The developer's conundrum - TechCentral

The developer’s conundrum

[By Steven Cohen]

We’re a quarter of the way into 2011 and I am still a bit confounded about the direction I should be steering my development team here at Softline Pastel. It’s imperative that our business software remains at the cutting edge, but with so many conflicting options how do I know we are making the right development decisions that will appeal to our clients and still benefit our business?

A decade ago there were limited options when it came to choosing an operating system. About 95% of the market operated machines that ran on Windows, Apple was considered a niche tool for the design space and offered no business applications, and Linux and Java were only used for speciality operations. So it was an easy decision: we all developed for Microsoft.

The problem facing software developers today is that there is currently no perfect solution to getting applications running with all the features that they’d like on the multitude of devices that are appearing in the market.

And it’s not just Pastel that is perplexed. No sooner had Hewlett-Packard announced in late 2010 the release of its Slate, an iPad rival, they delayed the release by acquiring smartphone maker Palm in order to gain access to WebOS, which is considered a very good smartphone operating system and a better option for its tablet devices.

Devices such as Apple’s iPad and iPhone have proven that applications written “natively” for the device’s operating system deliver a far richer experience as opposed to delivering these same applications through a browser interface. But the challenge with “native” applications is they require a developer to learn the platform’s language and development environment as well as have an intimate understanding of each device.

The advantage to this is that teams are challenged and stretched. And, technically, the applications are responsive and you know that they interact well with the device hardware and operating system. It’s exciting stuff, but the number of devices is already too great for a single developer or small team to master.

Even Microsoft is confused. It’s been punting Silverlight as the silver bullet to running Microsoft applications across myriad devices. But now the software giant is turning its attention instead to HTML 5, much to the chagrin of many software developers.

In theory, HTML will run on any device. In practice, you still have to test it and work around different implementations of the HTML standards, although the new HTML 5 promises to render better than before. But the Web requires an active Internet connection and can be very slow to download — a real challenge facing local businesses.

So what to do? We’re taking the middle ground for now and developing our programs for the Web in conjunction with the PC. Users will have the option of running their applications in the cloud or on their desktops. It’s a case of old meets new as young — as tech-friendly entrepreneurs embark on new business ventures they tend to prefer cloud accounting over PC-based software. We’re keenly waiting for the demand.

  • Steven Cohen is MD of Softline Pastel
  • Mitch Wong Ho

    I’m not quite buying the slow connectivity excuse. How much downloading will Pastel customers be doing? or is it that most are still running IE6 (lol)

  • Joel Job

    You can’t fault Microsoft’s Silverlight mission – it will become the web development standard of the future. It does not compete with HTML5, but rather complement it. WP7 and Windows 8 are deloped on a Silverlight platform.

  • @Joel I’m not sure what you mean Windows 8 is developed on a Silverlight platform? Win8 is simply Win7 that will scale up and down better, add an ARM version and probably add a touch interface.

    The article comment about MS shifting its focus to HTML5 from Slverlight is incorrect, it’s based on a bad article that spread through the tech sites. MS issured a formal correction, which wasn’t nearly as exciting so didn’t spread as far, so lots of people still think MS is pulling back on Silverlight, but they’re aggressively expanding it and repositioning it as a OS-agnostic app platform (not web platform, @joel 🙂 ) which will be key when they roll out their tablet offering… eventually.

  • Facebook has not suffered from any speed limitations SA users might have. The biggest challenge here is a user base with patterns entrenched by the legacy application.
    You will increasingly see your user base being eroded by the new generation of cloud (aka web) apps like Freshbooks, Curdbee, Xero… These developers carry no legacy, they focus on areas of significant user benefit and do it really well. These apps don’t suffer feature bloat.
    The real issue is that the cloud is typically and paradigm shift for application developers. This is more than a conundrum and has nothing to do (really) with OS.
    Interfaces need to be pulled apart and re-thought – think modules that represent user processes. Technically it means looking at a language often dismissed and difficult to really master by application developers – javascript.
    Another significant issue in moving to the cloud is architecture to support these interface modules. Interfaces need to be completely decoupled via an API.
    Moving from application to cloud will also require the expertise of running a data centre as you now take responsibility for the availability and integrity of customers data…

  • Another take on the same subject but from a different perspective: which asks the question, “Will the Web have a place in a world full of apps?” This, in turn, examines some of the issues raised by Wired Magazine’s editor Chris Anderson’s (in)famous article in the Sept 2010 edition entitled “The Web is Dead”.