It’s true.

We built our startup in Flex, having spent six months wrestling with HTML/JavaScript and browser compatibility. Flex offered enormous advantages that we just couldn’t overlook, so we dumped six months of development effort and started again .

Adobe have been a great supporter of us for obvious reasons, and we presented during a keynote at one of their recent roadshows.  Like any startup we study the competition, but never did we think we would have to worry about Apple.

Apple vs Adobe

The current Apple vs. Adobe saga has us in a bit of a quandary. With each new revelation – culminating in the change to Section 3.3.1 of its Developer Licensing Agreement – it has become clear that Apple is not just dismissing Flash as buggy or slow.  Adobe can improve Flash and address the concerns of Apple all they want, but it won’t make a lick of difference..

Make no mistake – Apple is making a strategic business decision to bar Flash from the walled garden entirely.

The licensing change canned the new CS5 feature which allowed the Flash platform to compile to a native iPhone (or iPad) binary. It was Adobe’s attempt at an end-run around Apple’s earlier exclusion of the Flash runtime in not only the Mobile Safari web browser, but in any apps submitted for approval to the App Store.

Sorry Adobe.  Apple says No.

Why should we care?

Since launching the iPhone, and in particular the App Store, Apple has rapidly gained around 25% of the US smart phone market. Developers and end-users alike have flocked to the platform. Apple have continued this march on the mobile computing arena with the launch of the iPad.

Like many startups, we had plans to join that wave. Even though our primary user interface was written in Flex, we planned on developing an iPhone app available for our users as a key part of our strategy.

The Flash CS5 export feature seemed the perfect choice. We could write an app for the iPhone with a limited feature-set and Apple UI components, but for the iPad we could build an app very similar to our current rich internet application. That grand scheme came crashing to a halt when Apple changed the Developer Agreement.

But perhaps more important than our ability to easily compile our product into a working iPad/iPhone application is the future of Flash in general.  Apple’s star continues to rise with significant share of the mobile market, and it is fairly clear Flash plays no part in their plans.

This raises two burning questions:

Should we pivot and embrace an Apple friendly technology, before we get wiped out in the stampede of mobile devices?

and

Should new startups avoid Flash/Flex as a technology choice?

Before we answer, let’s consider why we chose Flex in the first place.

You chose Flex?  Are you nuts?

It was a conscious and educated decision to use Flex.  We don’t regret it.

Our decision to use Flex for our user interface was based on quite a few factors, including the Flash experience of our lead developer. Flex offered us the best opportunity to rapidly progress from an idea to a product. For a startup, this is critical.

The interface we were aiming for was advanced in terms of the functionality we wanted to deliver to the customer, and the interactivity of Flex was an excellent fit.

At first we were concerned about the availability of Flash.  It turns out however that Flash has made leaps and bounds in market penetration in recent years, hitting 98% penetration of internet connected computers. Compare this to the number of people using outdated browsers that makes rich internet application development in HTML/Javascript a compatibility nightmare, and Flex becomes a serious contender.

Right tool, right job.

We still had some concerns with Flex. We have always been proponents of open source platforms, doing nearly all of our server side web application development in PHP, MySQL, Linux, Kohana and Eclipse. Although most of Flex is open source, some is not (the Flex Builder application and of course the Flash Player itself). Add to this the fact that even we have our share of Adobe Support horror stories and you can see why we might have balked at implementing our baby on their platform.

Overall though we strongly believe we made the right choice.  Our speed of development has been quicker, our features richer and our browser testing regime smaller.  Flex was and is the right tool for our particular needs.

Flash does have a use by date however. Sooner or later, browsers will be able to do what previously needed a plugin. It has already begun with the adoption of of HTML5 by some browsers.

Is this use by date going to be in the next couple of years? It’s highly unlikely. While HTML5 is an exciting open standard in theory, in practice there are browser compatibility issues even amongst early adopters like Firefox, Safari and Chrome.

Okay you built it in Flex – what’s your next move?

If like us you built your startup in Flex or Flash, then you have at least five areas for which you need a solution.

Browser based via a Desktop Computer

You already have a solution, and it’s called your existing application.  Yes Flash has a use by date, but it’s not likely to be in the next five years.  Stick with what you have and build the best damn product you possibly can.  Focus on the customer, not the tech.

Status: No change required

Desktop Program

While not everyone needs to consider this, we think it has potential for our ticketing application.  Imagine processing entries to a music festival set in a farmers paddock – wouldn’t it be nice to do that on a desktop app and not have to worry about dodgy 3G connections?

Here our choice of Flex is in a class of its own.  It is relatively trivial to port our code base to an Air application, giving us the best of both words.

Status: No change required

Mobile device running Android

Android will support flash, and it is clear they intend to support it for the foreseeable future.  Like us you may be better off writing a specific mobile view for an android enabled device, but you would have had to do this anyway if you had written your app in HTML/JavaScript.

Status: No change required

iPhone

Here you have two possible scenarios.

1.  Write a dedicated app

2.  Write a specific web view

Either option requires development, but both also would have been required if you had written your app in JavaScript.  It is technically true the iPhone will render a native HTML/JS app right now, but anyone serious about this platform is doing the work to make it iPhone specific.

Status: Write a specific app or web view

iPad

This one has the potential for the most amount of rework.  The larger screen size makes it likely that an existing HTML/JavaScript web application will look perfectly fine on an iPad without any extra work.  Yours and our flex/flash apps will look completely pants.

In the interim, you can take the work you have already done on the iPhone (that you would have had to do anyway) and port that across.  Job done.

But you know what the real answer is?

iPad Schmipad.

It’s a device that has been out for under a month and is still only available in the US.  No-one but early adopters have one, and the actual success of the device is still up for grabs.

We’re not saying it has no chance of completely disrupting the laptop/mobile device market, only that you should wait until it does before making any critical development decisions.

Take the Public Enemy approach until the numbers are in.  Don’t believe the hype.

And if we are wrong about the lifespan of Flash?

We were wrong once back in 1991, but no-one remembers that and all records were suppressed.

But okay…on the off chance that we are wrong for the second time in our lives, the solution is easy.  We make a technology pivot, and move away from flash to whatever is the best fit for our model and our customers.

As we were developing our application, we paid special attention to how it interacted with the back end. From the start we felt the best way to do that was to create our own API, making even our own application interact with our data through the API. This left our Flex application as responsible for the user interface, not the application itself.

How does this help?

It allows us to change the interface if needed without a complete re-write. It also allows us to build complementary applications on other platforms – as we were intending for Apple’s iPhone. This should mean that even if Flex/Flash does have a shelf-life, we can adapt.

“There is no standing still on the web“.

What should others do?

If you already have a Flex/Flash app, be happy you got off your bum and launched something.  Good for you!  Flash remains  pervasive and will be for the foreseeable future, so unless mobile is your entire business model, stick with it.

If you are just starting out and still to choose your platform, our advice is this.  Choose the right tool for the right job.  Don’t just jump onto Ruby on Rails or HTML5 because all the cool kids are.  If you have a specific problem to solve, think about the best way to solve it and then pick the right tool.

If you don’t have a specific problem to solve, then you have much bigger problems and we can’t help you.  Go straight to the Deadpool, do not pass go, do not collect $200.

Otherwise, enjoy startup life.  It’s never boring!