Whitelisting URLs in PhoneGap

I've built a couple of HTML5 (jQuery Mobile) applications that have also targeted native app stores using PhoneGap. When building these applications, it becomes clear that to get desired navigation in iOS, one must carefully consider the PhoneGap settings for url handling "OpenAllWhitelistURLsInWebView" and "ExternalHosts". In many configurations this results in setting the "OpenAllWhitelistURLsInWebView" to "YES" and Subsequently setting a single "ExternalHost" as "*" in order to keep your application functioning within the WebView exclusively. This is great for everything except where you wish to open an external link that would break your HTML5 applications navigation (due to a lack of back button). What follows are a few tricks I have picked up to help with whitelisting. In many cases you know when you're building your application where you'd like to have the external links open, in which case you can set these external links within your Objective-C source code letting the applicaiton know which links to open. This would look something like the following Gist (which looks to the prefix of the URL in order to open a link outside of the WebView): In other cases, your application may allow for some dynamic client created content to come from the server. This can mean either a maintenance nightmare for setting up this prefixed whitelist, where you get to update the whitelist and resubmit the app to Apple everytime the list changes (icky).... Or we can come up with a solution that works better. One way in which this can be accomplished is by looking to the suffix of the URL to find a querystring parameter that would indicate that we need to open external. A perhaps more robust solution, allow for the conversion of any link with target="_blank" to open in Safari. This solution is inspired by this where you tell your iOS app to find a specified document fragment, and also tell your web application to ensure that target="_blank" links always have such a fragment. The Obj-C code looks like this: and telling the webapp to set this fragment (using jQuery) looks as follows:

I read a book: JavaScript Enlightenment

I dunno if this is a full fledged review or not, but I recently picked up and read the new book JavaScript Enlightenment by Cody Lindley, and I found it ranks in my list of JavaScript books I would direct others to read in their quest to JavaScript wizardry. Lindley states he writes his book first and foremost for himself, as a personal reference, which I find handy because that is exactly how I like to read JavaScript books. His approach to teaching the language might strike some as contrary to what is found in the wild - he spends plenty of time not using object literals when creating objects in order to hammer home the underlying principles of JavaScript objects - but he does also hammer home that using literal notation is preferred for its terseness and readability.

What else did I like about this book? Well I enjoyed the use of code examples, as its always good to see JavaScript in action. He also makes many great succinct notes that help make good distinctions such as "Remember: object instances created from constructor functions are just plain old objects". Another concept that those in search of JavaScript enlightenment should find well laid out in "JavaScript Enlightenment" is how the prototype chain is traversed whilst accessing a property on an object. Along with traversing the chain, Lindley describes through an example of what not to do, how prototypes can be altered even on the Object.prototype, promptly reminding the reader that this is a "forbidden" example. Since the prototype is key in JS, the book dedicates a considerable portion to describing and outlining how it all works.

The book describes in great detail and examples many concepts that any front-end engineer should be familiar with. Things such as the Function objects instance properties and methods - you know: arguments, constructor, length, apply(), call(), toString() - including going into specific examples on how to use arguments.callee when calling a function recursively. The discussion of JavaScript scope and closures is straightforward and to the point without over complicating concepts that are sometimes hard for some developers to grasp.

Other parts of this book I would describe as desk reference material. These are items that are common and I think are generally understood by many, but in the case someone runs into a gotcha during development, they may find it useful to hit the chapters on the various objects: Array(), String(), Number(), Boolean(), as well as then utilizing String, Number, and Boolean primitives. Other desk references: Math(), null, and undefined.

I'd totally recommend this book on the same JavaScript binge handed to new developers when the get underway (things like The Good Parts, Eloquent JavaScript, JavaScript Web Applications, or any of the many many useful links to the plethora of great resources available to the modern developer). I guess this did turn into a full fledged review.

Openness

On 11 November 2012, my first patch to jQuery-UI landed in the project. It was not the world's greatest pull request, nor was it the best code ever written (by a long shot).  But it was important. Not because I get my name on a change log somewhere, or that I will now conquer to world with my open source prowess. Its important because it is another tiny step in the massive open source snowball that wins. It wins big. The open web as we know it has spiraled out of openness and open source projects. The documentation that I turn to regularly and have contributed to is open and invites new contributors each day. There are countless open initiatives in the world these days - and by no means are these all software related - these encourage the scientist in us all to submit pull requests, change direction and assert for the greater good. If you have a passion, pursue it. That is all.

Homework

Recently I was tasked by my son's pre-k teacher to present a paragraph that describes my occupation... in a way that 4 year olds will understand. This really put things into perspective for me, more than I thought it would. At first I think of what I might describe to my parents: "Software Engineer: I write code for web applications and databases"... No. So what is it that I do?

I was then reminded of an interview question I once had where I was asked: "How would you describe a database to your grandmother?"  These sorts of questions may seem silly to some developers, but I find them to be a great exercise. These questions help us to target and focus on what it is we really do, and how our end users or clients might perceive what someone in the software profession might be thinking. So I thought of the answer provided in that interview where I made the analogy of a jewelry box that helps to organize and sort jewelry for efficient placement and retrieval.

How does a jewelry box translate to pre kindergarten children? Well I tried to avoid an abstract analogy and simplified it by stating that I teach computers to do neat things and give us information when we ask them for it by typing or clicking the mouse.... Our children don't care about CSS, HTML5, JavaScript, objective-C, C++, Java, Ruby, Python, C#, or if we built our app for Android, iOS, Windows Phone 8 or whatever,
they care if they can learn from and enjoy whatever it is we built.

So a fresh perspective of "how can this actually benefit someone" has been bouncing in my head lately. I'm not saying one shouldn't try new cool tech... Still create experiments and things that push development forward, but perhaps a deeper look at what is truly needed should be a priority.

TypeScript

This week Microsoft released its shiny new JavaScript pre-processor - TypeScript. It represents a superset of JavaScript that is designed for large-scale application development. It utilizes concepts that most developers are comfortable with in their non JavaScript languages. Things like Classes, Modules, and Type checking.

These are all things that many people complain that JavaScript does not have built in.  With TypeScript, JavaScript does not magically gain these things. TypeScript makes these things accessible to a developer through its pre processing. Developers and JavaScript do not need TypeScript, and it is not a requirement to build large scale applications. It won't magically make crappy programming better, but it may help catch a gotcha for someone every now and again, just as JSHint would. I for one think that it will take off rather well, especially in the ASP.Net community. I also will probably try it out in a project or two or three. I like that it looks and feels like JavaScript, which is a syntax that I am comfortable with.  That is one reason why I have not adopted CoffeeScript, it just looks so foreign to me. I suppose if I were a full time Ruby on Rails developer, CoffeeScript would feel right, but I'm not, so it doesn't.

I'm going to end this by just showing the difference between TypeScript and JavaScript. It is subtle and that is appealing to me. See the comparisons below:
typescript javascript