Don’t delay because places are filling up fast.

The first thing to note is that there are two types of validation: binary and meta – in that order. Binary focuses on the app itself and meta focuses on the meta data you added as part of the submission process, including the app description etc.

The binary tests are as follows:
- Download, install, launch – your app should launch from the store and from a native location such as a desktop icon or program menu
- Primary functionality – buttons, features, menu items, levels, game play should all function correctly
- Check for viruses and malware – obviously, these should be absent
- An actual person will test and validate the app fully
- You can check guidelines on the Intel AppUp developer program website here

Meta validation test, which takes place after the binary validation, includes:
- A check on the content rating, which should be appropriate to the content of your app. Specific guidelines exist to help with this – for example, if the app accesses social media accounts such as Facebook or Twitter, it must be rated as 17+
- A check on the long and short descriptions you provided on your app -these should describe the content
- The validation team then assigns categories for your app to feature in (taking into account the suggestions you made during submission)
- If you have only changed the meta data since your last app submission, the binary validation is not needed and your app will be validated much faster

Common pitfalls to avoid:
- Apply for Microsoft code signing as soon as you join the program to avoid the five day wait that this entails when you want to submit your app
- Beta test your app on clean devices, not on development machine, and make sure you use basic factory settings
- Windows 7 starter has basic drivers, so apps cannot assume any resource is pre-installed
- Your app doesn’t launch from store due to relative path in the source code. Use the full path for application drivers and resource files. Beta testing will help to reveal this issue
- Creating MSI with bad shortcuts

Stefan Englet, European developer program manager at Intel, spoke at AppUp Elements 2011 on one of the things developers can do to increase their potential market and monetise their app more effectively: localisation.

It shouldn’t be a surprise to learn that the top 50 apps on the AppUp Center get 50% of the total revenue through the store. The top 1000 apps get 98% of the revenue. The wealth is not being shared – great for the top few but not for anyone else. The need to find ways to sell to more users is great, but why should localisation, rather than better marketing or better quality apps, be the route to achieving this?

The reason is in the numbers: 52 percent of AppUp Center consumers speak English as a first language. This is significantly more than any other language
and yet by only building apps in English, developers are cutting out half of their potential market. The total breakdown for the AppUp Center installed base runs as follows:

English – 52%
Spanish – 13%
French – 12%
German – 10%
Italian – 7%

Some apps need only change language in order to localise – Englet used the example of My Little Artist, a drawing app by Dmitri Rizshkov which supports English, German, French, Spanish and Russian.

Other apps need more effort to localise, such as Lugdulo’V by Corentin Chary, an app for finding bicycle rentals in the local area. To extend this app to new audiences has required more extensive modification – new maps and information for different cities around the world. Once updated for new regions, however, the app is instantly useful to people in those locations.

It would be good to understand what the above developers achieved in terms of additional downloads from those regions but unfortunately this wasn’t part of the session.

Given the potential market increase with localisation and the fact that Intel has now localised the AppUp Center and developer program, the company is offering developers some support in getting their apps viewed by non-English language consumers. Support offered includes:

- Validation of localised-only apps – the AppUp Center used to require English language meta data but this is no longer the case
- Support for entering new markets – advice about localisation
- Localisation tools for developers

Certainly something for mobile coders to think about but some case studies showing increases in downloads might help convince developers that it’s worthwhile putting time and effort into localising their apps.

In case anybody missed it, yesterday the Linux Foundation and the LiMo Foundation announced Tizen – a standards-based, cross-architecture, open source software platform. The blend of MeeGo and LiMo, along with some ‘new ingredients’, will support smartphones in the first instance, before extending to tablets, smart TVs and in-vehicle infotainment in the future.

So far, so MeeGo. What, I hear you ask, differentiates Tizen from Intel’s previous efforts?

There are a few things, as it turns out. In a lunchtime discussion with Apeland and developer champion Peter Biddle, it transpires that Intel knows it has stretched developers’ patience in some areas but really believe they’ve cracked it this time. Tizen’s LiMo heritage, for example, meets the needs and business models of operators and carriers, giving it broad support from this community. Intel also pledges an SDK in early 2012, with devices to follow in the second half of the year – a close enough timeframe to gain developer interest.

Tizen APIs are based on HTML5 – a great thing for getting your app across multiple devices and platforms, but less fun for coders who have poured their efforts so far into Qt. Apeland was quick to assure the Elements crowd that any software so far created would continue to be compatible, however at the same time, Intel is handing out free HTML5 support kits to its developers to help them get started on the new technology. These include books, educational DVDs and a month’s subscription to tutorials website Lynda.com – and some red liquorice, oddly.

