I am worried about the future of native UI technologies on Windows. Traditionally at least the developers of operating systems have eaten their own dogfood and have at least tried to implement well-performing & visually consistent native applications to serve as an example to others. Windows 11 has largely done the opposite. Windows has had minimal but perfectly functional native email and calendar apps at least since Windows 10 (could have been in 8, never used that). Windows 11 originally shipped with those apps, but they were removed in a later update and replaced with laggy webview wrappers that take seconds to start.
From the WinUI community calls, I would assert all new employees have zero Windows experience and management doesn't care to give them proper skills.
Too many questions that any Windows developer would know why the question was being asked, where they either couldn't answer or had puzzled looks on why the questions were being asked in first place.
That is also a reason why now there are Webview2 instances all over the place on Windows 11.
> That is also a reason why now there are Webview2 instances all over the place on Windows 11.
I can't help but feel like there is a happy median between Native UI and WebUI that we haven't quite figured out yet.
In the gaming space there is this library called RMLUI that I have used in anger quite a bit. It gives you something that is shaped a lot like HTML and CSS, but with a built-in data binding layer, and a scripting layer that supports Lua by default but gives you the flexability to roll your own language and API. It also is a much lighter dependency than Chromium and V8.
It's missing a couple of features from vanilla HTML and CSS, but also has a bunch of unique featurea that make it far more useful in other ways. For example, it doesn't have CSS background images, but image decorators are so much more useful. And don't get me started on sprite sheets and theme media queries.
I can't help but think that something similarly shaped and designed more for general desktop use would blow the doors off of electron. My workstation has 32 gigs of RAM and yet it's often a half to three quarters full mostly because of Firefox, Discord, and Visual Studio Code instead of more lightweight apps like Visual Studio 2022.....which sounds crazy when you say it out loud.
FWIW, the start menu is a React Native app. I think that’s a pretty good compromise (I think the performance criticism is solvable, they just don’t care). Too bad RN isn’t that well supported on other platforms.
The coolest part was that it was baked into the .NET Framework installed on all newer Windows versions, so I could make a useful GUI app in like 30 KB.
I dislike some superficial things about it (and some less superficial things about WPF), but a lot of the ideas around how properties are set are pretty cool to me!
Also how (IIRC) it’s compiled to pretty standard .NET view code (“partial”) you could extend in other parts of the app.
I’m still happy to have left it behind for the aforementioned, ecosystem & tooling reasons.
I heard that the WPF code under the hood was jank, which might be part of why MS never went forward with it. From a developer perspective it was quite nice. I started in winform and I miss UI frameworks being obvious and straight forward. I have a rant about how to darken an image using HTML + css that despairs in how unintuitable and bodge-job the solution ends up being. WPF and WinForm were not like that, you draw what someone sees, inline, as they see it and that's quite nice.
Well, to be fair, Microsoft decided to kill the Windows API that everybody knew, and spent about a decade and half creating a replacement every few years that couldn't actually replace the original thing.
It's hard to survive that. Honestly, I don't even know what the GP is talking about when they say the devs don't know "Windows".
Yup. I remember 4 or 5 different frameworks that were supposed to be the future of native UI on windows, but each one after WinForms was harder to use, slower, and less capable than what came before.
nah WPF made way more sense than WinForm in its abstract design, as it supported flow/stack layout ideas instead of anchors like WinForm did. WinForm was much quicker to get started but had a fundamental flaw in that it couldn't really do transparency.
Yes, mshtml was insanely powerful. As an undergrad I was surprised how easy it was to build capable UIs with powerful visual effects[1] using mshtml and JS. Even for C++ Windows apps.
It took a long time for this to become a cross-platform reality, but it did inspire me to ignore distractions like XAML and focus on the web.
This was already a thing 20 years ago. Students weren’t have any experience with windows, it was something companies used, and today even that has gone away.
Windows is still enforced and mandated on most corporate Fortune 100/500 companies computers.
Their "cloud lift" consists of putting vms into the cloud to run at 10x the cost. (Nothing else changed from The Old Ways)
And that's still where we are today for most enterprises.
That ecosystem has had them for 30+ years and shows no signs of going away anytime soon. If it doesn't have Active Directory and Office 365 or whatever they decided to call it today, they aren't interested.
I think web views do make sense in situations where you’re presenting lots of remote content that may frequently change. After all that’s what the web is, and store content, and to an extent emails many of which are HTML anyway, are reasonable candidates.
yeah in that regard it seems that apple tastefully does so on apps where there's mostly remote roundtrips already like the app store or music app, so I agree that there makes sense to reuse the web infra
but swift ui apps are great and fast cause they're not electron monsters!
thankfully you can use safari webkit inside them, but that doesnt work cross-platform
SwiftUI apps are not great and they're not fast. A lot of Apple's new apps are considered rather poor. Theo has a video where some devs switched to a webview because the text rendering performed better!
Mostly agreed... I'd love to see MS just ditch MAUI and/or otherwise buy out UNO or Avalonia and create a cross-platform XAML based UI library that works everywhere (including Linux). As far as Linux goes, they could just go with Gnome/GTK for underlying support... not sure if QT's license would be compatible at this point, MS mostly uses MIT for modern FLOSS.
Bonus points if it worked with C, C++ and Rust for those that want it in addition to .Net (C#).
I honestly think there's still too many old guard managers/exec at Microsoft to let that legitimately happen though. They're too busy trying to wrangle every last penny of value out of every Windows user along the way.
The issue I see with Windows 11's UI is they seem to focus too much on pushing new apps / features and not enough focus on catching up some of the older tools within Windows. Take for example the Control Panel which is a reskinned version of the same one that shipped with Windows 7. And I'm sure there are tools buried within the OS that probably date back to the 2000/XP days.
Windows 11 looks great if you just look at the press photos and stay a "very happy path" while using it but as soon as you start digging deeper you realize it's like that meme of Homer Simpson with clips on his back.
It's done when you're not visiting old control panel anymore. I wonder what people are even doing regularly enough in control panel for that to be such a problem.
I don't use them terribly often, but one thing that old control panel windows did was pack a lot of information into a small amount of space.
Device Manager has that nice tree setup allowing me to see problematic devices a lot quicker.
Drive management has that nice table of physical devices on the top and then partitions on the bottom.
The old-school network adapter page shows me every network device in a single window, real or virtual, and I can mess around with their settings a lot faster with right-click properties.
And of course the services manager allows me to make sure that certain services arent running, like Windows search indexing - I use voidtools Everything instead.
That said, those tend to be the exceptions and not the rules. I've actually been quite pleased with the progression of the settings window over the course of Windows 10 and 11. A special shout out to the new fonts page as it is light years better than what came before.
Funny, I recently wanted to uninstall a bunch of fonts I tested for programming. Turns out, in the new font settings widget you cannot uninstall all fonts or even all styles of a font at once. I would have needed to select a font, then click "Uninstall" for the bold style, the semibold style, the medium style, the italic bold style etc.
Opening \Windows\Fonts instead gives you a nice view of all fonts, you Ctrl-select all you want, right click, uninstall, done.
I won't be surprised if there is an effort to rewrite MSO in something like Dart and WASM, and make it independent from any native toolkits altogether. Yes, to reproduce all of the Excel power, and make it available everywhere as a premium plan of O365.
Then Windows could pull a ChromeOS. The only place where a native UI is really needed is the lock / login screen; a tiny subset of any current UI toolkit would suffice.
Office is a completely separate team divorced from Windows proper. Unless Office deemed it wise to rewrite their UI, they're not going to do so (and it's a frankenstein of a Win32 UI).
Office on Windows relies heavily on COM and other Win32-only libraries to function.
I can't think of a valid reason to rewrite Office to that extent. They already have Office for the web and Mac Office; while not identical in features, they're often good enough outside of BI scenarios or highly complex Excel work.
Outlook is the lone exception where that team decided to have Outlook for the web, Windows Outlook, and Mac Outlook be identical, so those are getting their rewrites with removal of Win32-specific features where applicable.
Before Project Reunion came to be, Office team was starting to adopt UWP.
See BUILD recordings from 2018, I think, where they demo the new UWP controls contributed by the team, similarly to what happened before with the ribbon on Windows 7.
I would vouch they got as happy as the rest of us.
> Outlook is the lone exception where that team decided to have Outlook for the web, Windows Outlook, and Mac Outlook be identical, so those are getting their rewrites with removal of Win32-specific features where applicable.
I wish they didn't. Outlook on macOS is abysmal nowadays and I still find myself resorting to the legacy view just to change some settings that both iterations can read but only one exposes.
I significantly prefer using Thunderbird or the web views for Gmail and Zoho Mail over any version of Outlook. Is the integration across O365 apps nice? Sure, but the platforms themselves are miserable to use.
In a similar vein, I was cautiously optimistic about Teams V2 for unifying the client. But then they completely dropped the Linux client for their PWA which does not have feature parity with the "native" platforms and has a significantly worse UX.
I think that Outlook as an experience has become significantly worse as o254 has advanced... I understand some of the reasons, such as managing millions of inboxes vs a few dozen on a private server. That said, it's kind of a mess.
I used Thunderbird for years, mostly to keep my NNTP feeds going, at some point, my mailboxes corrupted between versions (the NNTP site I was using didn't work right anymore) and I just kind of gave up at that point. It was in the dark time, before the more recent resurgence of dev activity, but NNTP didn't seem to be even a tangential focus.
I'd love to see a relatively easy open-source server for mail+contacts+calender that allowed the level of visibility, management and sharing that Outlook+Exchange/o365 offers. Right now, I have some of that, but not nearly the same.
Edit: Also, really sick of getting meeting change notifications in Teams for a meeting that is months away... just leave any meeting notifications beyond today/tomorrow in outlook, I don't need them in my workgroup app.
No, MSO should not be rewritten. It might be adapted to a more universal UI toolkit, if Win32 UI becomes problematic as is. COM is also not going to go anywhere, but I wonder if WASM-compiled components and native code-compiled components could interact via DCOM over the internet.
If you understood the power struggles within Microsoft and the cut throat office politics, you’d understand. Orgs are fighting orgs trying to over throw one another.
It’s as if it’s taught in business school or something, how to throw colleagues under the bus. I’ve seen it at every large corporation except for one.
My take on why not ui? though is because they are so busy trying to justify saving their own skins. How could they possibly look outward into the future when they’re busy recreating lord of the flies?
>It’s as if it’s taught in business school or something, how to throw colleagues under the bus.
You're overreaching. Business school doesn't need to teach that, that's just human nature being driven by greed and survival of the fittest, the base blocks of capitalism.
I refuse to believe that. I’ve seen completely normal humans treat each other with respect in the workplace, using data for decision making and accepting the data. It doesn’t have to be King of the Hill.
If I see an org start playing squid games, I’m out. I don’t care how much money I lose or what. I’m out.
> I’ve seen completely normal humans treat each other with respect in the workplace, using data for decision making and accepting the data.
Yes, and once the VC money dries out or the line goes down and the layoffs are on the table, they'll be at each other's throats to make sure the others go on the chopping block and not them.
Human nature is all about survival. People who are elevated beyond human nature are privileged people who don't live in a zero sum environment where every day is a fight for survival or those who are spiritually rich and can just as well adapt to homelessness as a lifestyle.
