Comment by lenkite
1 day ago
The problem is that they just could not commit to anything for more than 2 years after Win32.
They had something reasonably good in WinRT. They should have stuck to that. But Nadella came in, said Azure Cloud is the future and abandoned the Windows platform.
At this point one must ask if Microsoft is still a software platform company - whether their products form a substrate where an ecosystem can form and build a coherent software environment for the users of their platform.
Microsoft used to be the Windows company (after being the BASIC company, then the DOS company). Then it became the Office company. Now it’s SharePoint and Office365 and Azure, a utility. Windows is a relatively small part. Office is both desktop and web (and spacecraft, where they have two versions of Outlook and none of them works). If you are confused at this point, so am I. There is no vision as to what Microsoft is. If Satya Nadella knows what Microsoft is, he isn’t communicating it properly. It’s not Azure, because there is also Office and Windows. And on-prem server products. And a line of hardware products. And stores (do they still exist?).
MS has never been a software platform company. That's the fundamental reason behind the issue the article talks about.
MS has always been a software application company. Windows was never anything more than a way to sell MS applications--and Windows 3.0 and later wouldn't even have existed in the first place if IBM hadn't dawdled so long over OS/2. Even in the MS-DOS days, when MS was reaping the benefits of IBM's previous bonehead decision to hand the PC OS market to them, MS was selling Office applications--on the Macintosh.
The basic Windows API, in all of its many incarnations, has always been a second-class citizen; MS Office applications have always done their own things that other Windows applications couldn't do without using undocumented features that MS could change at any time (and often did). One could argue that the only reason MS even allowed third-party Windows developers to exist was so that they would, in the words of one of PG's essays, do market research for MS. When a third-party dev came up with something that got enough traction, MS would simply incorporate it into their apps.
Microsoft was a language company at the start - they had a huge share in 8-bit computers and their BASIC made into the ROMs of almost every computer sold in the 70s and 80s. Then they branched out to applications, with little success (I remember Multiplan on CP/M, DOS, and Mac). When they started selling PC-DOS and MS-DOS they had no applications play to speak of. Office only came much later, and the apps that appeared for Mac, Word and Excel, were ported to Windows starting on Windows 2. Word for DOS struggled in the market and never reached a significant share.
1 reply →
How do you reconcile this with their history of bending over backwards to achieve backwards compatibility for third parties?
1 reply →
This makes sense, because even in the best times Windows was not the biggest money maker for Microsoft, it was Office. So MS was never fully behind Windows, it was only the means to an end, which was selling the most software for enterprises.
1 reply →
Microsoft has always had a broad vision of itself as a technology company; I feel it's perfectly fine to not be able to describe Microsoft in one sentence without using platitudes like "empower every person on Earth to achieve more" or "put a computer in every home and every office" (both paraphrases of actual MSFT company mission statements), and I suspect many other current and former Microsoft employees would feel the same way.
IMO Microsoft's best long-lived products have always been both finished solutions to your problems and platforms to help you develop more solutions, and Microsoft leadership has always recognized this. Examples: Windows. Office. Dynamics (their Salesforce competitor).
But even if a product doesn't meet that "why not both?" ideal, there is always going to be room for it at Microsoft, as long as it is not only a good or at least mediocre product by itself, but also works to sell you on the whole Microsoft ecosystem. Sometimes that is a bad thing (see all the Windows adware for Bing, Copilot, and M365). But that at least is where Microsoft remains consistent.
> "put a computer in every home and every office"
That was such an amazing mission statement. It was a real measurable goal, and progress towards it was quantifiable. And Microsoft actually did it! That mission statement drove actual strategies (lower costs, don't complete with Apple on the high end, force OEMs to compete against each other on price, etc) that resulted in its ultimate fulfillment.
Nadella thought he could take the reins and start yelling “Cloud! Cloud! Cloud!” and that would be successful. He doesn’t have a strategy and now that’s becoming apparent.
He had a strategy and it worked very well. But every strategy must be updated. It's basic BCG matrix stuff every MBA graduate knows by heart: sooner or later your star product becomes a cash cow and then a dog. To keep your company growing, you need to identify your next potential stars among question marks, boost them with cash from the cash cows, put competent managers in charge and remove those who get in the way.
Gates did this with Windows, Office, XBox among other things. Ballmer failed to do this this with Windows Phone. Nadella did it with Azure, but he needs to do it once again with AI. You can see that he's pushing hard with Copilots everywhere, what's missing is a manager that has a coherent vision of what AI at MS should look like. ScottGu is in charge of both Azure and AI at MS, but I don't know if he can deliver.
1 reply →
Cloud is now massively larger than any other part of Microsoft. It's why he became the CEO in the first place.
He maybe never had a strategy for Windows but he wasn't hired to have a strategy for Windows.
Nadella took the reins in 2014 and the stock has 10x’d since then. In the same timeframe, the sp500 has 2.5x’d. Sounds pretty successful to me?
11 replies →
WinRT was technologically terrible (which immediately flows from "no one at Microsoft was actually using it to make anything useful"). But that wasn't even what sunk it - the whole requirements around "of course your WinRT app is going to be in the Microsoft Store^TM its the future" did that. The fucking store is a joke, and those requirements existed solely to boost a bunch of idiots internal careers.
The Windows Store thing was so terrible that I would argue the only good thing that came out of it was that it made Valve/Steam invest in Linux.
I still don't understand why the windows store search sucked so badly. It isn't like they had billions of apps. So why did it suck?
Most probably it was on purpose. MS is famous for the infighting of internal groups and how the management doesn't know how to control their divisions.
Correct for the Windows 8 and 8.1 initial versions, it was already quite good with UWP on Windows 10.
But then they rioted internally to kill C++/CX (the only time they had something comparable to C++ Builder), Project Reunion got announced and misused from the original goal, porting WinRT back into Win32 killed .NET Native as well, most of the key team members left to Amazon and Google, Azure or AI teams, the team is now mostly interns or juniors from Microsoft India, no direction, and is a mess, naturally.
I went from a WinRT advocate, to pointing out devs to stay away from it, this is how bad they treated those that actually believed WinRT could be it.
Part of the reason was that WinRT was aimed at tablets and other low power devices, where unrestricted Win32 API use could drain the battery too quickly. They were trying to put the genie back in the bottle to have more control in this new ecosystem, like Apple enjoyed on iPad, etc. Much of the weird signing and app store evaluation was to make sure your app used only the blessed APIs and wouldn't, for example, stay awake in the background listening on an arbitrary port and drain the little tablet battery.
Painful and nonsensical from a desktop standpoint but also kind of impressive in a way.
My favorite example of that was when WinRT app .exe files could not be launched from the command line. Only via some Windows Store voodoo dance with approvals, signatures and "security" that made WinRT for developers essentially a dead-on-arrival technology.
I would not be surprised if you still cannot launch a fricking .exe.
This was sorted out in Windows 10 with the unification of WinRT programming models under UWP.
1 reply →