yeoman, we are also given ready made solutions for how to combine technologies such as Angular, Bootstrap and Jasmine. We type yo angular, and we are given a single solution that integrates all these tools. It is a huge help, at least for me.

As John has pointed out, we give up some things by using Angular, but I think the advantages outweigh the drawbacks.

We don't like Angular because:

Those are the drawbacks. But we like Angular, because it combines

Overall, I think Angular is a win for us. But I will continue to switch back and forth between Angular and other solutions quarter by quarter to help us see that Angular is more an arrow pointing toward how to build apps rather than a single solution that we should all adopt.

Yes, there are problems inherent in open source solutions that force us to confront problems like this. It is a rough and tumble world, but oh my gosh, right now it is working very well indeed.