Let's take for example Nordic countries which are relatively peaceful and low crime due to welfare state, but if for example, all the food in their supermarkets were to suddenly vanish overnight with no way to replenish it quickly, then that peaceful society would flip into a survival of the fittest society where police and army would be the first to use their guns to secure food for themselves and their families.
You see this in the animal studies, where if they only give food to some monkeys, then the other monkey beat the others to steal their food.
The behavior you're seeing in the corporate world is an extension of that, not something needed to be taught. Never bet against human nature and the survival instinct, as all those who did, have been eliminated from the gene pool.
Doesn't windows have like 10 different UI frameworks? Using windows 11 is like going to the natural history museum. Certain apps like MSC still look and feel like windows 2000, then you've got some of the metro influence stuff that is big and bold and blocky colors, and then you have the win11 stuff that is more modern but honestly just looks like lipstick on a pig (right click context menu is a perfect example, it looks modern but has limited functionality and so you do the show more button and yep you guessed it just shows the older style and more functional menu)
I do not think that has ever been true. Features in Office and other Microsoft products were always more advanced than what was possible in the SDKs. Third party UI toolkits would often fill the gaps until MS decided to allow everyone to use those features, often years later. I guess they did it to make Office feel fresh.
> We are being thoughtful about resourcing. This effort is happening alongside other critical responsibilities like security, platform stability, and support for existing products. Our current focus is on foundational work that unlocks value for contributors and increase transparency. We are aligning this work with Microsoft’s broader business priorities to ensure long-term support and impact.
I don't sense any benevolence in their words. They are just pulling off their resources and dumping the framework on the public, hoping passionate losers will contribute.
This is unduly meanspirited. Your passion projects are not even considered to probably the vast majority of the world; that doesn't make you a loser.
I have zero interest in the Win11 UI, and am even on board with the cynical view that this is purely a bean counter cost savings for MS rather than some benevolent outreach.
But I respect the people that take this on and want to keep it going.
> Your passion projects are not even considered to probably the vast majority of the world; that doesn't make you a loser.
Not OP but I understood this as "contributing for free to a project owned by a corporation worth more money than you could realistically spend in a lifetime is what makes you a loser".
That is the self-interested feeling that Open Source preys on.
And I do mean "prey" with a negative connotation. One of the biggest perks of Open Source from a company's perspective is that you can get developers to work on your project for free without paying them. However, those same developers have very little say in the direction of your product, and any forking of your project would have to compete the economies of scale that come from being a company. The only downside is that you have to worry about being out-scaled by a bigger company, as the developers of ElasticSearch, Redis, Docker, and others found out first-hand.
This is distinct from Free Software, which has different dynamics that are much more friendly to mutual benefit, collaboration, and forking, especially if there's no CLA that pools all of the copyright into one corporate or non-profit entity. But then again, this sort of Free Software moralizing is expressly the reason why Open Source was created as an alternative in the first place. The OSI even used to admit as such on their website:
That’s great for you, but don’t underestimate the asymmetric value here. You’re giving them free labour, and they get _way more_ out of that than you do - especially in aggregate.
As has the rest of the world, and we will just put it on the list of UI frameworks Microsoft did not completly implement, fully support or considering "the default".
So we stay stuck with the status quo: There's no official UI for Windows, still.
There is Win32, that has made us company for several decades, is kind of ok, definitely much better than Motif, the official UI for much UNIXes, before Linux took over.
Thanks for calling it out. I get OPs passionate disdain for Microsoft but one must remember that the world is built on contributions from such people. Take the whole Linux and GNU ecosystem for example. We’d be lost without them.
Maybe the biggest beneficiary will be AI/LLMs - which will become way better at creating Windows UX after this.
I don't know if that's quite what you meant, but I believe that this is the first time that I've heard of AIs being treated as a beneficiary and an "end" in themselves.
It's sad how OSS enables predatory practices by the mega corps. It's definitely not just Microsoft, Google, Amazon at least also behave the same way, and the worst part is how the audience is captive - if they're stuck on any platform (since all the platform owners behave the same way), despite paying for it, the only way to fix broken parts of the platform is to fix it yourself.
For any time a legitimate bug gets a "we always accept pull requests" comment from a $XXXK paid engineer, I really wish I had control of an orbital laser, or a deathnote with eyes that work with text.
My favorite one was when the docs and helm template of Grafana was broken and I sent an email about it, I was told that they only have a single person working on the docs and that they were out of office, and that I would be free to do a PR
Then I learnt that this company is valued at >$6 Billion
People do what they want to do, and might get enjoyment out of it. It can be as simple as that. That MS gets benefit from it, sure, and that's probably even their goal, but to say "I'm not going to do a thing that I enjoy, just because someone else might get benefit from it" is just a spiteful, mercenary way to go through life. If that's what keeps you going though, more power to you.
More seriously, a desktop UI toolkit is hardly a moat by now, especially a Windows toolkit, Windows having 3-4 very different look-and-feels mixed and shipped with the official distribution.
OTOH security and stability are things that Windows critically depend on to stay on the laptops and desktops in medical, governmental, and financial institutions, and on devices of executives.
This feels like Windows itself is no longer producing enough growth for Microsoft relative to its other efforts. Even the enterprise sales lock-in isn't compelling enough for the cloud/AI-centric future Microsoft envisions. So Microsoft is slowly pulling resources that it can instead invest into Azure and AI and other high-growth business units.
I don't watch Windows too closely. Have there been any other signals of waning investment into Windows? Has Nadella or the other leadership admitted to this?
Hasn't Microsoft also been pulling back from Xbox? IIRC, haven't they been trying to consolidate and use gaming to lionize Windows as a platform? After spending billions on multiple AAA studios? That would seem counter to a Windows pullback strategy. Is this a case of the left hand not talking to the right hand?
> Have there been any other signals of waning investment into Windows?
Wasn't there a story some while back about them crawling the web for PWAs and putting them on the Microsoft Store (or is it Windows store?) to make it into less of a ghost town? And if you go to their official website, browsing vaguely in the direction of UI development, you will see them advertising PWAs as first-class citizens of the Windows ecosystem. I also vaguely remember that Windows+Edge offers special APIs to PWAs for things like file system access and so forth that are unparalleled on other platforms.
I take this push for PWAs (combined with their own lack of dogfooding -- Office is not written with Win UI) as them basically throwing in the towel on Windows-native desktop software (outside of games, maybe).
But, to be fair, native desktop development has seen a lack of investment on all desktop platforms. JavaFX is a ghost town too.
All of that could change, depending on what happens next with Chrome, Bing, and Mozilla. -- The future of each of those seems to be hanging in the balance at the moment.
The web could become a mere implementation detail of the Google monopoly, rather than the open thing it is today. Couple that with a government-level push for digital sovereignty in the E.U. and other places (certainly China). Then, maybe, you will see renewed interest in desktop GUI apps.
On a side note: I think it's amazing what has happened in the open-source space with Rust-based UI frameworks (iced, egui, slint) and COSMIC. The future for cross-platform desktop UI development hasn't looked so bright, maybe since the introduction of Java Swing (was that in the early 00s?)
> Qt and VCL/FireMonkey are pretty much alive, as 3rd party .NET like Avalonia and Uno.
Last time I looked into Uno, it seemed to me more like a mobile-first play (similar to Flutter), but maybe I missed something there.
Qt is still joined at the hips with C++, and I'm finding it hard to imagine that the next generation of developers will go in for C++ over alternatives like Rust. There is a QML/JavaScript story. But why, on earth, would you go there, if history hasn't forced your hand, as it did with the Web.
I hear good things about Avalonia, but: Why throw out the baby with the bathwater and build something "open" within a monopolist's walled garden.
> Apple is also doing just fine.
But "lack of investment" is still not entirely improper as a characterization of how they prioritize desktop versus mobile.
...the amout of code we've been writing has increased with the progress of technology, just comparing what it was like to write desktop UIs with Delphi in the 90s versus the amount of code it takes for an equivalent app in 2025. Low-code and vibecoded software from the 2020s are going to be the unemployment insurance for all those greybeards who will still know how to code in the 2030s :-)
That’s wishful thinking. Just because things haven’t changed in the past doesn’t mean it won’t change in the future.
I don’t doubt that the amount of code will not reduce, it’ll just be easier and easier to get AI to fix it.
We are still less than two years into widespread use of this technology, and it’s surprising how good it is.
I am a ‘greybeard’ compiler guy and modern LLMs fix compiler bugs better than me, to a large extent. And it keeps slightly getting better every few weeks.
As compiler guy, how do you see direct machine code generation?
I firmly believe having LLMs generate code for current languages is a transition step, just like Assembly devs had to be convinced optimising compilers were generating the same kind of code they would write themselves.
LLMs’ design is around novelty and creativity. It was the missing left brain of computation. We should stick to traditional styles of computation for compilation.
Presumably this would need some advancements in formal verification to make sure the results are correct.
But ignoring that for now, LLMs can already do better than modern compilers when it comes to optimizing code.
It has become a neat way of finding opportunities to implement in a compiler.
Give the LLM a compiler generated piece of assembly and it will sometimes spit out a slightly better version. Then you try and figure out if it's possible to implement it in the compiler. This works really well for blind spots in compilers like generating good vectorized code.
Sure, it took only what, 40 years of intensive hardware improvements* for assembly to move to the fringe? And we still reach out to it more often that I would like to because reasons?
Yep, I guess you can train an LLM on a bunch of binaries to get it to mimic a SotA compiler with some accuracy, which may or may not improve over time, but come on. Times where there were free performance increases are gone, and this is not the area where shipping any bullshit real fast will get you any sort of advantage.
* Which are unlikely to happen again in the foreseeable future.
> I think it's amazing what has happened in the open-source space with Rust-based UI frameworks (iced, egui, slint) and COSMIC
Do you mean it is amazing how poor is an experience and how many features, widgets and controls they lack when compared with something like FPC/Lazarus and Qt?
Rust is a nice thing, but its UI ecosystem simply cannot even compare.
Rust is - by design - antithetical to pretty much every idea of rapid application development paradigm Delphi/VCL and to lesser extent Qt adhere to.
It doesn't matter how many of Rust UI toolkits there are. Consider that there are a lot of Rust game engines, and pretty much zero games written in it, because even C++ gives you better trade-offs in that particular space.
IMO it's not ideal for Rust to have only Rust specific toolkits (at least they need to have other language bindings)
I've tried using Qt (non QML, using rust-qt) last week... initially things were looking fine (despite unsafe everywhere)... but I found quickly that I need unsupported things (deriving from C++ classes), and the project isn't really maintained anymore, so I'll need to so some low level work myself if I want to proceed.
Linux desktop share is up 20-35% in the past year, depending on sourcing. Still only 5-6% total for US at least, but definitely a shift. MS is almost definitely making more money for o365 and Azure than Windows (though I don't follow shareholder releases).
From google's ai result Windows ($25b) is 1/2 the size of Office (54.9b) and 1/3 the size of Azure ($75b) in terms of revenue.
Windows is definitely not the cash cow for Microsoft it once was.
Usually I get why companies release their UI frameworks. I've strongly considered using Atlassian's and AWS's frameworks in the past to build web apps because if it's good enough for Jira/AWS, it's probably good enough for my B2B saas app.
But I personally don't know why anyone would reach for this framework. Maybe if you're building a Windows app and you want a very consistent look and for your app to feel "native", but aren't there better options out there for doing this already?
