News Blog

Salesforce has purchased a leading Ruby cloud platform provider Heroku for $212m. Heroku provides a scalable cloud environment to run Web apps written using Ruby – compared to Web hosting providers that offer Ruby on Rails as one development option (like the provider that Darteco uses; as you may recall, these pages as well as most Darteco-created Web apps use Ruby on Rails), Heroku is a true cloud system with full-featured scaling. It also uses the cloud business model, charging per usage. The announcement came only one day after Salesforce had acquired Database.com, a cloud database provider. There is a very clear synergy between the two deals: Most Ruby on Rails Web apps use databases heavily, enabling a straightforward cross-selling business model for Salesforce. With these deals, Salesforce is further strengthening its grip on the cloud computing business. Note that many of the competitors do not currently offer Ruby support, giving Salesforce a unique selling point at least for the time being.

No comments yet - write one!

Research in Motion has purchased the Swedish UX company The Astonishing Tribe (TAT) for an undisclosed sum. For a cool user experience, TAT has always been one of my favorites. TAT started with a graphical UI widget library already long time ago, and later extended its offering to include a full-blown declarative UI framework. During the past years, TAT has been working in close partnerships with many of the top-10 handset vendors, and participated in creating the original Android UI. Today, many of their design ideas still look fresh, and fit well to the large touch screens of smart phones and tablets alike. Not surprisingly, competitors have also emerged: most recently, Nokia's (Trolltech-originating) Qt framework with QML language addresses pretty much the same targets.

This move further emphasizes RIM's commitment to operate in the mobile platform market. HP's ex-Palm chief Jon Rubinstein just commented that the mobile OS race will be unlike the (Windows-dominated) PC OS race – hoping also that HP could be #3 in this race. iOS and especially Android are progressing steadily, Symbian still has a very strong position due to the large installed base, WebOS remains a bit question mark, and Windows Mobile and MeeGo will show their potential during 2011. With the TAT acquisition, RIM is definitely able and ready to play in this league.

Image © TAT

No comments yet - write one!

This is the second part of the blog, focusing on the technology behind the Hydra framework. I know there are still many bugs and areas for improvement, so I'll get back to the topic later with more learning points. From the very beginning, I knew that writing a cross-browser framework without jQuery or other similar libraries is a hard task. One function these libraries provide is, after all, to abstract away the browser peculiarities. However, as I left out IE from the supported browsers (the market share is currently small in the mobile context), the task became somewhat easier. Right now, Hydra works reasonably well on iPhone and different versions of Android; with Nokia browsers, I still have some work to do. And, with desktop browsers (Safari, Chrome and Firefox) there were very little challenges, indicating that things will get better over time, when desktop technology is adopted by the mobile community.

Overall, most challenges are currently related to placeholder handling. Placeholders are like labels which are shown within text input fields when no text is entered; some modern browsers support these natively, but Hydra provides emulation for those browsers that don't. The difficult part is password entry: As the placeholder needs to be shown in plain text, I first tried using input field type modification. This I quickly found to be subject to errors in many browsers, so the correct way is to use two input fields, and switch between these. Then, the next problem turned out to be a focus()-method bug in Nokia's Webkit engine; This I have not fixed yet. Another challenge currently is the lack of opacity support in some browsers: The styled version of select box relies on this, and I am looking for a good workaround – one which does not lead to the focus()-method bug mentioned above.

So if you haven't tried it out yourself, do it today. Remember to get always the latest version, as the package is continuously updated. And, remember to send me all feedback you have on Hydra!

No comments yet - write one!

After having created a number of mobile-optimized Web sites, I decided to invest some serious thinking and work on the underlying framework itself. In most cases, I had divided the system into two parts: an iPhone-optimized version, and a generic version. There exist several good frameworks for iPhone: I have mainly used jQTouch, an extension of jQuery, as outlined in my earlier blogs. Frameworks such as jQuery and Script.aculo.us work great on the desktop environment, and offer useful functionality for mobile implementations as well; however, they are fairly heavy, adding plenty of kilobytes on top of the actual Web app payload. I decided to go the hard route, and write a generic mobile-optimized Web app framework by myself. The result is Hydra; in this blog entry, I will go through my initial target setting. I will return to the technology part in my next blog.

