I've had many conversations recently on the topic of application refactoring, and it appears that 2015 is shaping...
up to be a year when both the concept and the vendors will be put to the test.
The concept is rather simple: Take existing applications that are made for large screens with a mouse and a keyboard, and make them work on mobile devices without any new application development.
Some people see app refactoring as the solution we've all been waiting for, whereas others look at it as a stopgap until the next phase of application modernization, whatever that may be. I can see both sides.
Regardless, some users are getting frustrated with the old-fashioned option of using a remote desktop client on their phones. They might take matters into their own hands and start using other cloud services outside IT's control. That is not a good solution.
Has the market for application refactoring reached a boiling point? Not necessarily, but that could very well come soon, so it's good that there are several companies contributing answers to these problems. They are all relatively young, with good (albeit different) approaches, but not a lot of customers. You have to start somewhere, though, and I think we'll see app refactoring technology blossom and find a home in 2015.
What follows is a rundown of the three main players in application refactoring: Reddo Mobility, PowWow and Capriza. They all have products that are aimed at just about any vertical and require no access to an application's source code. Each has a different approach, and each focuses on refactoring a specific type of application.
Reddo Mobility was born from a company called Gizmox, which created a plugin for Microsoft Visual Studio to add HTML5 capabilities to traditional Win32 applications. Gizmox split into two companies after it realized that it could use the same approach for application refactoring, and Reddo Mobility was formed.
Reddo works by inserting itself between the operating system and presentation layers of Windows. It learns what the app tells the OS to draw on the screen and vice versa. The information gleaned from this approach is then disassembled into individual objects that an IT administrator -- or even a power user -- can reorganize to be more mobile-friendly.
Reddo's application designer shows you the Windows interface, as well as the mobile interface. From there, you can drag and drop different screen elements to the mobile device. The designer is also device-aware, so you can create different configurations and workflows for tablets and smartphones. You can even skin the app to look native.
End users see an application, and when they launch it, what looks like a native app is really an HTML5 rendering of what is happening on a computer in the data center (or even at their desk).
The approach of capturing control information between the presentation layer and the OS means that the typical use case for Reddo is geared toward Win32, data-driven applications. It can optimize window controls, data fields, buttons, and so on, but has a harder time with graphics. Graphics work with Reddo's product, but they're simply sent as they are rendered using a technique known as pixel-pushing. It does not work for Web apps.
PowWow was one of the darlings of BriForum 2014 because it was the first time many people had ever seen app refactoring. PowWow's approach is also immediately identifiable to anyone in the remote desktop community because it detects what is happening on the backend by establishing a Remote Desktop Protocol (RDP) connection and reading what is sent to the remote display.
PowWow does not use RDP for end users. It establishes an RDP session from its own servers to an RDP server or VDI desktop. When the end user's device talks to PowWow, they're communicating via HTML5. Because PowWow uses this approach, it can offer a collaboration feature that lets multiple users connect to the same HTML5 session. Picture it just like using Google Docs; workers can interact and collaborate, or the session can be view only.
This dependence on RDP is both good and bad, because PowWow uses RDP protocol information to identify and transform screen elements. There could possibly be some RDP weirdness -- such as compression, blocking, network issues and more -- that get in the way of the transformation. It's not likely, but it's also not unthinkable. On the flip side, it works with any application. PowWow's transformation is actually a two-pronged approach, with a server-side component that hooks into the OS to generate metadata about what the app is trying to do and how it should be interacted with. Then it merges that data with the transformed screen information that comes from scrubbing RDP. In this way, PowWow can deliver full Windows interfaces with mobile controls or transform everything into a mobile interface.*
Like Reddo, PowWow has a designer that lets you modify elements from the native Windows app to make them more mobile-friendly. It preserves more of the Windows feel of an application because it's easier to just pass stuff through, but it has the ability to make buttons larger and move screen elements around, or to different tabs.
Capriza is the old man in this sea, having been around for two years. I'll give it credit for being the first to market, however its focus is solely on Web applications.
The product works by using a headless virtual browser -- either in the cloud or on-premises -- to connect to Web applications that you can deliver from on-premises servers or from software as a service vendors. The transformation happens at the virtual browser and is then passed to the Capriza cloud service. That service handles connections to the mobile devices, which is via HTML5.
Capriza decided to remain in the Web app niche for the time being because it has seen quite a bit of success with refactoring these types of applications. For platforms that the company encounters on a regular basis, such as SAP, it added specific optimizations to the platform, though I'm told that 95% of what they do is generic and works on all Web apps. Capriza even has optimizations in place for modern apps, such as Salesforce.com. Though Salesforce has an app, Capriza has customers who want to streamline the interface and the processes to make it more appropriate for a mobile user, leaving the more complex workflows to the full app interface on a desktop or laptop.
Capriza's designer is very similar to Reddo's and PowWow's. The Web app and mobile template are in front of you, and you can drag, drop, and tweak until you get the mobile interface the way you want it.
Three companies, three approaches, three use-cases. We'll have to wait and see which one catches on. Though Web apps are important, I suspect that application refactoring will be very popular amongst Windows users as way of modernizing how users interact with their applications and data. Whether or not these tools are used permanently or as a stopgap until something else is assembled remains to be seen, and is largely up to how fast the technology matures. I suppose by this time next year we'll have a much different view of the landscape.
*Paragraph was edited after publication.