No one in Windows development community cares about WinUI, other than those with sunken costs that bought into the WinRT/UWP dream and now are stuck with a dead technology.
Too many burned bridges since Windows 8 came out.
If anything, this is Microsoft confirmation that they are unwilling to fix all the broken issues, and hoping the community will somehow still care.
Very true. We just developed a brand new LOB desktop app and settled on sticking with WPF. WinUI has been dead for years imo.
On a side note, I still love WPF after working in it for 10 years. Maybe it's just familiarity, and it's a little verbose at times, but man it's a great framework when you know that you're doing.
We also settled on WPF for a new LOB Desktop application, this validates the decision. If you combine WPF with the CommunityToolkit MVVM, it’s a very nice framework to develop with.
DevExpress supported WinUI for a little while but decided to abandon support.
One of the biggest problems with WinUI compared to WPF is that DependencyProroperty is implemented as native code, so for .NET developers, there is a huge performance penalty getting or setting any property on control.
Honestly at this point who would seriously use any Microsoft UI framework? They've abandoned 100% of their previous UI frameworks unfinished when they get distracted by a new, shinier framework.
Why use a busted incomplete framework missing basic features when there's entire ecosystems of open source cross-platform frameworks being actively maintained and which actually have all the features you need?
Really this is just another UWP destined to be forgotten and scorned.
WPF uses DirectX 9, in any case it is good enough for many businesses use cases.
UWP never had feature parity with neither Forms nor WPF, so already it has a hard sell to businesses.
Microsoft marketing always cool about cool experiences and design, without the actual meat in features.
That is why there were API reboots between Windows 8, Windows 8.1 and Windows 10.
Then comes along Project Reunion, which is supposed to properly unify Win32 and UWP execution models, instead a few months in, they do yet another reboot, that to this day is a shadow of UWP features, endless bugs, and a team that everyone thought was also part of the layoffs, has they were radio silence for months.
Forms is still my go-to for quick one-off and disposable programs. But maily because it's simple and I have a big library of controls and extensions I've built over years to make Forms more complete.
I've also tried WPF, but it's missing very basic controls that I just don't feel like recreating.
I already lost count how many UI frameworks are in windows. It looks like complete chaos and mess.
I really wonder what they expect from open-sourcing it. Just to pretend how open they are? Or is there any real benefit to developers who target windows?
WinUI is an evolution of UWP which is an evolution of WinRT. WinUI has been around for years.
MAUI is not exactly a competing product and is more about enabling cross platform UI development. Different intent.
WinUI is actually ok tech. It’s evolved over the years through a few iterations, now on WinUI 3.
Im mostly with you though. Until they rebuild the entire OS in it, including all of the administrative controls and tools, I don’t trust the longevity.
They already do, though. The big UI refresh in Win10 is all XAML, and the new Win11 taskbar (the one we all hate) is now a totally normal XAML app.
WinUI 3's big changes (to get a 3.0 version number) is not with the XAML stack itself, but its new ability to be called by unmanaged apps as a normal UI toolkit, so it can finally be used by all apps. No more using Shell UI like we're writing Win 3.1 apps.
And yes, some stuff in Win11 still isn't WinUI, which is kind of annoying, but some of those dialogs hidden away in Windows are at least 20 years old, and probably would need to be entirely rewritten, not merely have their UI's updated.
Also, fun fact: The Win8/10 taskbar's code predates Avalon (the prototype/codename for WPF), and trying to change/fix it at all usually ended up breaking it. It's one of the few binaries on Windows that would not be recompiled to build a new release image in fear of breaking it. Rewriting the taskbar made sense, GETTING RID OF SMALL MODE DID NOT, GODDAMNIT MICROSOFT.
Lets not forget that after five years Project Reunion was announced, WinUI 3.0 is yet to achieve feature parity with the Visual Studio experience developing C# or C++ applications, or UWP components features.
The WinUI map component is a Webview2 instead of proper native component, Win2D is only a subset of the UWP one, ink is yet to be migrated, and lots of other issues.
Github repos are filled with thousands of issues, and they already did a cleanup a year ago where they simply closed enough tickets to bring it under 2000.
> The Win8/10 taskbar's code predates Avalon (the prototype/codename for WPF), and trying to change/fix it at all usually ended up breaking it. It's one of the few binaries on Windows that would not be recompiled to build a new release image in fear of breaking it.
The taskbar that underwent a major redesign in Windows 7 (released after WPF)? Also, that binary is explorer.exe, surely it got rebuilt quite often for new ads. features, and fixes?
The taskbar exists in its own timeline, apparently. (Ex-)Microsoft engineers working in the UI/UX team for Windows itself have spoken about how much of a nightmare dealing with that thing was.
Also, the taskbar technically exists as a dll, not an exe, explorer.exe would link to it and run it if it was being ran as the UI shell. This is now split out to its own exe (Shell Experience Host iirc), and explorer.exe is now only File Explorer.
Since it exists as a dll, btw, this is how the Win10 taskbar injectors work, they just call the dll (which still ships in Win11) instead of letting Shell Experience Host do the Win11 thing.
The sane way of handling this, btw, is just use Shell Experience Host injectors to get the desired behaviors, such as using Windhawk and use m417z's taskbar height and icon size plugin (the third most popular Windhawk plugin); to match Vista/7 era small, set to height 32, icon size 16, taskbar button width 28.
> And yes, some stuff in Win11 still isn't WinUI, which is kind of annoying, but some of those dialogs hidden away in Windows are at least 20 years old, and probably would need to be entirely rewritten, not merely have their UI's updated.
And this is hard to do. These dialogs often are _dynamic_, with third-party settings rendered as ActiveX controls.
I recently noticed that they introduced an option for small icons. Not that it changes much, as the height of the bar stays the same, but hey. Personally I've been fine since they added back the option not to combine buttons unless full.
I was going to ask about Win32. I haven't had to do it in a while, but if I had to write a desktop app in windows, that would be what I would reach for. It's still supported... is their any indication that it won't be for many years to come?
Also, it looks better, in my humble opinion. It's probably lacking features that I'm uninterested in.
I implemented a WPF control to look like a winform control and ran into a issue of the vintage control looking perfect at an odd (9px by 9px) native pixel size that gets distorted by WPF when rendered at different scaling.
Visual Studio's + sign expanders have had this bug for years that was finally fixed by changing to a chevron shape in 2022. Clearly this problem bugged someone on the dev team hence the fix. =)
WinUI 3 still supports WinRT. It ALSO supports more. It's an evolution of WinUI 2, not just a simple version bump, but also not a completely new tech. It's probably a closer evolution to go from WinUI 2 to 3 than it was to go from Angular 1 to 2.
I think this is completely independent. You can simply use WinRT APIs, because Win32 Apps can use them now.
WinUI3 apps are win32 apps.
> but also not a completely new tech.
Not sure about this.
UWP APIs work out of the box. For WinUI3 you need the Windows App SDK, and it is much slower and heavy than UWP (out of curiosity I created a very simple app and it was fast just a few dozen kbs big)
You have that backwards. WinRT is the managed languages runtime for Windows, introduced in Win8. Its sort of the replacement for COM/OLE but also defines the ABI dialect in a way that allows managed languages to call unmanaged code without an FFI penalty.
UWP is built on WinRT, and acts as a fully managed app container, similarly to how phone apps exist on your phone. It allows WinRT apps to be deployed to any Microsoft platform, Windows, XBox, Windows Phone, etc, but also Android and iOS, and also as PWA, and are guaranteed to run identically on any of those platforms. UWP apps must be written a fully managed language that runs on the CLR (ex: C# runs on the CLR, but C++/WinRT does not). UWP also uses the second generation of WinUI-family XAML UIs, which means all UWP apps use completely native UIs, instead of slow non-native Javascript shit in a web canvas.
The WinUI family of XAML UIs started with WPF, and a slightly incompatible version of it also appeared in Silverlight (WPF = WinUI 1.0), then was brought to UWP (= WinUI 2.0), and is now its own stand alone thing that any app can use, managed or not, as 3.0.
WinRT is not an attempt to move beyond .NET, instead it is their way of allowing .NET to natively call code, and make .NET languages first class in Windows.
Yeah but I think when it was introduced it wasn't a thing you could use separate from the rest of UWP. What changed in Win10 was you could use WinRT APIs from regular Win32 apps too. They started breaking UWP up into independent pieces.
Or not. I haven't thought about this stuff for years. Definitely possible I forgot the ordering of things.
UWP most certainly did not exist when WinRT came into existence. WinRT, itself, is also more or less the third version of this attempt: Managed C++ came first, then C++/CLI, then the WinRT era and C++/CX.
C++/CLI and C++/CX are semi-managed, you only need to use the extensions when interacting with WinRT, although you're free to write your entire app in that dialect.
WinRT is not the same thing as managed .NET code. There is no requirement that a UWP is .NET. There are many examples of unmanaged C++ UWPs, including the open source Windows Terminal.
WinRT is a mechanism to express APIs in a way that is amenable to cross-language usage. It is built on top of COM, and is not a replacement for COM.
Windows Terminal is sort of a clusterfuck of multiple programs, only one of which is a UWP program (the actual visible WinUI shell that runs) which you're not allowed to do as a UWP-era Microsoft Store app, but they got internal permission to do that.
Its now a WinUI 3.x program, apparently, and now the Store no longer requires UWP programs, so WT now only needs to make the "ships inside of Windows release images" guys happy, which apparently is harder than making the UWP-era Store guys happy.
I genuinely thought a lot of these sibling comments were satire at first! The acronyms, the lengthy and confused explanations of versions, frameworks etc:
> WinUI is an evolution of UWP which is an evolution of WinRT
They probably started something new and shiny (Now with AI!) and want to get rid of the old baggage without causing too much of a user revolt (all dozens of them) ;)
There are three UI frameworks in Windows, and only two actively used/developed.
All the other "countless" frameworks are iterations of one of two lines: Win32/Native (WinAPI, MFC, WinRT, WinUI3, etc) and WPF/Managed (Avalon, WinUI2-3, etc). WinUI3 exists to bridge the gap.
You could also go for wxWidgets as it is kinda MFC-y but better and cross-platform, though like MFC you can combine it with Win32 API code (almost) seamlessly.
Or go with Qt, though that doesn't use native controls.
No, it implements its own functionality. As an example consider one of the most basic controls which is available pretty much everywhere (i.e. no need for polyfill), the push button: the source code[0] for QPushButton clearly implements the behavior itself, it does not rely on any native button.
Compare with wxWidgets' equivalent to QPushButton, wxButton, where there is a backend-specific header[1] and implementation[2] where 99% of the wxButton functionality is there (there is a `btncmn.cpp` under `common` that is shared across backends but that has very little code itself).
I don't even need to dig through the paint loop. I've written QT software, and you can do it yourself.
Add a QButton with no styling, you will have a standard button widget in macOS and Windows (or a KDE styled one in Linux). Now have an event that modifies the button in some way (adding a CSS style to it is common, but also changing the sizing manually will usually do it) and the style of the button will immediately change to a QT-style one. The difference is so obvious in Windows and macOS, that it would be laughable to claim otherwise; but you can also just ask the QT Group devs or the community, they've reiterated/experienced this behavior plenty of times.
