Synaptics and AMD today announced that they’re teaming up on a biometric security solution for consumer and business PCs built on AMD platforms. But for Microsoft watchers, the most curious portion of the announcement is that the biometric tech is squarely focused on a mysterious “next-generation operating system” from Microsoft.
[…]
It’s not entirely clear what the biometric security OS is that Synaptics is referring to, as Microsoft itself hasn’t announced any forthcoming releases. However, it could be related to a Microsoft project called Polaris, a more modern version of Windows 10 for desktops that Windows Central senior editor Zac Bowden reported on earlier this year.
Built on an internal project called Windows Core OS, which aims to turn Windows into a modular OS, Polaris is said to focus on desktop, laptop, and 2-in-1 form factors. The goal of Polaris is to provide a shell that Windows users are familiar with, but while leaving behind legacy components in favor of UWP apps. According to our reporting, Polaris would still be able to utilize some form of virtualization to run Win32 programs. However, dropping legacy cruft would, in theory, allow Microsoft to create a more secure version of Windows 10.
That’s basically what I’ve been wanting Microsoft to do for a decade now, so I hope this is actually true. It’d be a big, bold move, but Win32 has run its course, and it needs to be contained and phased out.
Every time Microsoft tries this it fails, miserably. Non x86 versions of windows from the 90s and 2000s, flopped. Windows RT UWP on ARM – flopped. Windows phone – flopped.
Win32 is the crown jewels. Without it, Microsoft doesn’t have a market.
Azure is probably part of the strategy to change this, making everything depend on web apis and SAAS, but the client world still depends on Win32.
Even Office – the other linchpin of the Microsoft ecosystem – is solidly win32. The “UWP” version of Office is just a store friendly wrapper around the Win32 version of office. The true UWP “Office Mobile” apps, and the web versions, are way behind the Win32 versions in terms of functionality, compatibility and usability and unlikely to ever fully catch up.
Finally, UWP is way too locked down to be fully useful. There’s a lot of things you can’t do by design and by policy – the only way to do many things is via Win32 (or something layered atop Win32).
Finally, you simply cannot do anything system level without Win32. Even the versions of Windows without user Win32 access had large chunks of the Win32 subsystem present to make it work. UWP itself is built upon Win32. If they port it to native NT API or to work on a BSD or GNU/Linux system or whatever you’ll have a whole bunch of compatibility issues and problems that will damage Microsoft and the ecosystem even more.
Win32 is not going, and currently cannot go, anywhere.
Edited 2018-07-13 00:50 UTC
(Yes, I said “finally” twice. I was distracted, sue me!)
I disagree. Most users need nothing more complex than a Chromebook. Even AutoCAD is now purely a subscription service. The future will probably consist of very cheap/free basic laptops and rented software, storage and processing power. A bit like the current mobile phone market.
Something being subscription doesn’t mean it can operate in a limited environment like UWP (natively) provides. Reimplementing apps like AutoCAD entirely in UWP would be a massive undertaking and probably not economical for the vendor.
Something still needs to provide a powerful API for more complex tasks, and as it stands Win32 has the existing software library, existing expertise, support and scope to pull it off. Moving away from it entirely is just not going to happen, at least in the next decade or two.
Geeks have the unfortunate habit of extrapolating their very specific needs to the general population. The reality that the vast majority of users need nothing more than a web browser and a very basic word processor. In fact a smartphone is the only computer the majority of people in the world have ever used.
All the things you claim to be totally impractical were the norm until the PC arrived. People used dumb terminals connected to mainframes which performed batch jobs on a timeshare basis. It worked sprisingly well from memory.
MS was primarily a large Unix vendor with thir own version (Xenix) until approached to provide an OS for the IBM PC.
Complex tasks like CAD were originally performed on graphics terminals. Even today profesional video rendering is done remotely on renderfarms.
For Google and MS the ability to run platform agnostic software remotely is a wet dream. For software vendors it prevents piracy and for most users it represents a hassle free experience.
Wasn’t MS, before the PC, primarily a maker of BASICs for home computers?…
MS was, by far, the biggest Unix distributor in the USA. They originally planned to port Xenix to PCs.
https://en.wikipedia.org/wiki/Xenix
Xenix being quite successfull doesn’t have to be at odds with how BASIC/firmwares for home computers were possibly the majority business for MS… (those were millions of units shipped)
Have you looked at the way Microsoft handles licensing? They definitely *dont* want everything delivered like that. they still are betting on the desktop.
Under Nadella MS has started a rapid transformation into a saas company. MS now earns far more from subscriptions and cloud services than it earns from Windows.
https://arstechnica.com/gadgets/2018/03/windows-leader-terry-myerson…
https://www.businessinsider.com.au/microsoft-q3-fy18-earnings-revenu…
And a detailed breakdown kinda proves my point.
The money in “cloud” services is Office 365, Windows Enterprise subscriptions and Azure based hosting of Windows server services (like AD, Exchange, etc).
Almost all of that is Win32
Add to that – until about a year ago it was impossible to license Windows desktop for shared hosting. Now you can, but only if you’re essentially a top-tier hosting company (1-tier CSP partner and the QMTH programme).
RDS as a platform (shared session hosts) is being neglected and new features and software rarely are compatible. Server 2016 build 1709 didn’t even have a desktop component available. Hell, the initial 2019 preview didn’t even include RDS (it was added in later). And on Server, there’s no Edge, no store, none of the new stuff Microsoft makes available.
Microsoft itself is pushing its Citrix based Windows VDI solution.
The “upper tiers” of 365 licensing are including Windows desktop licenses.
