Ideas You Can Use

Jan Ole Suhr, from Mobileways.de in Berlin, Germany, is the developer behind the smash-hit Twitter app called Gravity. If you have not heard of Gravity before, it is likely that you work for a company that does not develop Symbian applications ... And if you Tweet from your Nokia S60-based device, Gravity has quickly become the app of choice for those in the know.

We recently had a chance to catch up with Jan, and learn more about his experience creating – and now continually refining – Gravity. His story demonstrates the potential that many of us see in "apps" for Nokia mobile devices, and serves to inspire developers around the world. Read the Q&A below to learn about his experience; discover his sources for inspiration; and learn what he suggests other developers working in the mobile space should try to do. His answers might surprise you! 

Q: Why did you decide to create Gravity? What was your inspiration?

A: I was using Twitter for quite a while before starting with Gravity.

Once I got more and more friends on Twitter, I figured that I needed a good client on my phone. With all the buzz around mobile Twitter clients for the iPhone, I thought there must be something similar or even better for S60, too. At the same time, a couple of friends asked me to write a native Twitter client for S60. I decided that Gravity could become a showcase product of what you can do with S60 and Symbian if you just tried hard - and at the same time I was programming it for my personal needs.

Q: How long did it take to finish the original version of the app?

A: I started working on Gravity around the end of January 2009 and released the first version on 30 March of 2009. I was reusing a lot of code from earlier projects, though. A lot of time went into designing the user interface and version 1.0 was lacking quite a lot of features I initially wanted to have in the first release. I still decided to publish early because of the tremendous feedback and interest spurred by the beta testers.

Q. Tell us about your ongoing efforts to improve/refine the app – we read recently that you are regularly working to fix minor bugs and add new features … Does it ever end?

A: Right after I released the first version of Gravity, I was flooded with feature requests and bug reports. At the same time, I received a lot of positive feedback and comments praising Gravity, so I realised I couldn't just stop with version 1.0. And, of course, there really were a lot of features missing that I needed myself. Luckily, the bug reports concentrated on three or four major flaws which I was able to fix pretty quickly. Fixing bugs isn't much fun, but adding new features is great!

And then a mutual process started which is still in place: users of Gravity were asking for features, and once some of the features were released, I received a lot of postive feedback and also, new feature requests. So the process started again: adding and releasing features, lots of positive comments, and new feature requests. This has become very encouraging and almost addictive.

Q. What helped you to decide that this needed to be a Symbian application – instead of Java, or other technology?

A: I've been programming in Symbian C++ for the past six years, so for me, there was never really a decision to be made about which platform to choose.

Q: Were there any particular problems that you encountered along the way? If so, how did you resolve those issues?

A: The biggest obstacle I'm still facing is the network connectivity.

In theory, it's pretty easy to write good wrappers for HTTP-based APIs in Symbian. But in practise, making an application "compatible" with all the wireless networks in the world seems to be very complex. Add some misconfigurations and inconsistencies with the Access Point / Network Destinations user interface to this, and you've got support work for years to come!

The other big issue was – and is – Symbian Signed. Some features are still missing or not working 100% correctly, because Gravity is self-signed and not Symbian Signed. I have chosen to avoid Symbian Signed for the moment because of my current pace releasing beta and “final” versions of Gravity. There's no solution to this right now, but I've heard that the Symbian Foundation is working on it.

One more issue is the user interface (UI) for Gravity. This is the most important part of the application, because the UI is what makes Gravity so exciting. It wanted it to look as polished as the UI of an iPhone application. But there are two problems: first, I am just a developer and "designing" a UI and picking colours just takes ages for me. Second, we've got dozens of different LCDs to design for, and also keypad and touch-based phones. As an example, a gradient might look great on the N95, but on the E61i, you barely see any difference between the dark and the light colours. Creating a UI that works on both keypad and touch-screen phones, that can be compiled from the same source, and still give the best user experience, was another tricky thing to overcome.

Q: Has the popularity of Gravity helped to expose consumers to other apps you have created?

A: That's very difficult to say. I'm currently not doing any statistics that could prove an increase of sales for the other main product I'm offering. The other application is a tool for remote controlling S60 phones, so maybe it's just a completely different group of potential users than people being interested in Gravity. On the other hand, my website did see an enormous increase in traffic after the release of Gravity.

Q. And, does work on Gravity mean less time for other projects? How do you balance that out?

A: I'm currently working 100% on Gravity because of its success and because it's so much pleasure to get the positive feedback when releasing new features.

Q: Tell me about your design decisions and the reasoning behind them. What did you want Gravity to do that you won’t find in other Twitter apps?

A: I wanted Gravity to become the best mobile Twitter client - not just on S60, but for all platforms; and I especially wanted it to be better than what you find on the iPhone. The iPhone has a superb UI/UX, so that's why I spent so much time on the user interface of Gravity. There's still a way to go to achieving this goal, but I think I'm on the right track. It's great to see so much interest in a Symbian/S60 application, because nowadays the iPhone platform tends to dominate so much of the tech news we see.

I also wanted Gravity to run in the background and notify me of new tweets or direct messages - a feature that was impossible to have on the iPhone (before v 3.0).

Feature-wise, when it came to releasing Gravity, there were no mobile Twitter clients with support for Groups and multiple saved searches. These two features are very important to me, because otherwise I wouldn't be able to keep up with the news on Twitter.

Q: Did you get advice or feedback from other developers in the Forum Nokia community - or through Nokia employees - as you worked on this app?

I regularly visit the Wiki at forum.nokia.com, because it's such a great repository of code examples. Of course, there was also feedback from Nokia employees, but this was more on a personal basis.

Q: What advice would you give to other developers - particularly those who are just starting out - when it comes to creating mobile apps for Nokia devices?

A: I would suggest several things:

1. Buy yourself an iPod Touch. Seriously! It is a great investment to check out the UI and UX. Then go create something similar or better on S60, please!
2. If you've got a networking application, the first thing to implement is an auto-update feature to check for the latest release of your app. Never, never ever release a version without this. Never! Especially if you're part of a small development company. You'll lose a lot of customers, because they will still manage to find version 1.0 from three years ago, even though you've already released version 4.x.
3. Programme the application for yourself and try to make money with it. Granted, that's not always possible, but try to use the application as long and as often as possible. Use all of the features to get an idea of how your app works – is it easy to use? What's missing and what's not working at all?
4. Think of how to make money with your app first (only if you need to make money with it, of course.) Do not postpone this decision. It's not going to be easier to release a "Pro" version after some time than to think about making money right from the start.
5. Try to get some beta testers that are as enthusiastic about your app as you are! The feedback will help a lot if you get yourself into a dead end or just feel frustrated by bugs, and whatnot.

Now, some ideas you can use:

• Learn more about Mobileways.de
• Take a look at Gravity in Ovi Store
• Follow Jan Ole Suhr on Twitter

Update (17 September): As noted in the first comment below, Gravity is one of the apps up for the CTIA 2009 Hot for the Holidays web vote award, for the Best Mobile Consumer Application. Show your support for Gravity by marking the app as a Favorite. Vote Now.

Update (18 September): 10-Day Free Trial of Gravity is now available in Ovi Store. Try before you buy.