Maybe people can cannibalize some of the rendering code and extrapolate the controls to a better class library than they already have. Like a kind of winforms but using modern rendering APIs. I know you already can create such controls but they often end up being very verbose and just look like xaml but in C#
I kind of wish Microsoft would just continue development of WPF. I've used it for years for various projects, and there is a learning curve but I've since enjoyed working with it. XAML, data bindings, ViewModels... all of it I actually like. But, WPF needs a few improvements to really make it perfect. I tried several of Microsoft's newer frameworks and the open source ones (Avalonia, Uno), but I either couldn't get the sample projects to even build successfully on my machine, or I never got comfortable with development workflow, and went back to what I know.
My big idea to fix WPF is to rebuild the data binding system to use the .NET compile-time code generation feature instead of run-time reflection. I think that would solve a lot of problems. For one, projects could do an actual AOT build of their applications (right now, you either need to rely on an installed .NET runtime or "publish" the project with a lot of .NET libraries included for self-extract, bloating the final file size). Code generation would probably improve performance quite a bit too, maybe open up the possibility to compile for cross-platform, introduce type safety for XAML bindings (rather than getting vague runtime binding errors), remove the need for so much class scaffolding, etc... I've thought about starting an open source project to do it myself, but seems like a pretty big task and I would essentially be starting a project to help with my other project which I already don't have enough time to work on...
Your second paragraph sounds like you're describing Avalonia. Avalonia has AOT, compile-time binding errors and cross-platform support. Maybe there have been some updates since you last tried it? I'm not very familiar with Avalonia or WPF though so maybe there's more to it than that.
Thanks, yes I'll probably have to give it another try some day. I might be confusing Avalonia and Uno, but I think I first attempted it a couple years ago, and then again last year. I remember spending a whole weekend trying to get it running but wasn't having success. Also, I was a bit turned off by how heavy the development environment was. I had to download and install a tool, then that installed more build tools and packages, and then there was also a "recommended" VS Code extension. With WPF, I've gotten used to writing XAML without a designer, so I can get by with just VSCode, the C# extension, and the .NET CLI.
It would also allow for assembly trimming, which would be a huge boon if you are trying to do a self contained deployment. Right now you either do framework dependant or have like a 200MB+ deployment.
Last I evaluated it WinUI3 was a terrible developer experience. The application had to be literally installed on the system to even debug it, which means you end up with a large number of useless start menu entries, not to mention registry entries and such. Another thing was that the example programs crashed when I clicked on a button.
All I want is something simple to work with to make applications for Windows, and so far I'm still using Win32 with WTL.
I want to be able to create self-contained GUI application that are relatively small and can be just copied and run on another computer. Installation should be an option, not a requirement. From my evaluation, WinUI3 doesn't offer that.
WinUI3 offers that, it's just not the recommended way of building+deploying.
What you are looking for is "unpackaged, framework-dependent". This will build an application like an old-school WinAPI executable, one which expects the relevant DLLs to be installed on the host system (or distributed in the same directory).
I know that these days a lot of game and application developers just render a custom UI to a 2d or 3d rendering device. I've been looking for a UI framework like that as I don't want to write my own, but so far most of them have been either web based or require their own mini-language and compilation step.
You mean just self-drawn controls or also immediate mode GUI?
I haven't tried it, but there's ImGui.NET (or a bit easier ImGui.Forms) immediate mode GUI. Downside is no accessibility.
Avalonia uses Skia to draw controls, although it primarily uses XAML to define UI.
The issue with self-drawn controls is that creating such framework is a lot of work, so smaller projects often have issues with things like accessibility.
EDIT: Just did hello world test with ImGui.Forms. Self-contained AOT compiled result was 7 MB.
Within game development it is predominantly immediate mode as it more closely resembles the main game loop. With other applications I'm not sure, some I know just use ImGui and run an immediate mode UI, others would be retained mode.
In my case I'm looking for a C++ UI library, not .Net. Also not a fan of immediate mode GUIs either, so would want something more traditional.
I don't want to write my own, so will probably try Qt again, but might switch to a custom library in time.
I did try to develop an unpackaged test application but I gave up trying to implement it and just went with Win32 instead as I wanted make something rather than messing around with a UI framework.
These days if I were to switch from Win32 I might try some custom rendered framework which a lot of apps seem to use, or Qt.
Many here point out, and rightly so, what an abominable mess UI frameworks have been on Windows for quite a while.
That being said, I'd point out that the situation has been (and still is) unfortunately far from satisfying on the cross-platform and open-source side as well: GTK has also been a mess for quite long; Qt has - aside from the bloat - an increasingly repulsive licensing model for professional use (too bad the hopes from the early Nokia adoption days got smashed by MS mole Elop and the later owners of Qt), …
… and while there are some nice solutions for specific niches (e.g. Dear Imgui), the overall situation in general-purpose cross-platform native UI frameworks looks quite bleak and unsatisfying imho. There's a huge gaping opening for a permissive-licensed open-source cross-platform UI/Widgets framework that compiles to native, has a good set of widgets and gives access to Vulkan contexts for 3D rendering.
People like to hate on Electron and React Native (because web bad!!!!) but tend to fail to realize there’s basically no alternative if you want to be platform flexible.
Microsoft could’ve made a real impact in this area with a great advantage to themselves, but both efforts I’m aware of were so half hearted they could never have worked out.
I hope this leads to having a native vertical taskbar, which has been absent in W11 despite being a taskbar feature dating back as early as Windows 98.
Third-party tools have tried to reimplement it but it's either been by bastardizing the native W11 horizontal taskbar to be vertical (eg: Windhawk) or just restoring the old W10 taskbar code (eg: StartAllBack).
How will making the UI framework open source lead to taskbar changes? For third party contributions in this area, they need to open source the taskbar, not the UI framework.
> Nit-pick: Windows 95, actually. The vertical taskbar was an option in its very first version.
Yes, but this was when computer were less capable. Now, you are not able to choose colors or fonts, the taskbar is fixed and the programs need at least 2 seconds to start.
Terrific evolution.
The Start Menu is apparently a React Native app, so I'm going to hazard a guess and just assume WinUI is built on top of React, and that the Start Menu at least is thus indeed built with WinUI. But it's also clear that some other parts aren't, so who knows what's what. I'm sure there are folks who spent time reverse engineering it all though who do.
The start menu is not a React Native app, but it's actually even worse. Only the recommended section (which is basically recently used files - plus probably advertisements in some scenarios) is. The rest of the start menu is WinUI, to my knowledge.
I cannot stand the latency using a local app. Same with rendering views of local file systems. Frontend reactivity as the expense of responsive performance is the problem with modern user interfaces in my opinion.
Like I’m searching for an installed app. I don’t need news articles about that and never expected a file system ui to be a web portal.
The talk in question showcases a widget in the start menu that's using react native. Apart from the meme itself, I have not found indication of the start menu being a react native app. Every time this comes up it seems to always lead back to that meme.
That'd be correct. See also the top comment in that thread that (evidently after I originally read that thread) also explained exactly this, and a sibling comment in this thread tree that also did so (more than two hours after I posted my original comment, meaning I was unable to amend or delete it).
For Windows UIs I've been getting into Win32/GDI/DirectDraw/etc.
Tools like CsWin32 and modern C# (ref returns) make working with these APIs a lot more approachable today. It used to be the case that you had to create a nasty C++ project to do any of this. Now you can just list the methods you need access to in your nativemethods.txt file and the codegen takes care of the rest.
Win32 is a lot lower level than other things you'd typically consider to be a "UI framework", but the important tradeoff is that it is also a lot harder for Microsoft to remove or screw with in any meaningful way. I cannot come up with something that has been more stable than these APIs. The web doesn't even come close if we are looking at the same timescales.
You still need C++ in many places, because of the COM rulez attitude within Windows team.
Windows Runtime Components was a lost opportunity to level up the play field for .NET.
As such, if you want to do something like a shell extension, or context menu extension, it is C++ as always, or having your little C++ stub that calls out into a .NET process.
Yeah this entire discussion about these high level framework would be cooler to discuss the lower level APIs. As far as I know, the entirety of Windows rendering stack is built on either GDI or DirectX? With Win32 even being built on GDI?
It would be much cooler to discuss building a ground-up Windows UI stack as close to metal as possible which I guess would necessarily be using DirectX...
I think windows needs a community effort to create an actually good framework for native development on windows. Unfortunately I just dont think such a community is big enough.
What I miss about UI development on Windows is the commitment to help me create applications that seamless blend in, and look like something that Microsoft themselves had created. The introduction of web technologies has allowed the visual language of the Windows experience to fragment. Not just because Microsoft hasn't kept older applications up to date (like the Control Panel), but also because the modern tooling isn't providing the libraries to match their style.
From memory, this first started during Vista, where 3rd party devs just weren't able to match the translucency of the new Aero UI. And the lack of articles being published by the MSDN group with titles like "Here's how you add this new feature to your app" with examples in C, C# and Visual Basic.
My analogy is every Microsoft UI framework was almost completed in the sense of someone being almost pregnant.
A framework that has just one show-stopper missing feature or problem is... unusable. You can't embark on a large, complex application development journey if you even suspect that you'll be painted into a corner.
For example, many of WPF-derived frameworks had atrocious performance, with fundamental mistakes in their design that made them incompatible with list virtualization. It wasn't until they had to eat their own dogfood and use WPF for Visual Studio that they started fixing these issues.
Win UI 3 meanwhile dropped all support for HDR, wide-gamut, etc... going backwards to SDR sRGB only in an era where all mobile phone manufacturers were starting to standardise on OLED HDR displays. The logic behind this decision? Microsoft wanted a UI framework that is "mobile compatible"!
From watching the community calls, long after I stopped caring, management doesn't seem to care to actually hire people that have Windows development background, many times they would ignore community questions or don't get where they were coming from.
As someone reading the comments here and never made a real Windows app outside of a visual basic hello world a pretty long time ago. Why doesn't Microsoft just stop making these? They already own GitHub and vscode so why not just admit that electron/typescript is the Windows UI framework now?
I wonder how much longer Microsoft stays committed to Windows as a whole. Windows is less than 10% of the company, users are migrating to phones, tablets, and Chromebooks (all of which can run Office), and with .NET on Linux, Windows servers are making less sense. It's a shrinking market.
But Windows is where most people run Office, including all the back-end implications like Exchange and all the "Power" stuff.
It is a lot more than 10% of Microsoft revenue really (just not directly).
I agree though that they do not really care about the platform anymore. In fact, they care less and less every day about everything that is not AI on Azure.
It will be a long time before Microsoft Exchange, Active Directory, and Excel are not important enough to enterprises to justify the Microsoft subscription. And Visual Studio is still a huge part of the Microsoft capture strategy.
Microsoft is working on separating all of the above from Windows without losing too much control but it is going to take a while.
Exchange should be SaaS for 99% of businesses. There's a web version of Excel that's probably OK-ish. I'm not sure what Active Directory looks like in a world without Windows.
Not every company is a startup that only needs gmail and Figma, though. The ecosystem of LoB apps and device controller software is ginormous, and not going away any time soon.
Microsoft has a long history of releasing numerous UI frameworks: VB, MFC, WTL, Silverlight, WPF, WinForms, and others. Yet despite this abundance, many of the core components Microsoft used in its own applications were never available to developers. They rarely ate their own dog food, and desktop UI development relied on third party components. For the past two decades, native desktop UIs have steadily declined in favor of web-based components, so it's unclear what the real benefit of another native framework would be today.
The "bad thing" is that it's effectively getting abandoned, open sourcing it won't make any difference.