The move to HTML5 came, Apeland described, from Intel’s efforts to rewrite its AppUp Center so that it would work across devices and platforms. Rather than keeping rewriting the client, the engineers decided to put it in HTML5 so that very little was left to do to optimise it for new and emerging device types. It seemed obvious that the same model would work best for AppUp developers too.

Apeland commented, “It means you can have an app that is local and takes advantage of those capabilities but also accesses cloud and web, taking advantage of those. This gives it both native and web advantages.”

It’s not all new, however. Intel Encapsulator already supports HTML5 -packaging it up and enabling it to run on Windows and MeeGo.

The keynote concluded by promising more info about Tizen and the associated SDK soon – the first to know will be existing AppUp developer program members.

Apologies for the camera quality – I was caught short on a decent camera at the time!

So far, progress has been swift:

• 15 February 2010: MeeGo was announced at Mobile World Congress in Barcelona.
• 26 May 2010: MeeGo 1.0 for netbooks was released.
• 7 July 2010: MeeGo 1.0 Update 1 for netbooks was released.
• 30 June 2010: Handset user experience project day 1.
• 27 October 2010: MeeGo 1.1 final release.

A single SDK supports the development of both netbook and handset applications.

Stylianou gave an overview of MeeGo compliance, a way to ensure that your app will run on any MeeGo compliant device. If you write an app for a netbook, it will run on any make of netbook, provided that device is certified as MeeGo compliant. The same applies to other device classes, such as smartphones and tablets. Each device will have to meet a minimum specification for screen resolution, sensors and so on to be awarded MeeGo-compliant status. While there are reference user experiences (UXs) provided for a wide range of devices, compliance with the reference UX is not required to achieve MeeGo compliance. There is a compliance test suite that you can use to verify applications, devices and MeeGo distributions. Compliance is overseen by the MeeGo Technical Steering Group and use of the MeeGo brand is granted based on compliance test results.

There were a couple of slides in the session which I thought would be particularly interesting to share with you here. Firstly, there was a slide outlining the MeeGo architecture (click the image below for the full size version). This shows how the MeeGo core covers most operating system functions, but how the user experience reference UIs sit outside of that so you can use them as-is or customise them:

The second slide highlights the tools and outlets available for MeeGo developers. As the slide shows, you can use the same code to develop apps for the Intel AppUp center and for Nokia’s Ovi store.

The MeeGo strategy is particularly interesting in the context of the Intel AppUp center. That app store is designed to sell apps across a wide range of devices, on both Windows and MeeGo (in future). MeeGo itself aims to provide a single operating system platform that underpins a wide range of devices. Together, the idea is that they enable developers to create apps for a wide range of devices with only minor changes to the user interface, and to sell those apps through a single store front.

MeeGo has a six-monthly release cycle, with the next major release expected in the Spring. At CES, there was a chance to see the netbook MeeGo experience and the WeTab MeeGo tablet. It will be interesting to see what the rest of 2011 brings!

The steps to getting your app in the store are:

1. Register for the Intel AppUp developer program. You’ll need to be a program member, which normally carries an annual fee of $99, but which is currently free. If you want to sell your apps, register using an email address that has a PayPal account attached to it. No other payment methods are currently supported.
2. Download the software development kit (SDK). The aim of the SDK is to enable your app to interact with the store architecture, so it doesn’t include (for example) video streaming or gaming features. The SDK enables authorisation, instrumentation and crash reporting. Authorisation is required and checks that the user has a valid licence to use your app. Instrumentation is optional and enables you to understand which parts of your app users spend most time in. Crash reporting is optional but recommended, and enables users to send crash reports directly to you through the AppUp architecture. The SDK is compatible with the Visual Studio 2008 IDE and includes an emulator you can use for testing your app’s compatibility with the Intel AppUp center.
3. Get a GUID. This is a 128bit authorisation code that uniquely identifies your app and is used for managing app licensing. When you start a new app in the Intel AppUp developer program, you can provide a unique name for your app and then receive your GUID. You need to copy this and paste it into your app’s source code. For debugging purposes, you can use a testing GUID, made up exclusively of 1s. Don’t forget to replace it with your real GUID later.
4. Develop your app. For the purposes of AppUp, the main differences are that you need to incorporate the authorisation code and GUID. The ATDS software debugger simulates the app store so you can test your app will validate okay.
5. Package your app. It’s ideal to use an MSI installer and you must ensure it performs a silent install. That means it’s okay to tell users what’s going on, but you can’t require them to make decisions or confirm options to complete the installation. You can find the full packaging requirements here, and they’re strictly enforced. You need to add at least one shortcut and have all shortcuts pointing to the same executable. You also need to check for the correct runtime and include it in the installer if necessary.
6. Upload your packaged app. You can add beta testers by email address, and they will be provided with access to the app through the Intel AppUp center. Beta testing will give you confidence that the app can be downloaded and installed correctly.
7. After testing, submit your app. Validation takes seven calendar days. You can set a start and end date for availability for your app if you want to, but otherwise it will go into the store until you decide otherwise.

