A little rant here.  I know Adobe AIR is new sexy and everyone’s trying to build apps that run naively on the desktop, but please, if you’re not making use of local filesystem, native windows, drag-and-drop, SQLLite, native menus, or any other of the useful AIR APIs, you should really consider not releasing it as an AIR application.  I’d prefer to use your application quickly via clicking a URL in my web browser.  I say this after having spent a few minutes uninstalling most of the AIR applications that have been cluttering my desktop that turned out to be remarkably useless, yet still required network connectivity.  There’s no need to clutter the desktop with applications that run perfectly well in the browser!  Let’s not forget about the great little thing called the World Wide Web, web applications, Web 2.0, AJAX, and regular-old Adobe Flash/Flex!

  This burgeoning Adobe AIR application market space reminds me quite a bit of the Visual Basic 3.0 days when all of a sudden tons of crappy shareware was developed in it.  Unlike those days, however, we already have a better way of running and trying out these applications, the web browser!  The user doesn’t need to bother installing/uninstalling your application!

  Here’s a technique that I’d love for all AIR/Flex developers to adhere to when creating their applications.  Make it run in both environments: the browser (via Flash), the desktop (via AIR).  Make the web-browser version limited in functionality — if you desire — showing the user the essence of your application.  Make the AIR full-blown sexy that a user might want to upgrade to or purchase (maybe the user wants to use their own personal files and data).  Prompt the user using your web-based version to upgrade to the full version at appropriate moments, like when they press the “Save” button.  This allows me to navigate to a URL, use your Flash/Flex application, and if it’s worthy, download the full AIR version.  It even allows for a way to monetize your application.

  Now, I fully understand the desire to take your existing Flex apps and port them to the AIR Framework so you can test the installer and see how it behaves on your desktop.  After all, it’s a great new tool, and you get that warm-and-fuzzy feeling thinking about all the new desktop applications that you can build.  But if there’s something to learn from yesteryear, it’s that there’s going to be a lot of CRAPPY Air Applications released — and I don’t want to waste the time installing and uninstalling what are essentially demo applications.  Nor should we want to pollute the world with a ton of weak AIR applications so that when they are prompted in the future to install the Adobe AIR Runtime or and .AIR file, they’re not going to want to install it because of all the negative association in their head.  (Any of you who’ve ever installed shareware made with tools like VB 4.0, etc. know what I mean with this.)  This is one reason why Web 2.0 applications are so great:  I navigate to a website, use the app, hate it, never go back.  Nothing further is required from me!  No installing/removing of software.

  Flash/Flex/AIR provides a very unique opportunity that allows people to try your software before purchasing via a web-browser, with nearly zero hassle.  I can’t think of any other tools out there that encourage this. 

  I promise you, all my applications, you’ll be trying out on the web first.  When it comes time to using your personal data, you’ll download the desktop version!

  If you’d like some ideas on how to structure your Flex/AIR projects in Flex Builder so that you can build both web-based and AIR-based applications from the same Eclipse workspace, may I suggest reading my little tid-bit on How to compile both Flex and AIR Application from the same codebase.