It's not like external contributions will suddenly turn it into something usable, and they'll just have a skeleton crew maintaining it, like they do WinForms and WPF.
People are tired of Microsoft and their ever growing graveyard of ill thought out, half-baked, "native" UI frameworks.
Native UI is effectively dead outside of Apple’s platforms, and even there it’s hanging on for dear life. HTML, CSS and JavaScript won the cross platform toolkit battle.
Sadly yes. And all the platforms are to blame. Microsoft and their 1000 half-working frameworks made writing a wrapper that was any better than wxWidgets impossible.
But also Apple "totally not deprecating" AppKit and pushing everyone to the mess that is SwiftUI, Gnome breaking backwards compatibility as a sport, and Qt messing around with QML, meant "native UI" became quicksand.
Even going HTML, CSS and JavaScript wouldn't be too bad if the browser engines provided by the OSes were any good, but it took Microsoft giving up and switching to rebranded Chromium as a browser for Windows to provide a usable one in WebView 2.
WebKitGTK is also terrible compared to the macOS version of WebKit, which hurts projects like Wails and Tauri. So everyone bundles a freaking copy of Chromium with their applications.
On the Apple side of things, AppKit and UIKit work as well as they always did, and they’ve been less pushy about SwiftUI lately probably because they realized that the old toolkits aren’t going away any time soon.
For Qt, the hard-coupling of C++ or Python for Widgets and Quick being JS-centric haven’t done it any favors. C++ and Python are fine, but not everybody wants to write either, and most people interested in writing JS are going to gravitate towards front end web stacks over anything else.
I think that for a cross platform desktop UI toolkit to see any degree of long term success, a high degree of bindability is non-optional even if it’s most capable when used with its native language. The toolkit needs to meet developers where they are, and that means being usable in the language of their choice.
The most important thing the web standards get right is their insistence on never ever breaking backwards compatibility. HTML, CSS, and JS accumulate a lot of cruft, but they do move forward into the future without leaving anyone behind.
> The most important thing the web standards get right is their insistence on never ever breaking backwards compatibility. HTML, CSS, and JS accumulate a lot of cruft, but they do move forward into the future without leaving anyone behind.
This is the comment you originally responded to. Flash never had anything to do with web standards, which do indeed strive for backwards compatibility, it's why that classic space jam website still works.
The comment was not that the "web" as a whole strives for backwards compatibility. If that were the case we would also be running ActiveX controls and Java web applets.
They only won because developers stopped giving a shit about anything except their own ease of work. There's no such thing as a good UI built with web tech, so anyone who cares about the user's experience will use a native toolkit despite the difficulties. But very few do, turns out.
That's extremely uncharitable. I think slack and discord have pretty great UIs, and that's all "web tech." Figma just IPO'd at a $58 billion valuation - is that not proof their UI is good? If it wasn't good, no one would use it. VS code became the preeminent text editor and IDE over the last decade - all web tech.
Too many reboots since WinRT was introduced in Windows 8, many battle scars and lost wars, only people without WinRT experience can think anything positive about WinUI.
Microsoft probably has no interest in WinUI anymore and intends to replace it with WebView, which is why they're making it 'open source.' Microsoft's open source is literally a graveyard.
If only they'd open source Windows Explorer and the taskbar/start menu, rather than resisting peoples attempts to customise them through other hackery.
The answer to this question depends on the knowledge and quality of engineers working on the kernel and the overall Executive. These continue to evolve with more advanced technologies, like VBS or the future usermode endpoints for EDR and possibly anti-cheat, pushing those out of the kernel, which presumably requires kernel work.
David Cutler is still there but working on other stuff, last I understood.
That does lead to the question of 'would they do it the same way and/or follow the NT OS/2 spec' to get a functionally identical Windows today.
If abandonware of Win11 UI means Win12 then maybe I like that direction :)
On the other hand it it is smpossibility that MS is working on something better, right ? They probably plan some worse UI prison...
ASCI was [business] enforced by some EU project, POSIX - the same and probably there are more examples of such things so we just need some sanity __OS with UI__ project from someone...
I really hope they do and the rendering engine is decently decoupled, I'll give a try building a framework on top of it.
I wish all platforms gave access to their rendering engine similar to DOM on the web, imo SwiftUI/WinUI (or WPF, but they are very similar) are not that good.
Haven't built anything native on Linux, though, no idea how good those are.
What do you mean by access? APIs to program against or fully open sourcing the rendering engine? Because you can mix SwiftUI with a few different rendering frameworks at varying abstraction levels that it itself renders to (AppKit, UIKit, Core Graphics, Metal etc.)
Basically I want an API available to build my own SwiftUI. Definitely not on the Core Graphics level :)
But good point, I actually think AppKit might be a good abstraction level. I'll play with it a bit and see if I can abstract it behind a good component model.
It is a ton of C++ for what is essentially something that an OS like Windows/MacOS/Android/iOS or the browser would provide anyway. Apps that use it ship with a substantial minimum amount of bloat, e.g. Flutter for web.
For a second there I read it as "Windows 3.11's UI" and it got me really nostalgic. I wonder if it's been long enough that it's ready to be hip again now.
I would be fine with really any version as open source or at least one that was free-as-in-free-beer to make it possible to maintain virtual machines running old software without having to rely on dodgy downloaded versions... I'd even pay Microsoft something reasonable if they put them up on GOG or some similar site for a few $.
Having had to setup some Windows systems for a client of mine, all I can say it that they've long abandoned all hope of ever shipping a somewhat coherent UI. It's just not going to happen. Lipstick upon lipstick on frankensteined pigs just aren't going to work. At some point they'll have to start from clean slate or else we'll see 1990s ui remnants leaking through into the 2040s. It's just not going to happen until they start bleeding market share on the desktop. iirc it took desktop linux ~10 years from 1 to 2% and 1 year from 3 to 4% so there's movement in the market.
DHH pushing desktop Linux for developers, Pewdiepie making videos about Linux and GrapheneOS, there's definitely an accelerationist argument to be made here.
I am worried about the future of native UI technologies on Windows. Traditionally at least the developers of operating systems have eaten their own dogfood and have at least tried to implement well-performing & visually consistent native applications to serve as an example to others. Windows 11 has largely done the opposite. Windows has had minimal but perfectly functional native email and calendar apps at least since Windows 10 (could have been in 8, never used that). Windows 11 originally shipped with those apps, but they were removed in a later update and replaced with laggy webview wrappers that take seconds to start.
From the WinUI community calls, I would assert all new employees have zero Windows experience and management doesn't care to give them proper skills.
Too many questions that any Windows developer would know why the question was being asked, where they either couldn't answer or had puzzled looks on why the questions were being asked in first place.
That is also a reason why now there are Webview2 instances all over the place on Windows 11.
> That is also a reason why now there are Webview2 instances all over the place on Windows 11.
I can't help but feel like there is a happy median between Native UI and WebUI that we haven't quite figured out yet.
In the gaming space there is this library called RMLUI that I have used in anger quite a bit. It gives you something that is shaped a lot like HTML and CSS, but with a built-in data binding layer, and a scripting layer that supports Lua by default but gives you the flexability to roll your own language and API. It also is a much lighter dependency than Chromium and V8.
It's missing a couple of features from vanilla HTML and CSS, but also has a bunch of unique featurea that make it far more useful in other ways. For example, it doesn't have CSS background images, but image decorators are so much more useful. And don't get me started on sprite sheets and theme media queries.
I can't help but think that something similarly shaped and designed more for general desktop use would blow the doors off of electron. My workstation has 32 gigs of RAM and yet it's often a half to three quarters full mostly because of Firefox, Discord, and Visual Studio Code instead of more lightweight apps like Visual Studio 2022.....which sounds crazy when you say it out loud.
FWIW, the start menu is a React Native app. I think that’s a pretty good compromise (I think the performance criticism is solvable, they just don’t care). Too bad RN isn’t that well supported on other platforms.
Are you describing XAML?
Possibly? I was a web developer before I started delving into desktop development, so I don't really know a ton about XAML.
I used XAML a bit at an old job
The coolest part was that it was baked into the .NET Framework installed on all newer Windows versions, so I could make a useful GUI app in like 30 KB.
Those were the days... I hated that job though.
Everything I've ever heard about XAML from folks who have used it has been positive.
Seems to me like the kind of thing that was forgotten about due to being joined at the hip with Microsoft, not necessarily because it was a bad idea.
I dislike some superficial things about it (and some less superficial things about WPF), but a lot of the ideas around how properties are set are pretty cool to me!
Also how (IIRC) it’s compiled to pretty standard .NET view code (“partial”) you could extend in other parts of the app.
I’m still happy to have left it behind for the aforementioned, ecosystem & tooling reasons.
I heard that the WPF code under the hood was jank, which might be part of why MS never went forward with it. From a developer perspective it was quite nice. I started in winform and I miss UI frameworks being obvious and straight forward. I have a rant about how to darken an image using HTML + css that despairs in how unintuitable and bodge-job the solution ends up being. WPF and WinForm were not like that, you draw what someone sees, inline, as they see it and that's quite nice.
RMLUI looks neat! closest I've seen to an open source sciter alternative.
so microsoft gave up and the web won?
swift ui apps have some webkit views like the app store, music app etc
Well, to be fair, Microsoft decided to kill the Windows API that everybody knew, and spent about a decade and half creating a replacement every few years that couldn't actually replace the original thing.
It's hard to survive that. Honestly, I don't even know what the GP is talking about when they say the devs don't know "Windows".
Yup. I remember 4 or 5 different frameworks that were supposed to be the future of native UI on windows, but each one after WinForms was harder to use, slower, and less capable than what came before.
I disagree, WPF is quite good to use, and WinForms approach to keeping UI as code is quite moronic. I would say WPF is where it should've stopped.
WPF sucks. I rather like expressing the UI in code. Having to deal with so much XML (XaML) is annoying as fuck.
The workflow on WPF was so much more painful than winforms, which had a decent WYSIWYG designer and solid integration with visual studio.
WPF in comparison was slow, memory hungry, and difficult to learn.
I tried 5 times to make a WPF application but it didn’t even have all the same basic controls that WinForms supported.
nah WPF made way more sense than WinForm in its abstract design, as it supported flow/stack layout ideas instead of anchors like WinForm did. WinForm was much quicker to get started but had a fundamental flaw in that it couldn't really do transparency.
>and spent about a decade and half creating a replacement
I often feels modern Apple is the same. I mean we are 10 years into Swift, 6 years into Swift UI.
And it feels they are still in beta.
And yet it feels more polished than WinUI, that is how bad it has gotten.
What to expect when new interns have no clue about Windows, have been educated with macOS and ChromeOS, and the design team carries Apple devices?
Windows 98 introduced Active Desktop, and still, not as many webviews all over the place.
MSHTML was the first Electron.
Yes, mshtml was insanely powerful. As an undergrad I was surprised how easy it was to build capable UIs with powerful visual effects[1] using mshtml and JS. Even for C++ Windows apps.
It took a long time for this to become a cross-platform reality, but it did inspire me to ignore distractions like XAML and focus on the web.
[1] https://learn.microsoft.com/en-us/previous-versions/windows/...
This was already a thing 20 years ago. Students weren’t have any experience with windows, it was something companies used, and today even that has gone away.
I promise you it has not "gone away" which is why Azure is so popular.
What does Azure have to do with windows these days?
