Relax, IT: Endpoint diversity is nothing to fear

Being able to choose the tools you prefer is great for users, but it drives many in IT nuts. After all, diversity means more work to manage. There’s a reason IT has driven companies to highly standardized monocultures, and it’s not based on some character defect.

Old IT hands remember the nascent days of widespread business computing — the early to mid-1990s — when every department had its own computers and software, each different than the rest. People bought “best of breed” tools that ddn’t work well together. That was OK at first, before corporate networking, much less the Internet, took off, and sneakernet — sharing information via paper memos and in meeting presentations — was the communications channel for most. As soon as real networks and the Internet became common, it became painfully clear how siloed businesses were, how incompatible data and processes were, and how much labor was involved in making the work products and technologies compatible across the systems.

“Best of breed” and “departmental computing” became dirty terms, and IT and business leaders went about transforming both their technologies and work processes into integrated, standardized, homogenized approaches. That helped businesses take advantage of the Internet and tap into what is now a global supply chain of goods, services, ideas — and customers. Ever since, IT has guarded against a return to that chaos of incompatibility and inconsistency.

Now IT is told to let it happen all over again. Or at least that’s how the consumerization phenomenon appears: Suddenly, it’s OK to use different mobile devices with different capabilities, different levels of management, different levels of security, and different types of expertise to support. It’s OK to bring in Macs into the pure Windows environment; never mind that the standard management tools often can’t manage them, and IT now has to be proficient in yet another technology. And those corporate ERP, Office, and other apps — they’re inferior or nonexistent on non-Windows devices. How do you expect a company to work efficiently, smoothly, and cost-effectively in such a sea of differences?

In many cases, IT is overreacting. SAP has plenty of mobile and Web apps for non-Windows platforms, and Microsoft is doing the same for its CRM software. There are plenty of common protocols that bridge many of the differences. Microsoft’s Exchange ActiveSync (EAS) for managing email, calendar, contacts, and device access for PCs, Macs, and mobile devices is one example. Another are the common file formats used by knowledge workers — .doc, .xls, .ppt, .zip, .pdf, .txt, and .html — that Windows, Macs, and most mobile devices support, even if with different tools.

As I’ve explained previously, IT has to move some of the support burden onto the users who choose their own technology; that’s only fair. However, it’s Microsoft’s fault that Office for Mac is a crippled version of what Windows users get and that mobile and cloud Office are even more hobbled, to the point of unusability in many cases. Buyers must demand for that business decision to be changed.

But IT still has to deal with certain changes, and in some cases they’ll only get worse as Apple, Google, and Microsoft — the three companies that matter the most in driving consumerization’s technology directions — favor their proprietary environments in other services, apps, and protocols.

On the other hand, such diversity is causing other vendors to address the problem. Consider the case of mobile management. Research in Motion’s successful BlackBerry Enterprise Server (BES) set the stage for that function being unrelated to managing PCs. When Apple’s iOS became popular, the moble device management (MDM) tools that arose likewise were independent of PCs — mainly because IT saw MDM as a facet of email administration (based on what BlackBerrys did) rather than as a facet of PC management. To be fair to IT, the notion that an iPad should be considered a type of computer has only recently gained broad, though not universal, acceptance.

But MDM is now moving out of its email ghetto into the big house of system administration. Microsoft’s System Center 2012, to be released later this year, now lets IT manage EAS policies from within; before, those policies were accessible only in Exchange. It’s a small step, perhaps, but in the right direction.

Third-party modules from companies such as Odyssey Software and Quest Software also bring Mac, Linux, and mobile capabilities to System Center’s Configuration Manager so that IT can manage a diversity of devices from one console and one set of underlying Active Directory policies in a user-centric way, not specific to devices. MobileIron has APIs to connect its MDM platform to System Center as well and may bring some connectors out this year. In addition, Symantec’s Altiris and Centrify’s Centrify Suite system management tools support policy-based management beyond Windows PCs, covering two or more of the following: Macs, Linux systems, Unix systems, and iOS and Android mobile devices (including apps and content to some extent).

At the same time, you’re seeing Good Technology and MobileIron rethink MDM more broadly as a user-centric management platform for modern OSes (Windows 7 and 8, Mac OS X, iOS, Android, and perhaps one day Windows Phone) with, respectively, their Dynamics and AppConnect efforts. Those MDM companies argue, with reason, that traditional systems management tools’ longtime focus on patch management and inventory control and optimization for legacy platforms such as Windows XP, Windows Mobile, and Symbian requires a significant rethinking and reworking. For both, the notion of MDM being about devices is already archaic; they’re thinking about mobile users, then the devices, data, and apps they use.

It doesn’t matter much whether systems management grows out of the Windows monoculture into a diverse world or MDM subsumes PCs in its mix. What matters is the shift to a user-based common management of diverse devices.

The heterogeneity of consumerization is also forcing out old, proprietary technologies such as Microsoft’s ActiveX that in recent years has worsened IT’s life by forcing it to support obsolete, insecure browsers such as Internet Explorer 6 because they had bought or created Web apps dependent on it. These apps, in turn, were hardwired not only to ActiveX but to IE6’s version of it (or IE7’s or IE8’s, depending on the app), causing a management and support nightmare for IT with no strategic benefit.

One IT pro told me that his group has been trying for years to get rid of such legacy ActiveX and ODBC apps, but business management didn’t see the point of spending money on completed projects — until they discovered their iPads didn’t work with these apps. Suddenly, business managers found the money to modernize those apps, giving that IT group the chance to move to more vendor-neutral and broadly supported technologies such as AJAX.

I suggest that IT actively look for such compatibility issues brought on by adoption of Android, iOS, Mac OS X, and so on — and use them to justify a move to more modern technologies and standards. Of course, trading one dependency for another is a foolish exercise, so the substitution has to move up the abstraction ladder. That means broad standards where available, with as few OS and hardware dependencies as possible (or broadly available clients to handle the various endpoints’ peculiarities).

Consumerization can be a catalyst for IT to get rid of the legacies that bedevil it, as well as the unnecessary silos that have grown over time. That should create space for the value-added aspects of consumerization’s diversity of apps, OSes, and devices, and even reduce the effort spent on the endpoint and low-level activities.

It should also make heterogeneity as positive a concept for IT as diversity has proven to be to be in personnel management.