So I’ve been thinking about the announcement of Apple Silicon a couple of weeks go along with reaching the commentary of those who have a good reputation within the information technology sphere about the future direction of companies. The one thing I overlooked was how this all fits in with Microsoft and their own future direction given that the future of Microsoft isn’t wedded to the success or failure of on the desktop – it’s all about the cloud and what devices that do exist are going to be light weight with the heavy lifting being done in the cloud.
Windows on ARM has been an abysmal failure but there is an opportunity for Microsoft to either do one of two things (relating to software) running on top of ARM. One option would be to create a rebranded version of ChromeOS in the form of ‘Edge OS’ which is heavily integrated into Microsoft’s own cloud service in much the same way ChromeOS is integrated into Google’s own cloud services (bundle it with low cost ARM based devices – get the device for cheap, make the money back on subscriptions). The alternative is for Microsoft to take Windows 10 and strip all the legacy code out of the system so essentially it is the NT Kernel, UWP, WinUI and then the standard bundled applications (gradually move Windows 10 to Edge OS by shipping Windows 10 for another 2 years then make Edge OS the default with 3 years support for those with Window 10 then eventually end all support with maybe a longer transition, say +5 years to end enterprise support) . Removing legacy support would free up Microsoft to be able to optimise the operating system without consideration for backwards compatibility along with making its operating system cheaper to maintain over the long run due to a smaller code base which is based on modern code (maybe jettison their own C/C++ Library along with compilers in favour of adopting LLVM/Clang (plus adopting libc and libc++), move from PE to ELF (if you’re going to brea compatibility you might as well go all the way) and make LLVM/Clang the details since the value of Microsoft’s developer tools is in the IDE not the compiler (maybe .NET utilising LLVM).
When it comes to the cloud side of the equation – end of the day as long as you’re using their cloud service I don’t think Microsoft actually cares whether it is being accessed from a Windows PC, Mac, iOS, iPadOS, ChromeOS, Android or any other platform (although controlling the underlying platform does give it a lot more leverage to make the experience better for users in terms of integration). Given how Microsoft is opening more data centres around the world with the recent (link) opening in New Zealand, I could imagine in 5-10 years time Microsoft will no longer sell on the premises software and everything will be delivered via the cloud with the barrier of ‘the illusion of control’ being overcome with a new generation of information technologies managers only ever knowing cloud computing thus seeing on the premises data centre being another legacy that can be gotten rid of.
It’s going to be a big culture change – Westpac has already started embracing the private cloud and seeing benefits of lower running costs and being able to get application upgrades to customers quicker. The recent move by Westpac NZ from Hogan to Celeriti along with a whole new internet banking website along with new mobile banking apps (having used it – the experience is pretty good from a customers perspective). I think at this point banks are asking themselves whether it makes sense to be so heavily invested into technology when in reality it a means to an end rather than an end in and of itself. The idea of moving to the cloud is to bring the focus back on what you’re doing as a business – the cloud is there to serve a purpose, to enable you conduct business rather than the focus of the business being on technology.