Windows is still enforced and mandated on most corporate Fortune 100/500 companies computers.
Their "cloud lift" consists of putting vms into the cloud to run at 10x the cost. (Nothing else changed from The Old Ways)
And that's still where we are today for most enterprises.
That ecosystem has had them for 30+ years and shows no signs of going away anytime soon. If it doesn't have Active Directory and Office 365 or whatever they decided to call it today, they aren't interested.
Still not seeing the connection to Azure.
Wasn't AD like one big security flaw after another?
I think web views do make sense in situations where you’re presenting lots of remote content that may frequently change. After all that’s what the web is, and store content, and to an extent emails many of which are HTML anyway, are reasonable candidates.
yeah in that regard it seems that apple tastefully does so on apps where there's mostly remote roundtrips already like the app store or music app, so I agree that there makes sense to reuse the web infra
but swift ui apps are great and fast cause they're not electron monsters!
thankfully you can use safari webkit inside them, but that doesnt work cross-platform
SwiftUI apps are not great and they're not fast. A lot of Apple's new apps are considered rather poor. Theo has a video where some devs switched to a webview because the text rendering performed better!
Do Windows developers use Windows or Mac at work?
Naturally they use Windows, Microsoft designers apparently not any longer.
It's impossible to develop Windows without Windows... All dev tools and the repository ecosystem designed to work under Windows.
Well, you can use Avalonia or UNO Platform on Linux or Mac and then deploy to Windows.
UNO Platform is basically WinUI.
Mostly agreed... I'd love to see MS just ditch MAUI and/or otherwise buy out UNO or Avalonia and create a cross-platform XAML based UI library that works everywhere (including Linux). As far as Linux goes, they could just go with Gnome/GTK for underlying support... not sure if QT's license would be compatible at this point, MS mostly uses MIT for modern FLOSS.
Bonus points if it worked with C, C++ and Rust for those that want it in addition to .Net (C#).
I honestly think there's still too many old guard managers/exec at Microsoft to let that legitimately happen though. They're too busy trying to wrangle every last penny of value out of every Windows user along the way.
The issue I see with Windows 11's UI is they seem to focus too much on pushing new apps / features and not enough focus on catching up some of the older tools within Windows. Take for example the Control Panel which is a reskinned version of the same one that shipped with Windows 7. And I'm sure there are tools buried within the OS that probably date back to the 2000/XP days.
Windows 11 looks great if you just look at the press photos and stay a "very happy path" while using it but as soon as you start digging deeper you realize it's like that meme of Homer Simpson with clips on his back.
You forgot our favourite windows 3.1 file dialog:
https://youtu.be/r549Zn74Xg8
Yikes!
Though nothing from Apple in 1993 still runs.
Control Panel is definitely a bad example, I do see them working little by little, update by update, to integrate everything into Settings
It's been like 10 years since that settings panel came out with Windows 8. How many more years until it's done?
It's done when you're not visiting old control panel anymore. I wonder what people are even doing regularly enough in control panel for that to be such a problem.
I don't use them terribly often, but one thing that old control panel windows did was pack a lot of information into a small amount of space.
Device Manager has that nice tree setup allowing me to see problematic devices a lot quicker.
Drive management has that nice table of physical devices on the top and then partitions on the bottom.
The old-school network adapter page shows me every network device in a single window, real or virtual, and I can mess around with their settings a lot faster with right-click properties.
And of course the services manager allows me to make sure that certain services arent running, like Windows search indexing - I use voidtools Everything instead.
That said, those tend to be the exceptions and not the rules. I've actually been quite pleased with the progression of the settings window over the course of Windows 10 and 11. A special shout out to the new fonts page as it is light years better than what came before.
Funny, I recently wanted to uninstall a bunch of fonts I tested for programming. Turns out, in the new font settings widget you cannot uninstall all fonts or even all styles of a font at once. I would have needed to select a font, then click "Uninstall" for the bold style, the semibold style, the medium style, the italic bold style etc.
Opening \Windows\Fonts instead gives you a nice view of all fonts, you Ctrl-select all you want, right click, uninstall, done.
I do agree they didn’t do much in the 8/10 days but definitely feels like there’s a big push since 11
last time I used Sql Server Management studio the "add database dialog" still had that poorly placed button that its had since its release.
I mean business or corpo world literally force microsoft to support backward compability
in fairness rewrite it from scratch is literally easier than supporting compability since dawn of times but they forced to do it
I won't be surprised if there is an effort to rewrite MSO in something like Dart and WASM, and make it independent from any native toolkits altogether. Yes, to reproduce all of the Excel power, and make it available everywhere as a premium plan of O365.
Then Windows could pull a ChromeOS. The only place where a native UI is really needed is the lock / login screen; a tiny subset of any current UI toolkit would suffice.
Office is a completely separate team divorced from Windows proper. Unless Office deemed it wise to rewrite their UI, they're not going to do so (and it's a frankenstein of a Win32 UI).
Office on Windows relies heavily on COM and other Win32-only libraries to function.
I can't think of a valid reason to rewrite Office to that extent. They already have Office for the web and Mac Office; while not identical in features, they're often good enough outside of BI scenarios or highly complex Excel work.
Outlook is the lone exception where that team decided to have Outlook for the web, Windows Outlook, and Mac Outlook be identical, so those are getting their rewrites with removal of Win32-specific features where applicable.
Before Project Reunion came to be, Office team was starting to adopt UWP.
See BUILD recordings from 2018, I think, where they demo the new UWP controls contributed by the team, similarly to what happened before with the ribbon on Windows 7.
I would vouch they got as happy as the rest of us.
Those versions in the Office store are d.e.d. Except for OneNote for Windows 10 which is shortly on it's way out.
Yes, because of how UWP and Project Reunion went down, right after they started looking into it.
> Outlook is the lone exception where that team decided to have Outlook for the web, Windows Outlook, and Mac Outlook be identical, so those are getting their rewrites with removal of Win32-specific features where applicable.
I wish they didn't. Outlook on macOS is abysmal nowadays and I still find myself resorting to the legacy view just to change some settings that both iterations can read but only one exposes.
I significantly prefer using Thunderbird or the web views for Gmail and Zoho Mail over any version of Outlook. Is the integration across O365 apps nice? Sure, but the platforms themselves are miserable to use.
In a similar vein, I was cautiously optimistic about Teams V2 for unifying the client. But then they completely dropped the Linux client for their PWA which does not have feature parity with the "native" platforms and has a significantly worse UX.
I think that Outlook as an experience has become significantly worse as o254 has advanced... I understand some of the reasons, such as managing millions of inboxes vs a few dozen on a private server. That said, it's kind of a mess.
I used Thunderbird for years, mostly to keep my NNTP feeds going, at some point, my mailboxes corrupted between versions (the NNTP site I was using didn't work right anymore) and I just kind of gave up at that point. It was in the dark time, before the more recent resurgence of dev activity, but NNTP didn't seem to be even a tangential focus.
I'd love to see a relatively easy open-source server for mail+contacts+calender that allowed the level of visibility, management and sharing that Outlook+Exchange/o365 offers. Right now, I have some of that, but not nearly the same.
Edit: Also, really sick of getting meeting change notifications in Teams for a meeting that is months away... just leave any meeting notifications beyond today/tomorrow in outlook, I don't need them in my workgroup app.
No, MSO should not be rewritten. It might be adapted to a more universal UI toolkit, if Win32 UI becomes problematic as is. COM is also not going to go anywhere, but I wonder if WASM-compiled components and native code-compiled components could interact via DCOM over the internet.
It makes a lot of sense to wire COM and WASM in some smart way.
If you understood the power struggles within Microsoft and the cut throat office politics, you’d understand. Orgs are fighting orgs trying to over throw one another.
>the power struggles within Microsoft and the cut throat office politics
That's most old large orgs who have been around for ~5 decades. Nothing special about Microsoft.
It’s as if it’s taught in business school or something, how to throw colleagues under the bus. I’ve seen it at every large corporation except for one.
My take on why not ui? though is because they are so busy trying to justify saving their own skins. How could they possibly look outward into the future when they’re busy recreating lord of the flies?
>It’s as if it’s taught in business school or something, how to throw colleagues under the bus.
You're overreaching. Business school doesn't need to teach that, that's just human nature being driven by greed and survival of the fittest, the base blocks of capitalism.
I refuse to believe that. I’ve seen completely normal humans treat each other with respect in the workplace, using data for decision making and accepting the data. It doesn’t have to be King of the Hill.
If I see an org start playing squid games, I’m out. I don’t care how much money I lose or what. I’m out.
> I’ve seen completely normal humans treat each other with respect in the workplace, using data for decision making and accepting the data.
Yes, and once the VC money dries out or the line goes down and the layoffs are on the table, they'll be at each other's throats to make sure the others go on the chopping block and not them.
Human nature is all about survival. People who are elevated beyond human nature are privileged people who don't live in a zero sum environment where every day is a fight for survival or those who are spiritually rich and can just as well adapt to homelessness as a lifestyle.
Let's take for example Nordic countries which are relatively peaceful and low crime due to welfare state, but if for example, all the food in their supermarkets were to suddenly vanish overnight with no way to replenish it quickly, then that peaceful society would flip into a survival of the fittest society where police and army would be the first to use their guns to secure food for themselves and their families.
You see this in the animal studies, where if they only give food to some monkeys, then the other monkey beat the others to steal their food.
The behavior you're seeing in the corporate world is an extension of that, not something needed to be taught. Never bet against human nature and the survival instinct, as all those who did, have been eliminated from the gene pool.
Doesn't windows have like 10 different UI frameworks? Using windows 11 is like going to the natural history museum. Certain apps like MSC still look and feel like windows 2000, then you've got some of the metro influence stuff that is big and bold and blocky colors, and then you have the win11 stuff that is more modern but honestly just looks like lipstick on a pig (right click context menu is a perfect example, it looks modern but has limited functionality and so you do the show more button and yep you guessed it just shows the older style and more functional menu)
What a cluster.
I do not think that has ever been true. Features in Office and other Microsoft products were always more advanced than what was possible in the SDKs. Third party UI toolkits would often fill the gaps until MS decided to allow everyone to use those features, often years later. I guess they did it to make Office feel fresh.
> Alignment with Microsoft Goals
> We are being thoughtful about resourcing. This effort is happening alongside other critical responsibilities like security, platform stability, and support for existing products. Our current focus is on foundational work that unlocks value for contributors and increase transparency. We are aligning this work with Microsoft’s broader business priorities to ensure long-term support and impact.
I don't sense any benevolence in their words. They are just pulling off their resources and dumping the framework on the public, hoping passionate losers will contribute.
> passionate losers
This is unduly meanspirited. Your passion projects are not even considered to probably the vast majority of the world; that doesn't make you a loser.
I have zero interest in the Win11 UI, and am even on board with the cynical view that this is purely a bean counter cost savings for MS rather than some benevolent outreach.
But I respect the people that take this on and want to keep it going.
> Your passion projects are not even considered to probably the vast majority of the world; that doesn't make you a loser.
Not OP but I understood this as "contributing for free to a project owned by a corporation worth more money than you could realistically spend in a lifetime is what makes you a loser".
what if that contribution benefits me personally in any way whatsoever?
That is the self-interested feeling that Open Source preys on.