The talk was similar to one presented in Berlin that same week, and you can find the slides here (AppUp SDK details, linked at the end of the article).

Nina Woolvett, business development manager for the Intel AppUp developer program, took to the stage to deliver a talk about turning your code into cash. She started by outlining Intel’s app store strategy, which involves making the Intel AppUp center available for free download online and pre-installing it on netbooks sold by leading manufacturers and retailers, including Dell, Toshiba, and Dixons Stores Group.  Customers can buy apps from the Intel AppUp center and then use them on up to five netbooks or Intel AppUp-compatible devices. Apps can be easily shared with family members and can be recovered if the netbook is lost or damaged. All apps are validated to run on netbooks, and customers have a 24-hour period to try the app before they commit to buying it.

The ultimate goal is to create a seamless experience across the new continuum of computing devices, all of them connected to the internet and always on. While the Intel AppUp center caters for netbooks at the moment, it will expand to serve tablets, set top boxes, and smart phones.

That’s the end user proposition, but what’s on offer for developers? They can get up to 70% of the revenue from app sales, and developer incentives including funding from Intel through the Million Dollar Development Fund. Its Accelerator Program enables developers to outline their ideas and get between $10,000 and $25,000 funding to support development.

Woolvett also shared some compelling numbers. Intel’s app store currently has 1,700 apps, 30,000 registered members and seven store fronts, including those provided by retailers using Intel’s AppUp center. Intel has had 500,000 unique visitors for its store already, and was projecting two million by the end of 2010, and over 50 million by the end of 2011. Total netbook sales worldwide were on track to reach 58 million in 2010 alone, with the UK accounting for 12% of that total.

Netbooks are the start of the story for the Intel AppUp developer program, but the store will grow to cover other Atom-powered devices as they emerge, so an investment made now could enable developers to seize early adopter advantages in new software markets in the next few years.

Woolvett’s message was that if you want to make money from your code, whether it’s for an existing app or a new idea, publishing through the Intel AppUp developer program could be a strategy that pays off.

What I will outline though is some trends that are already in flow which developers need to be mindful of as they look ahead to create new applications.

Perhaps one of the more obvious trends will be the continued morphing of form factors.  For a long time, computers all looked more or less alike. Lately we’ve seen new shapes and sizes blurring together to create ever more powerful and smaller devices. That said, the netbook will hold and grow its ground. It’s just that we will see more devices spinning out of this platform such as tablet PCs.

I mentioned this in an earlier posting referring to Intel’s CEO Paul Otellini who was keen to highlight the benefits of the soon to be released ‘Oak Trail’ Atom-based System on a Chip, designed specifically for netbooks and tablet PCs.

The DeviceGuru blog has done a good job in highlighting what it sees as the top eight mobile trends for 2011. For example, it emphasises good visual design and predicts that the market fragmentation will continue. These predictions are not rocket science and are points we’ve been emphasising on this blog for some time, but they do illustrate where we are in the market today.

However, I’m going to stick my neck out and advise developers to consider the MeeGo operating system as a platform to develop on. At the moment it’s the only truly cross-platform OS that spans a wide range of devices. Because of this, and the current market fragmentation, I’m betting we will begin to see greater adoption in 2011.

There are a few interesting posts on the Focus Q&A page about smart phone trends for 2011. One comment says there is “a need for app devs of all sizes to put more emphasis on [usability] design and the full user journey pre-download, post-download, and how the app delivers true value and impact.”

This is sound advice and ideally should inform the development of any application. Another interesting point made in the feedback on the Focus page is a reference to maps integrated into applications. I believe location-based services are barely tapped in mobile applications and there is much more to be leveraged. They can be hugely powerful within the right context, particularly for organisations who want to advertise their services to mobile users.

From the hardware perspective we’re also going to see better quality images on mobile devices as processing power enables more pixels per square centimetre. This will dovetail with an even greater surge towards the development of mobile applications. However, as the mobile app market continues to mushroom, and almost chaotically at times, the emphasis is going to be on apps that matter; that is, apps that people want.

I’m going to focus on this latter point in future posts as well as other relevant areas. So watch this space and have a good 2011!