5 keys to developing and maintaining a long tail of enterprise apps #AppsWorld
Picture credit: iStockPhoto
Simon Pitt, mobile team lead at the BBC, has told an audience at Apps World Europe of the five S's that helped the broadcasting giant build internal enterprise applications.
The trick, he told delegates, was that none of these methods was revolutionary, or even new: just best practice development ideas from 20 and 30 years ago, particularly if you've got a long tail of enterprise apps to manage.
The talk cued in nicely with the opening salvo from Cliff Saran, Computer Weekly technology managing editor, whose idea was 'back to the future', by showing off long dead tech products such as the Apple Newton and Nokia Communicator, comparing its feature set, such as multimedia, voice control and touchscreen, and relaying their relevance in today's smartphone market.
Pitt's five maxims for enterprise app building included simplicity, size and standardisation:
1. Simplicity. "We know that when you're building apps, and especially as an enterprise that has a long tail, to manage that level of complexity, you need to keep it simple," Pitt explained. With over 50 apps to manage in his workforce, it's a common sense tip to ease the workload and the pressure on the dev team.
2. Small. It's linked to simplicity: don't overcomplicate it and don't burden your apps with excessive code. Pitt explained the story of the first internal app built by the BBC - an app to let employees search the corporate directory. Originally there was a long string of code to allow users to search by last name, department and their start date. "And then we realised - nobody wants to do this," Pitt said.
The result: a single search box, inspired by Google, but much more effective for BBC employees. "Sometimes it's about telling our developers not to put more features in our apps - and it's quite difficult sometimes because some of those features are cool," Pitt added.
3. Standardisation. There's an awful lot of good information in the Apple, Android, and Windows Phone handbooks. You'd be crazy to pass it up. As Pitt put it: "Some of the richest companies in the world have done loads of work on developing these standards, and I'm not clever enough to come up with a better solution."
4. Single source of truth. This again overlaps with standardisation. It's the classic cartoon; if there are 14 standards, and a group is determined to create a definitive list, what happens? There are 15 standards. "This is really important if your data is in different places, it's all over the place," Pitt explained. "You limit what you can do."
5. ReuSable components. This is linked to simplicity. With 50 or 60 apps to manage, if they're all bespoke it's increasingly difficult to make each one of the same high standard. If two apps are likely to use the same data source or API, take advantage of that. Pitt explained: "Nothing is faster than not writing code because you've already written it."
Pitt concluded: "Those are the things that we've learned, and influenced how we can build apps. And none of these things are new, they're all things we learned 20 or 30 years ago. But sometimes we forget that.
"Sometimes we think we can implement hybrid apps, or Google Glass, and that will solve our problems. But it won't."
Interested in hearing industry leaders discuss subjects like this and sharing their use-cases? Attend the co-located IoT Tech Expo, Blockchain Expo, AI & Big Data Expo and Cyber Security & Cloud Expo World Series with upcoming events in Silicon Valley, London and Amsterdam and explore the future of enterprise technology.
- » Why businesses must bite the bullet and upgrade their legacy ERP systems
- » The quest for the perfect PC: Travelling through the cloud, via VDI and desktop centralisation
- » How the future of AI will put new demands on ERP systems – and how to overcome them
- » How the CIO needs to see the evolution of no-code platforms: Security, ML, and democratising data
- » How to deal with technical debt to fully go through the gears of digital transformation