▲ 835 points
back
580 comments
What I find must puzzling is that everyone seems to just be violating basic rules that had been in place for ages.
Things like:
- If you can't respond to a UI event wait until you can
- Menus should be tree structures
- Pressing alt should underline the hotkeys you need to access anything clickable
As well as just basic responsiveness or predictability. A 2000 era windows application may not have been pretty, and may well have several different styles all imitated from office, but at least I knew what everything did and when it was slow at least it did what I expected.
This meant I could start the computer, log in, potentially start and use several applications and only then turn on the screen. Nowadays that has no chance of working because even to log in I need to press enter or click some button (which one depends on how I logged in previously, maybe) before I can even start typing and doing so eats a random amount of keystrokes while the damn log in screen loads to do its one damn job.
It's because we went from the Desktop environment, where rules were well-documented and standardized, to the Web/Mobile environment, where rules had to be reinvented and, for the most part, were not.
We've lost design idioms, which is a huge tax on users everywhere. I've been mad about this for years: https://essays.johnloeber.com/p/4-bring-back-idiomatic-desig...
WinForms - It just works.
Any system that needs a straightforward UI for kicking things off, stopping them, logging them, and dragging data files into them..... WinForms.
Bugfree, hardened by the test of time, works on Windows X, Y and Z.
Everything else is just consumer silver sprinkles, and involves faffing around with multiple config files and obscure layout issues.
> basic rules that had been in place for ages. Things like: [...]
I am going to add my favorite here, just to rant into the void. A dialog box's options must never be Ok/Cancel. These are not the same sorts of things. "Cancel" is a verb, "Ok" is a noun (in this context). Even if "Ok" is taken to mean the verb "acknowledge", it is still not an alternative to cancelling.
99% of these dialogs should be "[Verb]/Cancel": Change "Ok" to a verb or short phrase indicating the action that will be taken if you press it. Don't do the action if the user hits "cancel". The verb should be something specific like "Delete file" or "Terminate process" and not something useless like "proceed".
- clicking a checkbox's label should select/deselect the checkbox
This is one that I hold my devs accountable for. No, I shouldn't have "put it in the spec", because it is the fucking spec.
Some Microsoft managers started suppressing people who reported internal defects, instead of leveling their teams up
We've ended up in a world where power users have been forgotten. Not out of malice, but out of a misguided aim to reduce complexity and achieve consistency with the web.
I would argue that desktop is the platform for power users, and its future depends on them. The keyboard shortcuts, the micro-interactions, the window management -- this stuff is all important when you're using a system for 8+ hours per day.
Yet we risk desktop experiences becoming less useful due to the UI becoming "dumber" as we keep shoehorning websites onto the desktop. Website UI is dumb. It's mouse driven, keyboard is an afterthought. There's no consistency, and you have to re-invent the wheel every time to get the details right (almost never happens).
Well, Alt+Tab in Windows is supposed to switch windows. That's unless you're in Microsoft Edge where obviously, it switches tabs. Inconsistent and annoying.
Because everything got made for smartphones and touchscreens while desktop users are left to suffer in the mediocrity.
> This meant I could start the computer, log in, potentially start and use several applications and only then turn on the screen.
I mean... well... responsiveness matters to me too, and I am impressed by such inspired productivity, but... I'm also confused. Why not turn on the screen - the monitor, right?
Now thinking about how gui lag might impact the sight-impaired, tangential as that is...
> WPF was good
As someone who saw what impact WPF had on average users running average hardware in the late 2000s to early 2010s, I disagree.
In 2011, my brother was in seminary, using an average Windows Vista-era laptop that he had been given in 2008. When he was home for Christmas in 2011, we were talking about his laptop, and he told me that the Logos Bible software ran sluggishly on that laptop. He said something about how, for reasons unknown to him, the current version of Logos required advanced graphics capabilities (I forget exactly how he phrased it, but he had learned that the slowness had something to do with graphics). Bear in mind, this is software that basically just displays text, presumably with some editing for adding notes and such. At the time, I just bought him another laptop.
A few years later, I happened to read that Logos version 4 was built on WPF. Then, remembering my brother, I found this Logos forum thread:
https://community.logos.com/discussion/6200
This shows that Logos users were discussing the performance of Logos on machines with different graphics hardware. For a program that was all about displaying and editing text, it shouldn't have mattered. WPF had made a bet on then-advanced graphics hardware for reasonable performance, and that was bad for these users. And that's just the one example I know about.
I remember when ChatGPT exploded and Bing had it integrated, the idea was brilliant because unlike ChatGPT it didn't have information cut-off since it can access the web. I was very excited to ditch Google for AI chat with web access.