Microsoft is still heavily invested in Windows and Win32 on the desktop (or indirectly on the desktop) and in the datacenter.
Its licensing, releases and product range reflect that – it’s still pushing hard for a desktop + server model (just they want to host the servers and perhaps the desktops too).
Microsoft has been quite effective killing Win16 with the release of 64 bit Windows, even if it took 20 years to get there.
Same will happen to Win32. It doesn’t matter if it will take 20 years as well to get rid of it, eventually it will be gone.
Every Windows 10 release is a step into that direction, specially now that they are pushing Win32 applications to use the same container model as UWP ones, as shown at BUILD 2018.
If you were actually paying any attention to BUILD 2018, there are quite a few UWP components that were originally developed by the Office team for Fluent interfaces, as they continue their transition to UWP.
Win32 is going the way of XP, no matter what.
In time, maybe. But
1) It will take a LOT of time, because:
2) Win32 is MUCH more useful and MUCH more widespread than Win16 ever was – and look how long it took to kill Win16. If it is even killed, some might consider it to be on an extreme level of life support still.
3) UWP is STILL not in the same ballpark of useful as Win32. It’s still mostly hype, half-assed “web technologies” and dreams. And the hurdles presented to an UWP dev are still more in number and stupidity than Win32, which has, let’s be honest, not been maintained for a long time since Microsoft started on a quest to develop a successor.
So, yes, Win32 will inevitably die some time in the future. But I have a sneaky suspicion it will not be UWP that is its successor, but Microsoft will have churned through about 3-5 new platforms/APIs by then.
That statement was ridiculous. When something has run its course it is already phased out.
Of course Win32 will disappear someday (as will all of us). Without putting a timeline on such a statement it is meaningless.
I am currently posting this from Chrome which is a Win32 application. 90% of the work I do today will be done in Win32 applications and the same goes for everyone around me. Win32 is used in almost all the business applications, both very tailored custom applications and generic tools like Windows/Office/Chrome.
What is replacing Win32-applications is web-applications. This process has been going on for about 2 decades now and we still haven’t reached the point where I can do most of my work in a browser or even from a Non-Windows device
…and you’re also forgetting another case: Games
Each game is a niche unto itself and you can’t just switch to a similar alternative on the new platform by some competitor.
Even in cases where you will accept substitutes, situations like Tetris, Bejeweled, or Minecraft (where decent-quality clones like Minetest arise) are exceedingly rare.
…a situation exacerbated by how games tend to be “write once, maintain for a short period of time, then move on to whatever is next” projects. Opportunities to port away from Win32, such as HD remakes, are the exception, not the rule.
We should also let Thom know that since Win32 “has run its course” we only expect nostalgia articles about Wine and ReactOS from now on
> Azure is probably part of the strategy to change this, making everything depend on web apis and SAAS, but the client world still depends on Win32.
It’s also a way to get people to pay Microsoft $1200 a year for running a pure FOSS stack on a Linux VM. Microsoft “loves open source” because the more stuff works on Linux on Azure, the more money they make.
I doubt that Windows Core will be that exciting. As I understand it (which of course could be wrong), Microsoft has been working for years to untangle the dependencies from the various parts of Windows that don’t belong together. They’ve succeeded with Windows Server, for the most part.
With Core, most everything will be modular including big chunks of Win32 legacy compatibility code. But it will still be available to be installed as needed.
They’ve already done this. Much was already accomplished by the time Vista was released, and it was basically finished by the time Win7 was released. At that point, they were already demoing MinWin, not as a product, but as a tech demo showing off the modularity that they had achieved.
Sounds very exciting to me. The first move from Microsoft that could move me to use Windows by choice instead of being forced.
The actual authentication solution which will be used it called Trusona.
https://www.trusona.com/product/
Frank Abignaile (the actual person portrayed in the movie Catch Me If You Can by Leonardo Dicaprio/Steven Spielberg) currently works in FBI and describes Mirosoft plans in his recent talk at a “Google Talks” session:
Source: https://www.youtube.com/watch?v=vsMydMDi3rI
Edited 2018-07-13 14:47 UTC
I could excuse the fact that the Trusona website’s design looks a little more “startup hype over substance” than I’d expect from a promising security-sector candidate (but only a little, given what I’ve seen for marketing to other startups), but having it built using Squarespace and having Abignale giving such a ridiculously optimistic time estimate without qualifying it (basically saying that passwords will be history in 2019 without clarifying) has my hype and BS detector pinging.
To be honest, that makes me feel relief, since I don’t own a smartphone by choice, while it’s easy to find inexpensive little USB-attached authenticators for U2F (Universal Second Factor, the more viable candidate that’s in the process of getting implemented by the major browsers).
Edited 2018-07-13 21:18 UTC
And this time, thread priorities work, in this “security” OS?
I think windows users only waited since…
Altair 8800? https://www.youtube.com/watch?v=BQq6ad9F3Xk&t=341s
Not quite that long, but the security question, is addressed..
Microsoft could ofcourse rebase themselves on the legit BSD, and fix this..
Edited 2018-07-15 00:46 UTC
with the lockdown.
Will they finally go all the way with the data slurp for Home and Pro users all in the interests of security you understand?
Just how far will they go?
Personally, I’m past caring and don’t run Windows other than a single XP system. This is never connected to the internet and is used as a controller for my Lathe.
I just see this as an ideal opportunity to really lockdown consumer editions while hiding behind the ‘all in the interests of security’ banner.
And while they are at it, they could make it subscription only. A snip at $9.99/month. Stop paying and it self destructs.