Microsoft hardware and software event.

Microsoft has recently held a big hardware event and it is clear that the claims by the talking heads in the technology media about Microsoft hardware division is only an inspiration to third parties is well and truly wrong. Microsoft is in the hardware and they’re in there to make money by developing the sorts of products that get people excited and inspired, they aren’t there to compete against the bargain basement products put out by high volume OEMs – they’re setting themselves up as the Apple of the Windows world then leaving the sub NZ$1000 market to Dell, Lenovo and HP to fight over.

What got my really excited were the Surface Duo and Surface Neo for different reasons. The Surface Duo being a Android based device could finally be the sort of device that, when coupled with Microsoft’s own support policy, will mean that upgrades and updates will be pushed out with the same sense of urgency that Microsoft do with Windows which would be a huge change in the Android world. At the moment you’ve got the likes of Samsung sometimes being 2-3 months behind Google’s security updates where security updates for their latest phones (at least in New Zealand) don’t come out until the end of the very month they were released by Google to its own phones.

The Surface Neo is interesting not necessarily because of the hardware, although the hardware is awesome, but rather because of the operating system that it uses which Microsoft labels Windows 10X. The attraction to me isn’t just the new shell that is present but also the fact that there is win32 support but it uses container technology which, in my book, is something that is desperately needed given the number of third party software titles with poorly written installer can hose an otherwise pristine Windows installation into a mess – the same reason why I got excited at the idea of win32 applications being in the store due to Project Centennial putting that line of demarcation between the operating system and the applications that are installed so that a badly written installer cannot hose Windows or when you uninstall a particular software title there aren’t hundreds of bits and pieces left over because the uninstaller is in worse shape than the installer.

Keeping in mind that all of this won’t be shipping until next year in December 2020 so it leaves me wondering whether part of this will involve the shell being very much built upon the work being done with WinUI 3.x and 4.x so I could imagine Windows 10X being made available to OEMs but long term I don’t see Microsoft wanting to support multiple branches. Long term I could see Windows 10X becoming mainstream then eventually being pushed out to enterprise customers but long story short the future is a Windows NT Core with two main frameworks, .NET and Unified Windows Platform, with legacy win32 libraries hanging around for backwards compatibility and those applications that need such backwards compatibility will be running inside a container. If Microsoft can nail this all down then the opportunity for Windows and Microsoft beyond the PC looks bright.

Personal · Technology

Getting life back on track.

One of the big focuses this year is getting my financial house in order and that involves paying down the debt I’ve accumulated. What I’ve done to knock some of the debt away is selling off my Apple Watch and that is because I had to be honest, I never really used it and it was money tied up in something I was never using that would be better spent reducing my debt that just sitting on the recharger. At some point you have to ask yourself whether the money is better off being used for something useful.

There has been a recent rumour come out that Samsung will be moving away from its own ‘in house’ customer CPU architecture in favour of utilising ARM designs (link). It is a rumour at this point but a lot of noise was made regarding this but one thing that needs to be kept in mind – for Samsung to have a profitable and thriving business they needed to be able to ship in volume enough units to make further investment into their own core design pay for itself and given the lack of third party enthusiasm I’d say at this point Samsung had to really ask whether sinking more money into it makes any sense.

Now, there was a point being made about it’s performance against Apple’s latest A generation of SoCs but to be honest I don’t see too many end users checking out the specifications – whether one phone loads something 2 seconds faster than another makes all that much of a difference in the grand scheme of things. I’d also argue that moving to a less bespoke design allows them to leverage the contributions that ARM have made to Clang/LLVM which should translate to improved optimisation and a better experience. The licensing of AMD GPU designs by Samsung also would allow them to close that gap so essentially you’ll have an SoC made of a big.LITTLE design straight from ARM, a GPU from AMD and make use of its own modem which does a pretty good job when compared to the other modems out there.