The first target was to make the complete framework small. I set the target to 32kB, including all the JavaScript code, stylesheets and graphical files – everything except for the actual Web app payload. The second target was to make the framework generic, yet not to go for the lowest common denominator. This requires a bit more explanation: Although the global share of smartphones is only around 20% (of all new phones sold, meaning the installed base has a much lower percentage), a substantially large portion on mobile Web access takes place from these devices. iPhone, phones running Android OS, and new Nokia phones, among others – these all include a browser that provides a good user experience. The third target was to focus on usability and graphical outlook. This means the layout and individual graphical elements, particularly in menus and forms. And finally, the fourth target was to make the framework as simple as possible to integrate into various Web apps. Have a look and see how I succeeded – and do not forget to send me your feedback!

No comments yet - write one!

Netbook is dead, long live the tablet! An article in FierceWireless summarized what we have seen during the past year: global sales forecast for tablets looks very promising, whereas the Netbook has almost disappeared as a product category. Particularly badly Netbooks have been selling with wireless plans – and the article asks what would be different with tablets. Indeed, the sales figures by AT&T are very low, compared to the global sales reported by Apple, just to take one example (well, iPad has around 95% market share). The potential reason is interesting: users' attitude towards multiple wireless plans. Some time ago, it looked like many people will have more than one contract and SIM card. On the opposite path, there were devices like Nokia's Maemo tablets (prior to N900) that heavily relied on wireless LAN – and a cellular connection using a Bluetooth-connected phone. Now, users find it difficult to justify just another wireless plan; They may have an unlimited data plan, or at least the capping limit is sufficiently high, so that it makes sense to have multiple devices sharing the same cellular connection. Some tablets manufacturers such as RIM are therefore focusing on the combination of Bluetooth and WLAN, instead of a cellular radio. Of course, carriers may change the rules of the game if they so decide, by providing simple but flexible multi-device data plans.

No comments yet - write one!

Today is the launch date of my second pet project: apzaar, the cloud-based mobile Web app bazaar. Just like the mobile app stores, apzaar presents a list of mobile-optimized Web apps, tailored according to criteria such as popularity and locality. Each mobile Web app that has been registered to apzaar has a small plus-sign on the top left-hand corner of the front page; this button opens up a mobile-optimized sharing dialogue, enabling single-click sharing of interesting mobile Web apps with friends. The sharing system is based on user-defined sharing profiles: these profiles define sharing through Email and sharing in social networking apps. Within the apzaar-registered apps, the user also gets the single-sign-on functionality: The app may request the username of the user (but nothing else) from apzaar and use that to tailor the presentation and content. For mobile Web app creators, integrating apzaar means adding a single line of code to their page; in addition to plain HTML content, apzaar can also handle DOM-manipulating libraries such as JQuery.

So, you may freely go and try it out - note that the whole package is very early Beta, so I am not too surprised if you will find a bug or two. I would be more than happy if you try and add your mobile Web app to apzaar. In case you later want to take it out for some reason, just drop me an Email and I will remove it. In the first version, not everything is finished yet: sharing works using either Email or Facebook, but expect to see one or two more options soon. The only supported JavaScript library is currently jQuery (including jQTouch plugin); again, there will be more in the coming versions. The implementation uses Ruby on Rails on the server side, and JavaScript combined with iFrames on the client side. I have tested apzaar so far with iPhone, Android emulator, selected Nokia phones, as well as Safari, Chrome and Firefox browsers, so all feedback is highly appreciated!

No comments yet - write one!

The dispute regarding the Java-like technology in Android's Dalvik virtual machine is still ongoing, and it will presumably take some time before it gets settled - one way or another. What is the real motivation behind the case? I find it rather unlikely that this would be primarily about licensing fees. Most if not all handset vendors have a Java license, but the total license revenue Oracle receives from this market segment cannot be much more than a rounding error at their bottom line. Also, it looks to me that Google has been careful not to tie Dalvik to Java branding - another potential source of trouble.

My own theory is that this is all about platform leadership. In the mobile context, Java has been losing market significance during the last couple of years. Although it can be found in most phones, apps - even games that used to be a Java stronghold - are increasingly developed on top of the native API's; Alternatively, they run in the browser and are based on Web technologies. With Android steadily gaining market share (18% of all open-OS phones now, and almost 30% in 2014, according to Gartner), it may lead Dalvik becoming the Java (without being branded as such). This would effectively bring the de-facto Java out of Oracle's control. Operators have been traditionally favoring Java, as it has provided a (relatively) uniform app environment across different handset vendors and models. Today, operators are more focusing their initiatives around Web technologies. If Oracle wants to keep the official mobile Java alive, they must act now: in addition to the Google case, they also need to create a positive spin around the brand and revitalize the technology, in order to make it a winner.