How did MS actually implemented it though? After a few messages the chat is blocked because MS did not choose to walk the extra mile and maybe compact the context so that their product can be actually usable.
Of course OpenAI, Perplexity and others later implemented that properly and its integral part of modern AI chat and I actually ditched Google for the most part. Had Microsoft done it, they might have had a shot in replacing Google and maybe becoming the AI Chat provider. But no, Microsoft can't have a well thought UI to provide a delightful UX.
IMHO it's a culture thing. Lack of cohesion is a result of it, I used to be annoyed by Apple that doesn't allow to ship its own UI libraries together with the app so to support old versions etc. but Apple had it right, thanks to the limitations UI is coherent.
The deeper problem is that Microsoft keeps trying to solve GUI consistency at the framework layer instead of the design system layer. WinForms, WPF, UWP, WinUI -- each one a new framework, each one eventually abandoned.
Apple solved this by treating the design system as the product and letting the framework be invisible. Microsoft has it backwards every time.
2 things Microsoft failed to do in the last 15 years are:
1) They abandoned their mobile phone, tablet, and wearable strategy. So, today if you develop a native Windows application, it will only work on desktops and laptops. That is it. It is not attractive for a developer to learn a whole new UI framework just to target a single form factor. And I don't know if there is any solution for this at this point, they shouldn't have completely abandoned those markets.
2) They did not back 1 UI framework for a long time (I mean 10 years+), instead they did significant changes to their UI framework strategy every 3-4 years. It takes a huge time for developers to trust, learn and develop complex and polished apps in a UI framework. Also it takes a long time for a UI framework to become mature. If you change your UI strategy every few years, you will never have complex and polished apps written with it.
To be honest I am not sure if Windows will ever be able to recover in the long term and keep its market share. The only reason it seems to be alive is because enterprise runs on Windows and it is hard to change that.
I feel like an Apple + Google dominance will be more likely in the long term for desktop operating systems. I am not sure if Google will be able to avoid the first mistake I wrote above but they are working on bringing Android to desktop. It is a good idea but it requires at least 10 years of supporting and polishing it despite not getting much traction. But if Google persists, we might be all using MacOS and Android on desktop 20 years from now.
The churn would have been much worse if Microsoft was rolling out successful GUI framework after GUI framework. As it is you can still write a Win32 app if that pleases you, or still write .NET (and damn that runtime download!)
Microsoft has bought into ‘make a web app’ since 1988, they introduced AJAX, they got flexbox and grid into CSS and numerous HTML 5 features to support application UIs. They ‘frikin bought npm!. I use Windows every day but I almost exclusively develop cross-platform systems based on the WWW, Java, Python, etc. Whenever I have developed with .NET it has been for a cross-platform front-end like Silverlight or Unity/itch.io.
I can’t say I have a desire to make a native Windows GUI app when I could make a web app: like if it worth doing from my computer isn’t it worth doing it on my iPad from anywhere with Tailscale? For all the complaints about modern JavaScript it gives you the pieces to make a very pleasant world in terms of DX and UX and you certainly don’t need to ship an Electron runtime for many applications.
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.
Desktop, especially Windows, is such a mess.
It's 2026. We're running 8+ cores and 32gb ram as standard. We can run super realistic video games at high frame-rates.
Yet on the same machine, resizing a window with rectangles in it is laggy on every platform except macOS (where there's a chance it's not laggy).
Steven Sinofsky wrote this piece a couple of weeks ago about the same topic:
Having spent some time kicking around the Delphi space I got quite into WPF in 2007ish. By 2010 I had not just sworn off it, I'd sworn off Windows entirely. The constant stream of rug-pulls as one bit of MS managed to pull off a political heist over another and - oh no - yet another "latest and greatest" technology was effectively deprecated within 18 months of launch, invalidating all the effort you put in to staying up to date just became a pointless treadmill.
Fortunately Rails was taking off at that point so it was fairly easy to change horses and just ignore it.
Java wanted to run your code everywhere so they basically wrote a byte code compiler and then wrote one vm per OS/architecture
I never understood.NET's purpose. What problem it exactly went out to solve? Did Microsoft want developers to be able to run their applications everywhere too? Absolutely not.
Sidenote - MFC is the ugliest thing you'll see. Yet they didn't mention another piece of work called ATL. Active Template Library.
WinForms were really decent and that was enough. Keep Win32 API and a managed wrapper around it as WinForms and that would have been more than enough.
.
The answer is, clearly, Qt.
(This isn't even a joke. Assuming you're not going Electron.)
Part of the reason here is that it's their main business. They can't afford letting it languish and/or changing direction every 6 months.
When a platform can’t answer “how should I build a UI?” in under ten seconds, it has failed its developers. Full stop.