And I do mean "prey" with a negative connotation. One of the biggest perks of Open Source from a company's perspective is that you can get developers to work on your project for free without paying them. However, those same developers have very little say in the direction of your product, and any forking of your project would have to compete the economies of scale that come from being a company. The only downside is that you have to worry about being out-scaled by a bigger company, as the developers of ElasticSearch, Redis, Docker, and others found out first-hand.
This is distinct from Free Software, which has different dynamics that are much more friendly to mutual benefit, collaboration, and forking, especially if there's no CLA that pools all of the copyright into one corporate or non-profit entity. But then again, this sort of Free Software moralizing is expressly the reason why Open Source was created as an alternative in the first place. The OSI even used to admit as such on their website:
https://web.archive.org/web/20021001164015/http://www.openso...
That’s great for you, but don’t underestimate the asymmetric value here. You’re giving them free labour, and they get _way more_ out of that than you do - especially in aggregate.
And? If I get value, I get value. If someone else gets MORE value, that doesn't diminish the value _I_ get.
Viewing ones gains only through the lens of how it affects others seems like a hard way to live.
You need to ask OP, I don't know what they meant, just how I personally understood it.
> I have zero interest in the Win11 UI [...]
As has the rest of the world, and we will just put it on the list of UI frameworks Microsoft did not completly implement, fully support or considering "the default".
So we stay stuck with the status quo: There's no official UI for Windows, still.
There is Win32, that has made us company for several decades, is kind of ok, definitely much better than Motif, the official UI for much UNIXes, before Linux took over.
In Motif, i can set the shadowThickness as much as i want. /s
I rather not deal with Xlib C API.
I mean, Apple gets a lot of flak for changing things and deprecating old frameworks, but I’ve lost count of the number of post-win32 UI frameworks?
Thanks for calling it out. I get OPs passionate disdain for Microsoft but one must remember that the world is built on contributions from such people. Take the whole Linux and GNU ecosystem for example. We’d be lost without them.
Maybe the biggest beneficiary will be AI/LLMs - which will become way better at creating Windows UX after this.
I don't know if that's quite what you meant, but I believe that this is the first time that I've heard of AIs being treated as a beneficiary and an "end" in themselves.
>This is unduly meanspirited.
No it's not? It's accurate.
>Your passion projects are not even considered to probably the vast majority of the world; that doesn't make you a loser.
A Windows UI project is not a "passion project" and the only """person""" that really benefits here is Microsoft.
>But I respect the people that take this on and want to keep it going.
Contribute to something else that doesn't only help the bottom line of a mega global corp?
Why are you defending this?
It's sad how OSS enables predatory practices by the mega corps. It's definitely not just Microsoft, Google, Amazon at least also behave the same way, and the worst part is how the audience is captive - if they're stuck on any platform (since all the platform owners behave the same way), despite paying for it, the only way to fix broken parts of the platform is to fix it yourself.
For any time a legitimate bug gets a "we always accept pull requests" comment from a $XXXK paid engineer, I really wish I had control of an orbital laser, or a deathnote with eyes that work with text.
My favorite one was when the docs and helm template of Grafana was broken and I sent an email about it, I was told that they only have a single person working on the docs and that they were out of office, and that I would be free to do a PR
Then I learnt that this company is valued at >$6 Billion
I don't get it.
People do what they want to do, and might get enjoyment out of it. It can be as simple as that. That MS gets benefit from it, sure, and that's probably even their goal, but to say "I'm not going to do a thing that I enjoy, just because someone else might get benefit from it" is just a spiteful, mercenary way to go through life. If that's what keeps you going though, more power to you.
> No it's not? It's accurate.
Ah, that settles it then I guess. <eye roll>
This is definitely corporate speak for "no guaranteed support, no planned further updates beyond critical security bugs, you are on your own".
Apache Windows when?
More seriously, a desktop UI toolkit is hardly a moat by now, especially a Windows toolkit, Windows having 3-4 very different look-and-feels mixed and shipped with the official distribution.
OTOH security and stability are things that Windows critically depend on to stay on the laptops and desktops in medical, governmental, and financial institutions, and on devices of executives.
This feels like Windows itself is no longer producing enough growth for Microsoft relative to its other efforts. Even the enterprise sales lock-in isn't compelling enough for the cloud/AI-centric future Microsoft envisions. So Microsoft is slowly pulling resources that it can instead invest into Azure and AI and other high-growth business units.
I don't watch Windows too closely. Have there been any other signals of waning investment into Windows? Has Nadella or the other leadership admitted to this?
Hasn't Microsoft also been pulling back from Xbox? IIRC, haven't they been trying to consolidate and use gaming to lionize Windows as a platform? After spending billions on multiple AAA studios? That would seem counter to a Windows pullback strategy. Is this a case of the left hand not talking to the right hand?
> Have there been any other signals of waning investment into Windows?
Wasn't there a story some while back about them crawling the web for PWAs and putting them on the Microsoft Store (or is it Windows store?) to make it into less of a ghost town? And if you go to their official website, browsing vaguely in the direction of UI development, you will see them advertising PWAs as first-class citizens of the Windows ecosystem. I also vaguely remember that Windows+Edge offers special APIs to PWAs for things like file system access and so forth that are unparalleled on other platforms.
I take this push for PWAs (combined with their own lack of dogfooding -- Office is not written with Win UI) as them basically throwing in the towel on Windows-native desktop software (outside of games, maybe).
But, to be fair, native desktop development has seen a lack of investment on all desktop platforms. JavaFX is a ghost town too.
All of that could change, depending on what happens next with Chrome, Bing, and Mozilla. -- The future of each of those seems to be hanging in the balance at the moment.
The web could become a mere implementation detail of the Google monopoly, rather than the open thing it is today. Couple that with a government-level push for digital sovereignty in the E.U. and other places (certainly China). Then, maybe, you will see renewed interest in desktop GUI apps.
On a side note: I think it's amazing what has happened in the open-source space with Rust-based UI frameworks (iced, egui, slint) and COSMIC. The future for cross-platform desktop UI development hasn't looked so bright, maybe since the introduction of Java Swing (was that in the early 00s?)
Qt and VCL/FireMonkey are pretty much alive, as 3rd party .NET like Avalonia and Uno.
Apple is also doing just fine.
It is Microsoft that went south, as Satya apparently sees no value on Windows.
Note the drama on XBox, the console not the Microsoft Games Studios brand, as suffering from the same lack of interest from Microsoft's management.
> Qt and VCL/FireMonkey are pretty much alive, as 3rd party .NET like Avalonia and Uno.
Last time I looked into Uno, it seemed to me more like a mobile-first play (similar to Flutter), but maybe I missed something there.
Qt is still joined at the hips with C++, and I'm finding it hard to imagine that the next generation of developers will go in for C++ over alternatives like Rust. There is a QML/JavaScript story. But why, on earth, would you go there, if history hasn't forced your hand, as it did with the Web.
I hear good things about Avalonia, but: Why throw out the baby with the bathwater and build something "open" within a monopolist's walled garden.
> Apple is also doing just fine.
But "lack of investment" is still not entirely improper as a characterization of how they prioritize desktop versus mobile.
> the next generation of developers will go in for C++ over alternatives like Rust.
The next generation will not choose a language that is even more cumbersome to use and ugly.
> the next generation of developers
are not going to write code.
...the amout of code we've been writing has increased with the progress of technology, just comparing what it was like to write desktop UIs with Delphi in the 90s versus the amount of code it takes for an equivalent app in 2025. Low-code and vibecoded software from the 2020s are going to be the unemployment insurance for all those greybeards who will still know how to code in the 2030s :-)
That’s wishful thinking. Just because things haven’t changed in the past doesn’t mean it won’t change in the future.
I don’t doubt that the amount of code will not reduce, it’ll just be easier and easier to get AI to fix it.
We are still less than two years into widespread use of this technology, and it’s surprising how good it is.
I am a ‘greybeard’ compiler guy and modern LLMs fix compiler bugs better than me, to a large extent. And it keeps slightly getting better every few weeks.
As compiler guy, how do you see direct machine code generation?
I firmly believe having LLMs generate code for current languages is a transition step, just like Assembly devs had to be convinced optimising compilers were generating the same kind of code they would write themselves.
They are not there yet, but the day will come.
LLMs’ design is around novelty and creativity. It was the missing left brain of computation. We should stick to traditional styles of computation for compilation.
Presumably this would need some advancements in formal verification to make sure the results are correct.
But ignoring that for now, LLMs can already do better than modern compilers when it comes to optimizing code.
It has become a neat way of finding opportunities to implement in a compiler.
Give the LLM a compiler generated piece of assembly and it will sometimes spit out a slightly better version. Then you try and figure out if it's possible to implement it in the compiler. This works really well for blind spots in compilers like generating good vectorized code.
Sure, it took only what, 40 years of intensive hardware improvements* for assembly to move to the fringe? And we still reach out to it more often that I would like to because reasons?
Yep, I guess you can train an LLM on a bunch of binaries to get it to mimic a SotA compiler with some accuracy, which may or may not improve over time, but come on. Times where there were free performance increases are gone, and this is not the area where shipping any bullshit real fast will get you any sort of advantage.
* Which are unlikely to happen again in the foreseeable future.
> I think it's amazing what has happened in the open-source space with Rust-based UI frameworks (iced, egui, slint) and COSMIC
Do you mean it is amazing how poor is an experience and how many features, widgets and controls they lack when compared with something like FPC/Lazarus and Qt?
Rust is a nice thing, but its UI ecosystem simply cannot even compare.
Qt and Lazarus had much longer time to grow and mature, and Lazarus can also take a number of pages from the Delphi components book.
What's interesting is the speed at which Rust UI toolkits develop and maybe even mature.
Rust is - by design - antithetical to pretty much every idea of rapid application development paradigm Delphi/VCL and to lesser extent Qt adhere to.
It doesn't matter how many of Rust UI toolkits there are. Consider that there are a lot of Rust game engines, and pretty much zero games written in it, because even C++ gives you better trade-offs in that particular space.
IMO it's not ideal for Rust to have only Rust specific toolkits (at least they need to have other language bindings)
I've tried using Qt (non QML, using rust-qt) last week... initially things were looking fine (despite unsafe everywhere)... but I found quickly that I need unsupported things (deriving from C++ classes), and the project isn't really maintained anymore, so I'll need to so some low level work myself if I want to proceed.
Linux desktop share is up 20-35% in the past year, depending on sourcing. Still only 5-6% total for US at least, but definitely a shift. MS is almost definitely making more money for o365 and Azure than Windows (though I don't follow shareholder releases).
From google's ai result Windows ($25b) is 1/2 the size of Office (54.9b) and 1/3 the size of Azure ($75b) in terms of revenue.
Windows is definitely not the cash cow for Microsoft it once was.
Usually I get why companies release their UI frameworks. I've strongly considered using Atlassian's and AWS's frameworks in the past to build web apps because if it's good enough for Jira/AWS, it's probably good enough for my B2B saas app.
But I personally don't know why anyone would reach for this framework. Maybe if you're building a Windows app and you want a very consistent look and for your app to feel "native", but aren't there better options out there for doing this already?
Yeah, WinUI has been a disaster.
> hoping passionate losers will contribute.
Offloading the work to their victims. Maybe they will even make it usable again.
No one in Windows development community cares about WinUI, other than those with sunken costs that bought into the WinRT/UWP dream and now are stuck with a dead technology.
Too many burned bridges since Windows 8 came out.