Image © BMW AG

No comments yet - write one!

A Palo Alto, CA -based startup has released a beta version of a HTML/CSS/JavaScript framework, dubbed Sencha Touch. The framework enables developers to create native-like Web apps for the iPhone and phones based on the Google Android platform. If you have been following my blog or listening to my presentations, this presumably sounds familiar. In fact, Sencha Inc. was formed by renaming Ext JS Inc. and getting onboard couple of experts and their products: jQTouch by David Kaneda, and Raphaël by Dmitry Baranovskiy. The latter one I haven't personally used, but the former one I am very familiar with. jQTouch is a Web app framework, comprising of iPhone-optimized CSS libraries and jQuery JavaScript extensions to support development of iPhone-flavored Web apps. Similarly to jQTouch, Sencha Touch includes API's to access selected phone features and peripherals such as the GPS location info.

Sencha Touch also signals the starting commercialization of mobile Web app frameworks: Unlike jQTouch which is available under MIT license, Sencha uses a dual-licensing scheme where you either purchase a commercial license for your products, or make the source code for your app available under a corresponding open source license. As I think the framework and the new team behind it are very interesting, I'll keep an eye on the development here. Let's see if any of the big vendors steps in and acquires the startup - it would definitely help them getting into the forefront of mobile Web app development.

No comments yet - write one!

Despite the fact that a very large part of business is within the lower product categories, smart phones continue to attract the largest media attention. Apple provided no real surprises with the launch of their latest iPhone; they have a product strategy of incremental innovation and improvement, much like the Japanese manufacturers had in the earlier days of digital cellular. Now the race is heating up especially in the enterprise market, following HP's Palm-acquisition and RIM's announcements and rumors regarding their next phone and OS generation. Microsoft's Ballmer openly admitted that they are now clearly behind the competition - but obviously also said that they will be back. In the past, MS's strongest asset in the mobile race has been the excellent integration with office back-end; this will still be needed, but I believe they will also have to excel in some other fields (like usability or third-party app development) in order to regain the lost market share.

A smallish but interesting data bit is the launch of Facebook lite version. Facebook has an agreement with a bunch of operators that when the user stays within the lite version, no network data charges apply. If the user, however, wants to access the full content, she is redirected to the standard version, with normal data rates. The business deal is a very simple partnership, involving no revenue share model. Instead, the operators are expected to attract new users for wireless data (many of the deals are in developing markets) - and get their pay-back when the user wants to get the full Facebook experience. Sounds familiar? The same model we have had in the Internet for many years, getting basic services and content for free, but paying for premium membership. Also, numerous apps in Apple's store and others have a simplified free version, as well as a non-free premium version. I'm sure we will see this same model applied to network charges as well - but including also some innovative revenue share models, in addition to this type of partnership deals.

Image © RIM

No comments yet - write one!

Apple's iPad has started shipping, and all news feeds are filled with articles on the new product. I will not discuss this particular gadget now; instead, I will take a helicopter view on the more generic topic. Tablet concept itself is not a new invention: various manufacturers have come up with the same idea in the past. So far, none of these products has been a success - though Amazon's Kindle has been selling reasonably well, especially for such a specific-purpose product. The technology itself has advanced somewhat, making the screens easier to read, and shrinking the overall size of the tablets. These improvements, however, are not sufficient alone to make a difference. So, providing we trust the market forecasts indicating strong sales in the coming years, what has actually changed?

First, the use case of eBooks is a new one. A traditional paper book has been perfected as a concept since the days of Johannes Gutenberg. Compared to paper books, eBooks still sell in small volumes - but this market is getting large enough to partially fuel the growth of the tablet market. In the same way as music and videos have been driving the small media player ("iPods") market, eBooks and videos will be content formats for the tablets. But I believe a bigger change is yet to come: advertisements. When Steve Jobs launched the iPhone OS 4.0, he showed a preview of the iAd platform. iAd (and other similar systems) permits developers to include ads within their applications, implemented in such a way that they should create a positive user experience. The platform owner takes care of selling the ads, sharing the revenue with the app developer. I can easily believe there will be loads of smart phone apps that use ads to collect revenue - but actually the tablet format is much more suitable for these. In a tablet device with the larger screen, the ads may either take more space and employ stunning visual content; or, they may select the other option and steal only part of the screen real estate, effectively embedding the ad content to the app content. Either way, the ad platforms are likely to accelerate the creation of third-party apps for tablet products, and become a major revenue generator and market catalyst in this product category.