That's fine, except no platform answers this.
Obviously Linux doesn't, but the Mac doesn't either. Apple of course has it's recommendation, but most developers do not take Apple's recommended path because of course, it's Apple-only, most developers make cross-platform apps these days.
Even if Microsoft decreed the one-and-one Windows development path, most developers are not taking that path.
It used to be the case that Mac developers used Apple tools, Windows developers used Microsoft tools, but those days are gone. Developers want to use Electron, or Qt, or some other system to support multiple platforms in one codebase.
Microsoft has less to do with this than the article makes out. I'm a desktop developer. I don't care what Microsoft recommends, or what Apple recommends, because neither work in the real world where supporting only their platform just isn't realistic.
As of today, I only use 2 solutions for Windows GUI :
- windows forms in .net
- flutter
All the rest always presents itself with a sheer aura of "It was a great idea but we couldn't finish it".
Without ever discussing with anyone from MS about it, I think they stopped improving/working on this because of electron.
Any web developer can build a good enough website and a good enough desktop app with electron.
My coherent GUI strategy: just use JUCE.
Invested in it, shipped it, seen it solve the cross-platform problem beautifully.
Can just write C++, and see it running everywhere.
The JUCE GUI capabilities are more than adequate for many, many things.
There are other platform-scaffolded cross-platform frameworks. JUCE is cromulantly FNORD.
This barely mentions Windows Forms, which is the cleanest and fastest way to code Windows GUI apps.
A few years ago, I wanted to prototype something quick and I wrote it in Windows Forms over C# (all code, no visual editor).
Wow, totally ignores the one framework Microsoft developed that was actually a pleasure to use (WinForms). Avalon sucked IMO; it was just a continuation of the fever dream at the time of making everything XML (or XAML in that case). MFC sucked because it was C++. WinForms was that brief Window of “modern programming language with simple, elegant GUI framework”.
I know it’s not a popular opinion, and I am sure there were reasons Microsoft abandoned it, but that was a brief few years when I actually enjoyed building GUIs on Windows.
> From their perspective, gambling on a new managed-code framework had produced the most embarrassing failure in the company’s history
Most embarrassing failure in the company’s history that far.
> WPF shipped in late 2006. It was remarkable – XAML, hardware-accelerated rendering, real data binding. If Microsoft had made it the definitive answer and invested relentlessly, the story might have ended differently.
Er… The author perhaps never used it? WPF was the worst framework I ever used. It was unbearably verbose, brutally unforgiving, used 2-way bindings that created updating nightmares, ans not the least it was incredibly slow.
WinForms was not the best for sure, but at least you can get stuff done. It was for a long time the right answer to the question the author asked. .Net + WinForms worked well.
When WPF shipped was when the shit hit the fan.
I remember I was trying to learn GUI programming in late 2000s. I took an attempt at MFC but no matter how hard I tried, it just didn't make any sense. Then I found out about Qt 4 and it felt like magic immediately. I ended up working as a professional Qt software developer for 3 years.
Fast forward to now, and KDE Plasma is state of the art Linux desktop that I've been daily driving for years. It's also built on Qt.
I used to be a software test engineer at Microsoft in the late 90's and early 2000's. What I learned is that 60% of their business was International because of their language support. Also most of their customers are business customers locked in to using their Windows Server Domain system for corporations. These corporations also run office and Sharepoint and build custom apps using office. This is why they have so many different ways to build apps on Windows. We are not the users they are targeting, they target the enterprise.
Programming Windows is THE authoritative source on Win32 programming:
https://www.charlespetzold.com/books.html
It is a fantastic book, I learned everything I know on Win32 from it. Wrote real time scientific software in windows for ~10y! We did it all, external hardware control, custom UIs, etc. Thanks Ryan Geiss for your timing info.
Right about VC6 was the sweet spot imo, C/C++ with lightning fast UI for docs and more. Tools got out of the way. Once other languages got involved (C#?) the docs got out of control and harder to use, and the UI started to get a little overloaded.
The snappiness of those old windows systems was pretty great.
I would bet in Avalonia UI. It's like WPF but cross-platform. https://avaloniaui.net/
Microsoft GUI development is a mess. They don't seem to care. Just look at the mishmash of different GUI styles in Windows 11.
Thankfully I have been mostly insulated from it by sticking to Qt and C++ for the last 25 years.
At Redmond, there was also the Patterns & Practices group (P&P) that tried to make sense of the dev products, and built extra libraries “to show customers how to use them”. They followed the bouncing ball of the frameworks releases from the main development teams. It suggests that it wasn’t clear exactly how you’d use the main products: so P&P said, ‘try it like this.’
