I’ve been covering PCs since the birth of the market back in the early 1980s. What I think is fascinating, given that IBM largely drove the early adoption of PCs into the enterprise, is that the company never grasped that the market didn’t want the PC at the time. The idea of everyone having their computer wasn’t the driver; it was that everyone wanted more control over their computing and, at the time, if you needed an application, you had to put in a request to MIS (which became IT) and then wait months, if not years, to get something that had little to do with what you’d requested. Our users were not happy at all.
The Long Wait till Windows 365
IBM responded to the threat Apple represented not by opening up their mainframes much like the cloud is today, but by creating an alternative to the Apple PC, the IBM PC. This error was a repetitive mistake that has plagued many companies. Apple almost went under trying to copy the IBM PC in the 1990s. Microsoft tried to copy the iPod with the Zune when they should have created the iPhone. Research in Motion tried to copy the iPhone, as did Palm. Palm failed, and BlackBerry almost did. Filling an unmet need when bringing out a competing product works; trying to copy that product badly doesn’t.
Both Oracle and Sun seemed to grasp this in the 1990s with their Thin Client effort, but they didn’t have the I/O of a mainframe, they didn’t embrace the then-dominant desktop platform Windows, and they tried to sell their solution to IT, and not the users who drove the PC effort in the first place.
We continued to have a Thin Client market. Still, it was crippled by the lack of high-performance wireless data connections (no 5G until recently), the inability of existing server architectures to handle the needed I/O and a massive number of sessions, and a distinct lack of user focus. Thin clients sold but mostly in areas and markets where users didn’t get a vote.
With the cloud, servers got the necessary upgrades to handle large numbers of users, provisioning advanced by leaps and bounds, and 4G LTE is being replaced with 5G finally, decades after the need was identified, supplying the right mix of hardware to create a mainframe-like product with a user focus. And thus, Windows 365 was born. The irony being this would likely be a perfect product for IBM to resell, but IBM finds itself partnered with Apple, the company that took them down the wrong path and cost them their dominant market position.
Why Windows 365 is Potentially a Perfect Storm Product
A product’s success is often predicated on the foundation set and how well it addresses an unmet need. The color TV didn’t sell until Disney’s Wonderful World of Color established a need around it a decade after it was created. Apple PCs didn’t sell at high volumes until the Mac addressed a missed need for more appliance-like products that ran productivity software. The cell phone needed to become affordable. The smartphone needed to embrace the two-way pager and then the iPod initially. The laptop PC needed to be light enough to be portable and wirelessly connected for success.
The Thin Client needed to fully embrace the PC while providing the cost, security, manageability, and reliability advantages of a terminal. Where this market fails is when vendors treat the list of requirements like multiple choice rather than all-of-the-above. Windows 365 on paper tied back into the Azure Cloud is the first offering of its type to approach (we won’t know for sure until it is in use at scale) this all-of-the-above set of requirements.
Security Could Improve if the Price is Right
While several elements are yet to be disclosed, like a cost that isn’t insignificant, and we don’t yet have deployments at enterprise scale to test loading, Windows 365 represents the closest thing to what we wanted back in the 1980s when the PC was created. Cloud technology allows it to scale; Microsoft owns the solution and can better assure its security, reliability, and availability. We have enough wireless bandwidth to cover mobile users, and it should be far easier to secure against the Nation-State level of malware and security threats we are now facing. If you need to, you could power off a cloud service, physically halting a ransomware attack and digitally isolate the infected nodes without involving the users, for instance. And you can far better assure compliance with patch updates if those patches are applied centrally and not locally.
We still have to sort hardware because, for a cloud-hosted desktop, you don’t need an entire OS or a potent processor or GPU, as those are provided in the cloud, but you do still need a good display and set of human interface accessories. PCs could drift toward Thin Client architectures, optimizing on bandwidth rather than performance, or smartphones could gain the ports or wireless accessory capabilities (for multiple monitors, for instance) needed for this solution. That’s coming, but we are finally back on the path we abandoned in the 1980s with the potential blend of the best aspects of terminals and PCs into Windows 365.