Image © Apple

No comments yet - write one!

Now I had time to update the mobile site of Darteco. The two versions, standard and iPhone-optimized, are combined so that the single URL mobile.darteco.com works for both. Each version uses a unique approach in the implementation: whereas the standard version minimizes the amount of extra code on top of the actual payload (there are just a few lines of JavaScript to steer the operation), the iPhone version utilizes jQTouch and jQuery libraries to achieve all the visual eye-candy and native-like application look-and-feel. In a similar manner, the mobile implementation of Kylansaitti.fi now has a version specifically for iPhone, in addition to the standard version. It is worth mentioning that I haven't performed any wide compatibility testing on the standard versions - therefore, I would highly appreciate any feedback, in case you run into incompatibilities with these Web apps.

Some years ago, the motivation for mobile-optimizing sites was not the same as today. Like in the era of WAP, mobile phones had small screens with limited display resolution. Now, the screens are somewhat bigger, and the display resolution is sufficient for most cases. Also, the amount of data to be transferred is seldom a major problem. However, the usability can be greatly improved by creating mobile-specific versions. Especially with touch-screen operation, standard Web sites tend to have far too small link tags; these can be scaled up by tapping, but the usability still falls behind a site that has full-width buttons, and application logic that is optimized for mobile usage. The increasing demand for mobile versions is also supported by market statistics; According to a study released by a mobile analytics firm Taptu earlier this month, there are 326 000 Web sites optimized for touch mobile usage. This can be compared to 200 000 mobile apps that are currently available from iPhone, Android, Ovi and Blackberry app stores.

No comments yet - write one!

2010 has certainly had a fast start. After a decline in shipping volumes in 2009 - that at the end turned out to be smaller than originally anticipated - this year is forecasted to provide a 5-10% rebound. More interesting than these numbers was the results round by Nokia and Apple: Steve Jobs started his iPad launch by digging into the figures, and proudly announced that Apple is now the biggest player in the mobile arena, bigger that even Nokia. Nokia responded later that according to the standard definition of mobile devices they are bigger. Hmm. I understand the need to be able to say that "we are #1, not #2", but something more interesting than this actual race is behind the numbers. The recent development has shown that you can make awesome profits by either focusing on the mass market (Nokia is very big in the developing markets), or focusing on the absolute top of the smartphone market (Apple's average selling price is very high). Many of the required tools and elements are different, but a high brand recognition is needed in both cases.

Of all the announcements of Barcelona MWC, two have particularly got a wide media coverage. First, Nokia and Intel joined forces to create MeeGo, a mix of Maemo and Moblin. This is a sign of some platform consolidation, and guarantees the position of MeeGo as the other serious global Linux platform (Android is naturally the other one - I don't count some other consortia having a truly global platform). The device categories will cover a wide spectrum from smartphones to tablets and netbooks, similarly to what Google targets with Android and Chrome OS. This is yet one more attempt by Intel to get into the mobile chip market they have largely failed to address in the past; actually, Intel is now in the both Linux camps, as they entered Android with the acquisition of Wind River last year. And, in a typical LG style, we saw today the announcement of the world's first MeeGo phone (GW990) coming from this vendor.

The other interesting announcement was by a group of leading operators, when they launched the "Wholesale Applications Initiative, WAI" (what a name!) The group certainly has the heavyweights supporting it, and a very large actual customer base to address. I have a bit of skepticism knowing that big consortia easily get slowed down by different business, political or geographical interests. But the underlying key promise seems to be in the app architecture and app store concept itself: WAI aims to create a universal support for Web apps (or cloud apps, whichever name you prefer). After a long fight on the mobile OS's and platforms, the operators are now focusing on the Web technologies and browser-based runtime systems. This is a rather obvious choice, and should also enable apps business models that fit well with the operators' billing relationship with the consumer. I hope the operators follow the market trends when setting the revenue share basics - 70/30 seems to be today a reasonable starting point accepted by the developers.

Image © MWC 2010 / GSMA

No comments yet - write one!