Mobile web development is tough especially when you're trying to offer native-like experiences to users. Several years ago, a small company called Nitobi took on the effort of simplifying building native mobile apps using traditional web development skills. Ambitious and sometimes controversial, the effort known as PhoneGap grew out of this need and one converts left and right.
One of the main masterminds behind the framework is Brian Leroux who apart from being well-respected for his development skills and incredibly likeable personality is also one of the savviest mobile developers around. Considering the number of mobile devices PhoneGap targets, you have to be pretty well-versed in a variety of devices and OSs.
Nitobi has since been acquired by Adobe and the PhoneGap codebase donated to the Apache Software Foundation to continue its development as the Apache Cordova project. Brian moved over to Adobe and continues to steward the codebase. In this interview, we'll chat with Brian about how PhoneGap came about and what the future of mobile web holds.
Hello, I'm Brian. I work on Apache Cordova, PhoneGap, and a new css library called Topcoat at Adobe. In my spare time I created a code joke site called http://wtfjs.com which kind of follows me around.
I've definitely been one of the stewards of PhoneGap but it is very important for me to say that MANY people of contributed to the creation and growth of it. No one person really decided to do anything it was a lot of forces coming together at once. PhoneGap was an outcome of the primordial soup that was the new Github model for open source, nascent mobile web browsers, and new generation smartphones. We started hacking, and did the whole thing in the open, and eventually more people subscribed to the project philosophy and utility. It grew from there.
Adobe acquired Nitobi in 2011! It is a little hard to believe that was close to two years ago already. Since our acquisition we donated the source of PhoneGap to Apache which is now known as Cordova. We are constantly improving project, adding features, polish, performance improvements, new tooling, and recently we shipped a vastly improved plugin architecture. PhoneGap has become as much about tooling and extension as it is a fancy embedded web browser for building apps.
We're also working closely with a new team at Adobe on a CSS library called Topcoat that is designed for building fast and clean apps. Of course, everything in Adobe Edge is growing mobile consciousness as a part of our focus on web technologies. Brackets is great for authoring web centric code. Reflow and Inspect are great new tools helping tame responsive design. We'll see more and deeper integrations between these tools and PhoneGap in the future.
Adobe PhoneGap is a downstream distribution of Apache Cordova. It is the same as the relationship of Safari to WebKit. When Adobe acquired Nitobi the original source of PhoneGap was donated to Apache to continue its open development, and encourage contribution from the wider developer community. It has been really great, and the community has grown exponentially since joining Apache. It was a great move for the project and has really matured the development.
Author Note: Brian goes into more detail about this in this blog post.
I think our popularity is owed, in part, to very clearly defined principles and goals. We want the web to be a first class platform and we often state a purpose of the project is to cease to exist. It's a powerful acknowledgement of our intention to get back to web development. This resonates with the web community.
PhoneGap is also just a really good name that clearly communicates the project succinctly. We got lucky there. I'm not sure if it was Brock Whitten or Andre Charland whom coined it. Rob Ellis was there but I doubt he'd remember either. I hated it at first but after five years of working on the thing I'm sort of used to it!
The adoption of PhoneGap was probably a little bit of dumb luck too. I'd like to think we made some of that luck with a regular release cadence and a strong testing philosophy. We rarely have regressions, and we ship quality releases continuously. That healthy activity has helped to build the confidence of our developer community, and the businesses and organizations that are using PhoneGap today.
Well, on one hand it is super easy to get started building a web app. On the other hand, web apps can grow complex quickly, and the devices we're talking about don't have a whole lot of horsepower to begin with. Software development is a balancing act. We're balancing all sorts of forces. Skill and code reuse. Adding more features or working on performance.
Most mobile operating system vendors are contributing directly to Cordova!
We have friends from Google bringing Chrome Packaged Apps to the fray. Mozilla is ramping up Firefox OS with us. Canonical has hackers working on Ubuntu Phone. Blackberry has a bunch of devs bringing us the Blackberry Webworks perspective. Intel and Samsung representing Tizen.
Early in the PhoneGap project, before all this glamorous Apache business <g>, we were temporarily blocked from the App Store by Apple. It was the best thing ever. It brought a tonne of attention the project. After much kerfuffle, Apple reviewed our code to our mutual satisfaction that we were not in violation of any of the App Store policies and PhoneGap apps have been shipping there ever since.
Well if you have an existing investment in web content or web developers then PhoneGap is worth looking at. If you are looking for portability then web technologies are obviously useful, and this can even mean on a single platform, but able to automatically target handset and tablet form factors with a single codebase.
I used to say that web technology isn't really the best choice for games. But this depends on the type of game. Mobile games in particular tend to be more puzzles, card, or two dimensional sorting things that do not require immersive graphics. Web technology is surprisingly good for these types of games. Until we get better support for WebGL I think going native is still compelling. The W3C and browser vendors are very aware of this shortcoming and it is only a matter of time before the Game Controller, Orientation Lock, Fullscreen API, and the Audio API are fully realized. The console will move into the browser and monetization will move towards a service model as a result.
Well, if you have the time to invest in a particular platform (sometimes proprietary too) then going native is a fine, if costly, route to invest in.
It is easy to write a crummy native app as it is with web tech but it is even easier to debug these things using the native platform tooling. That tooling integration makes most development environments very comfortable, with great documentation, and distribution is kind of built in. (You get these advantages with PhoneGap too as that we do not hide these details.) I encourage developers to always be learning as much you can and native mobile development is super fun stuff to learn.
That said, I'm not as convinced about the business benefits of going native. You inherit a reliance on (often) proprietary tooling and distribution channels which is inherently risky. When the vendor makes a change so do you. If they chose to shut down, deprecate, or otherwise abandon infrastructure you rely on for revenue you will have no say or recourse. I personally would not build a business in that way, but I can also respect that some do, and either way you can use PhoneGap to mitigate that risk.
Offline is still messy, we have App Cache but it is really complex and creates a janky user experience. When a new version is available you have to prompt the user to reload. But I have high hopes for the Navigation Controller effort to fix it.
Push notifications are another thing the web needs to get right. Notifications are crucial for user engagement. Those standards and support are starting emerge in desktop web browsers but we need those capabilities to rise up into the mobile web browsers.
The security models for packaged apps is in need of refinement. But that is happening. As a result we will win more and better device APIs. Firefox OS and Chrome OS are going to point the way. We're going to do everything we can to help by providing a quick prototyping surface for browsers.
Developer tooling experience needs love. It is getting pretty good and there is a real healthy competition between Firefox, Chrome, Opera and to a lesser extent IE and Safari. Performance instrumentation for monitoring and especially post deployment crash reporting would be particularly nice.
I want to thank Brian for taking the time to provide us with the history of PhoneGap and his insights into mobile web. If you're interested in building mobile applications using your web development skills, be sure to check out Apache Cordova and Adobe PhoneGap.