I also think the article didn’t say much about MS in the web era. The company survived webdev IMO, but it definitely wasn’t the leader
This article smells so strongly of AI that I'd be surprised if the author did much of any writing.
So the windows team hates .NET, so they use react webviews and put them in explorer and start menu???
Windows frameworks got web envy.
Xaml and styling and all sorts to try and compete.
Trouble is, it made desktop development harder, and it didn't win against the web. It left the simple and safe formula of standard and common controls for a designer's playground, but the designers preferred the web. And if you make something for the web, you can package it in Electron and get cross platform cheaply.
A decade or so ago, I had a clear idea of what a "native ui" should look and feel like, even if there were multiple routes to get there. I don't know any more.
There’s a poster on here who keeps posting and re-posting about their dinner with a Microsoft executive and how they were told Microsoft is all-in on enterprise. Waiting for that copy-paste to make its way over here.
Microsoft keeps footgunning things so hard I think even enterprise might be reluctant to go with them moving forward [0]. I don’t have Netcraft numbers in front of me but I doubt things have notably improved even if they do have a strategy shift to enterprise which includes crapping all over Windows for no good reason.
I’m personally glad FOSS is going strong but that’s a complete aside.
[0] We got burned by Azure as I’m sure many other enterprises have, and they did exactly nothing to remedy/compensate the situation, SLAs be damned. At this point our strategy is to move off of reliance on any Microsoft/windows tech. We moved off of ActiveDirectory not too long ago. Bing/Edge/etc honestly who cares.
The biggest issue in my opinion is that Microsoft is chasing web devs. XAML is pita, just like css and html markup. They need to pick a lane and just use it. Allow other frameworks access to the default low level api such as Win32, but document it to the same level as Google does on Android and Apple with SwiftUI
Given how bad windows has become since windows 7, I’ve been wondering. Does Bill Gates still use Windows? Does he put up with the horribleness?
I very much dislike WPF. If I have to do a windows UI (and usually when I do it's a simulator for some piece of hardware), I honestly just grab WinForms. It's stupid simple.
I've worked inside the Microsoft 365/SharePoint ecosystem for over 20 years, and the incoherence runs deeper than the GUI layer.
Microsoft has been developer first, then cloud first, now AI first. It has never been design first. And when something does work, the pattern is always the same: force it everywhere. Metro was genuinely good on mobile, then they pushed it onto desktop where it didn't fit. Now Fluent UI is repeating the cycle.
Fluent UI is on its third major rewrite (Fabric, then Fluent UI React v8, then v9). Each one brought breaking changes. v9 appeared around the same time Microsoft released their new Teams client, and it shipped with performance and accessibility issues that are still being worked through. Signs already point to the next throwaway Fluent UI version. In practice, Fluent seems to be first-party only. It's 'open source' in the sense that you can look but don't touch. In SharePoint alone, the same UI elements look different across pages, lists, document libraries, and settings panels.
The irony is that the web platform itself has caught up. You can build a fully themed, accessible component library in plain HTML and CSS today. No framework dependency, no version lock-in. I did exactly that with hTWOo (https://htwoo.io/), an open-source alternative that matches the Fluent design language without any of the React coupling.
The real problem isn't that Microsoft can't pick a GUI strategy. It's that they keep rebuilding from scratch instead of creating a design system flexible enough to sustain across platforms and product generations. The web platform already provides the primitives. You just have to commit to them.
Today i still use C# with Windows.Forms. All my old knowledge is still usefull. People know how to use a Windows.Forms program.
Microsoft's biggest mistake was .Net being a Java competitor when it should have just been like golang producing native binaries. Especially since .Net was realistically only going to succeed on x86/64 at that point (late 90s and 2000s). This shut the door on C# for consumer UIs, and people stuck to Visual Basic and MFC.
It took them more than 2 decades to finally support pure native binaries (via NativeAOT). And it's fantastic for servers on Linux.
As an outsider, it always felt that 'internal team politics' were the root cause of Microsoft's technological downfall. This went far beyond the GUI. In 'flagship' offerings like SharePoint or Skype/Lync/Teams you could see the unmixable partial code-bases brutally forced together under a single product name an foisted upon a user-base baffled by how this could have gone so wrong.
This is what happens when one's performance is measured by "impact".
Microsoft has a very coherent strategy, described aptly in "fire and motion" by Joel spolsky 24 years ago: keep changing the API/GUI every 2 years:
Fun and sad. Garbage collected languages like .NET, Java, and Go are slow compared DIY languages like C, Rust, and Swift. So, no wonder the Windows team dislike them. However, it seems the Windows team had to live with a React start menu for quite some time. How does that fit into the over all battle between the teams and UI frameworks?