Adobe or HTML5?

I know, some people will say it’s apples to oranges. Some people will say Adobe is also pushing for HTML5. But I’m writing this for people who read the title of this blog correctly: Flex or JavaScript.

Disclaimer 1. I do both.

Disclaimer 2. I don’t care about game programming. I work for a company that does enterprise development.

In the enterprise IT circles Adobe is a whipping boy of 2012. Adobe has accepted this award, and only their cleaning ladies and janitors didn’t apologize for what their CTO and CEO did last November hiding behind the wide back of Danny Winokur. But let’s put emotions behind and get real.

Today, there are two types of the enterprise CIO’s: those who don’t care about developing for mobile and those who do. CIO’s who don’t care about mobile can be divided into two groups: smart and typical.
Smarts will start new Web projects selecting a technology that would allow them to easily go mobile if the life will force them to do so. And not so smart will just go HTML5 because this buzzword is an easy sell to CEO’s.

Let’s continue splitting hairs. The enterprise Web projects can be divided into two categories – Web pages and Web applications. Web pages can be divided into (you guessed it right) two categories – those don’t have to be blazing fast (just show them that Donate or Pay button), and those that must be blazing fast.
Further splits would involve Intranet vs Internet, and by reliability requirements.

Now the most important question, “Who do you care about the most?” As the saying goes, men who shave in the evening love their wives, and those who shave in the morning love their work. Who can you afford to love more – developers or the users?

Seven years ago the most popular examples of successful AJAX applications were Google’s gmail and maps. Most of the people were drawn into this Google state of mind forgetting that their own teams had less skillful developers and more modest budget than Google’s.

Today, people are choosing the technology to go mobile. In a meeting with a prospective client I’d start drawing on the white board. You have two choices… I’m not going to do it here, but learn from the past – don’t assume that what the big guys do is affordable to you too.

I’ve learned from a Facebook engineer presentation that their users login from 2500 different mobile devices. On the other hand, this article states that Facebook has bad experience with their HTML5 mobile application and they are creating an iPhone version in Objective-C. Sounds great. What about people carrying 2499 other mobile devices.

Do you have to make such choices? Do you have skillful teams of Objective-C, C++, and Java programmers or prefer to have the same code base in HTML5? How many concurrent users do you have? Do you have enough money to cater to all possible devices your users carry?

It’s up to you to decide. But for most of the enterprise mobile Web applications the choice comes down to selecting between Adobe AIR and JavaScript. If you select AIR, it’ll cover all your today’s intranet desktop needs and the same team will be able to port this app to iOS, Android and Blackberry devices. If you are ready to dive into JavaScript, just double the time allocated to the development cycle of your project. But you may save time in the long run cause your Web application will already work on lots of mobile devices, especially if you’ve been using responsive design principles.

As to myself, my typical week is split between Adobe AIR and JavaScript. And yours?

Yakov Fain

8 thoughts on “Adobe or HTML5?

  1. I actually have great hope for Apache Flex and Adobe Air. Going from Adobe to the Apache model has been much harder than I think most would have projected. However, they are almost ready for a parity release.

    Once they reach parity, my guess is a divorce from all the bulk and weight of having to work with Adobe Flash Catalyst will be one of the first big steps. Once that is done and Adobe Air 4.0 comes out with support for Windows Metro, I believe (on my more pessimistic days, I would have to say I hope) we will see a resurgence of interest in Apache Flex/Adobe Air.

    It really is much better for application development…IMHO.

  2. Well put, Yakov and funny as usual. In the enterprise space, technical decisions should be made as rationally as possible. No one technology – regardless of “buzz” or fashion – is the answer to all needs. There are a number of technical options that a good consultant must match to his clients’ needs/budget. JavaScript is certainly more expensive (development and QA) and has benefits and trade-offs. Flex is powerful and is built for the enterprise but has some platform limitations. In the end, you need to take each for what it is (not what you wish it to be) and apply it where it fits best while never promising things that the technology cannot deliver.

  3. I think “double” is being overly generous, once you start to factor in maintenance, and continuous testing costs.

    AIR remains relatively stable, whereas HTML continues to break for the entire lifecycle of the application, as new devices, OS’s and Browser’s are introduced.

    It may double your initial cost to get the product out the door. But then maintain it for two years, you’ll see much more than a 2x difference in cost.

  4. Michael, JavaFX will deliver on its promise when Java Runtime will be available on various platform. As of today, JVM is not available on iOS. Oracle and Google are not buddies either.

  5. Hi Yakov,
    great post, I personally spend my time on AIR projects because I think it’s a good technology that is growing well.
    In my company we tried a development with Phonegap for BB, Android and iOS but we had to write 3 times the same app because it’s not true that you can write once and deploy everywhere with Phonegap.
    ANother big pain of Phonegap are the performance, with AIR is totally different, you can have good performance if you write good code!
    Personally I’m not so happy to come back work with JS… it’s a deep dive to the past and there are many interesting technologies that are growing fast and well, betting in them could be a risk but probably that’s why I like them 😛

  6. It’d be great if somebody would take attention on Haxe, especially for enterprise development.

  7. I would use js (even if I’m flash dev)
    simply because I prefer linux for development and I consider this OS which run on most of our server – desire a better attitude.

Comments are closed.