If anything, this is Microsoft confirmation that they are unwilling to fix all the broken issues, and hoping the community will somehow still care.
This. Also DevExpress and Progress Telerik do not invest into their WinUI Controls at all, and that’s a sign they don’t buy into WinUI neither.
WinForms and WPF are currently the only viable frameworks for Line of Business application. I have yet to see a WinUI3 application in the wild.
Very true. We just developed a brand new LOB desktop app and settled on sticking with WPF. WinUI has been dead for years imo.
On a side note, I still love WPF after working in it for 10 years. Maybe it's just familiarity, and it's a little verbose at times, but man it's a great framework when you know that you're doing.
We also settled on WPF for a new LOB Desktop application, this validates the decision. If you combine WPF with the CommunityToolkit MVVM, it’s a very nice framework to develop with.
DevExpress supported WinUI for a little while but decided to abandon support.
One of the biggest problems with WinUI compared to WPF is that DependencyProroperty is implemented as native code, so for .NET developers, there is a huge performance penalty getting or setting any property on control.
https://github.com/microsoft/microsoft-ui-xaml/issues/1633#i...
It's like saying "a UI toolkit for Rust" and then making all the Rust functions call into a java codebase haha.
I read through that github issue a few years ago and with a mix of surprise and disgust knew that I would not be learning WinUI.
Honestly at this point who would seriously use any Microsoft UI framework? They've abandoned 100% of their previous UI frameworks unfinished when they get distracted by a new, shinier framework.
Why use a busted incomplete framework missing basic features when there's entire ecosystems of open source cross-platform frameworks being actively maintained and which actually have all the features you need?
Really this is just another UWP destined to be forgotten and scorned.
Forms and WPF are still reasonable options.
GDI32 only works as well as it ever did.
WPF uses DirectX 9, in any case it is good enough for many businesses use cases.
UWP never had feature parity with neither Forms nor WPF, so already it has a hard sell to businesses.
Microsoft marketing always cool about cool experiences and design, without the actual meat in features.
That is why there were API reboots between Windows 8, Windows 8.1 and Windows 10.
Then comes along Project Reunion, which is supposed to properly unify Win32 and UWP execution models, instead a few months in, they do yet another reboot, that to this day is a shadow of UWP features, endless bugs, and a team that everyone thought was also part of the layoffs, has they were radio silence for months.
GDI32 you say? At least it works.
Forms is still my go-to for quick one-off and disposable programs. But maily because it's simple and I have a big library of controls and extensions I've built over years to make Forms more complete.
I've also tried WPF, but it's missing very basic controls that I just don't feel like recreating.
Can you give examples? I've been looking for some open-source project ideas...
I dunno, they are still doing doing bug fixes to Winforms.
I already lost count how many UI frameworks are in windows. It looks like complete chaos and mess.
I really wonder what they expect from open-sourcing it. Just to pretend how open they are? Or is there any real benefit to developers who target windows?
WinUI is an evolution of UWP which is an evolution of WinRT. WinUI has been around for years.
MAUI is not exactly a competing product and is more about enabling cross platform UI development. Different intent.
WinUI is actually ok tech. It’s evolved over the years through a few iterations, now on WinUI 3.
Im mostly with you though. Until they rebuild the entire OS in it, including all of the administrative controls and tools, I don’t trust the longevity.
They already do, though. The big UI refresh in Win10 is all XAML, and the new Win11 taskbar (the one we all hate) is now a totally normal XAML app.
WinUI 3's big changes (to get a 3.0 version number) is not with the XAML stack itself, but its new ability to be called by unmanaged apps as a normal UI toolkit, so it can finally be used by all apps. No more using Shell UI like we're writing Win 3.1 apps.
And yes, some stuff in Win11 still isn't WinUI, which is kind of annoying, but some of those dialogs hidden away in Windows are at least 20 years old, and probably would need to be entirely rewritten, not merely have their UI's updated.
Also, fun fact: The Win8/10 taskbar's code predates Avalon (the prototype/codename for WPF), and trying to change/fix it at all usually ended up breaking it. It's one of the few binaries on Windows that would not be recompiled to build a new release image in fear of breaking it. Rewriting the taskbar made sense, GETTING RID OF SMALL MODE DID NOT, GODDAMNIT MICROSOFT.
Lets not forget that after five years Project Reunion was announced, WinUI 3.0 is yet to achieve feature parity with the Visual Studio experience developing C# or C++ applications, or UWP components features.
The WinUI map component is a Webview2 instead of proper native component, Win2D is only a subset of the UWP one, ink is yet to be migrated, and lots of other issues.
Github repos are filled with thousands of issues, and they already did a cleanup a year ago where they simply closed enough tickets to bring it under 2000.
> The Win8/10 taskbar's code predates Avalon (the prototype/codename for WPF), and trying to change/fix it at all usually ended up breaking it. It's one of the few binaries on Windows that would not be recompiled to build a new release image in fear of breaking it.
The taskbar that underwent a major redesign in Windows 7 (released after WPF)? Also, that binary is explorer.exe, surely it got rebuilt quite often for new ads. features, and fixes?
The taskbar exists in its own timeline, apparently. (Ex-)Microsoft engineers working in the UI/UX team for Windows itself have spoken about how much of a nightmare dealing with that thing was.
Also, the taskbar technically exists as a dll, not an exe, explorer.exe would link to it and run it if it was being ran as the UI shell. This is now split out to its own exe (Shell Experience Host iirc), and explorer.exe is now only File Explorer.
Since it exists as a dll, btw, this is how the Win10 taskbar injectors work, they just call the dll (which still ships in Win11) instead of letting Shell Experience Host do the Win11 thing.
The sane way of handling this, btw, is just use Shell Experience Host injectors to get the desired behaviors, such as using Windhawk and use m417z's taskbar height and icon size plugin (the third most popular Windhawk plugin); to match Vista/7 era small, set to height 32, icon size 16, taskbar button width 28.
The detail missing is that those DLLs are COM servers, so it isn't really doing injection rather the standard COM RPC.
> And yes, some stuff in Win11 still isn't WinUI, which is kind of annoying, but some of those dialogs hidden away in Windows are at least 20 years old, and probably would need to be entirely rewritten, not merely have their UI's updated.
And this is hard to do. These dialogs often are _dynamic_, with third-party settings rendered as ActiveX controls.
Thanks for the informative comment.
> small mode
I recently noticed that they introduced an option for small icons. Not that it changes much, as the height of the bar stays the same, but hey. Personally I've been fine since they added back the option not to combine buttons unless full.
Still writing win32 stuff like it’s 1995 here. We have bits of ATL/MFC hanging out which are throughly abandoned.
I don’t trust WinUI at all.
I was surprised, when I spoke to a former colleague, to find that an internal tool I wrote 25 years ago is still being maintained. Win32 as well.
Software that solves an actual problem has the tendency to stick around, no matter how much time elapsed.
Just remember, cobol is still in active use, today
I was going to ask about Win32. I haven't had to do it in a while, but if I had to write a desktop app in windows, that would be what I would reach for. It's still supported... is their any indication that it won't be for many years to come?
Also, it looks better, in my humble opinion. It's probably lacking features that I'm uninterested in.
Still works fine. High DPI stuff is a dick but that’s about it.
Eh, high dpi has always been a mess given the hacks adobe and windows assumed in the past.
I implemented a WPF control to look like a winform control and ran into a issue of the vintage control looking perfect at an odd (9px by 9px) native pixel size that gets distorted by WPF when rendered at different scaling.
Visual Studio's + sign expanders have had this bug for years that was finally fixed by changing to a chevron shape in 2022. Clearly this problem bugged someone on the dev team hence the fix. =)
MFC support is still in the latest Visual Studio, and it looks like ATL as well.
I was surprised to see ATL/MFC received security updates such as Spectre mitigation. So there is still some support for these 30 year old components.
That’s only because half office hangs off it. If it didn’t they wouldn’t have just left it.
Which so much better tooling than XAML C++ with C++/WinRT, it is a tragedy.
It’s in there because it’s a cockroach. It is throughly abandoned though, frozen in time.
You can observe this looking at the state of old outlook.
The state is wonderful, versus the new kind of Electron junk.
I think only WinUI2 (deprecated) is an evolution of UWP (uses WinRT APIs). WinUI3 is something different.
WinUI 3 still supports WinRT. It ALSO supports more. It's an evolution of WinUI 2, not just a simple version bump, but also not a completely new tech. It's probably a closer evolution to go from WinUI 2 to 3 than it was to go from Angular 1 to 2.
https://learn.microsoft.com/en-us/windows/apps/winui/winui3/...
https://learn.microsoft.com/en-us/windows/apps/develop/platf...
> WinUI 3 still supports WinRT.
I think this is completely independent. You can simply use WinRT APIs, because Win32 Apps can use them now. WinUI3 apps are win32 apps.
> but also not a completely new tech.
Not sure about this. UWP APIs work out of the box. For WinUI3 you need the Windows App SDK, and it is much slower and heavy than UWP (out of curiosity I created a very simple app and it was fast just a few dozen kbs big)
Hardly an evolution, that is how it is sold, reality is something else, trailing behind UWP with half the tooling.
WinRT came out of UWP I think. UWP was their first attempt to move beyond .NET
You have that backwards. WinRT is the managed languages runtime for Windows, introduced in Win8. Its sort of the replacement for COM/OLE but also defines the ABI dialect in a way that allows managed languages to call unmanaged code without an FFI penalty.
UWP is built on WinRT, and acts as a fully managed app container, similarly to how phone apps exist on your phone. It allows WinRT apps to be deployed to any Microsoft platform, Windows, XBox, Windows Phone, etc, but also Android and iOS, and also as PWA, and are guaranteed to run identically on any of those platforms. UWP apps must be written a fully managed language that runs on the CLR (ex: C# runs on the CLR, but C++/WinRT does not). UWP also uses the second generation of WinUI-family XAML UIs, which means all UWP apps use completely native UIs, instead of slow non-native Javascript shit in a web canvas.
The WinUI family of XAML UIs started with WPF, and a slightly incompatible version of it also appeared in Silverlight (WPF = WinUI 1.0), then was brought to UWP (= WinUI 2.0), and is now its own stand alone thing that any app can use, managed or not, as 3.0.
WinRT is not an attempt to move beyond .NET, instead it is their way of allowing .NET to natively call code, and make .NET languages first class in Windows.
Yeah but I think when it was introduced it wasn't a thing you could use separate from the rest of UWP. What changed in Win10 was you could use WinRT APIs from regular Win32 apps too. They started breaking UWP up into independent pieces.
Or not. I haven't thought about this stuff for years. Definitely possible I forgot the ordering of things.
UWP most certainly did not exist when WinRT came into existence. WinRT, itself, is also more or less the third version of this attempt: Managed C++ came first, then C++/CLI, then the WinRT era and C++/CX.
C++/CLI and C++/CX are semi-managed, you only need to use the extensions when interacting with WinRT, although you're free to write your entire app in that dialect.
WinRT is not the same thing as managed .NET code. There is no requirement that a UWP is .NET. There are many examples of unmanaged C++ UWPs, including the open source Windows Terminal.
WinRT is a mechanism to express APIs in a way that is amenable to cross-language usage. It is built on top of COM, and is not a replacement for COM.
Windows Terminal is sort of a clusterfuck of multiple programs, only one of which is a UWP program (the actual visible WinUI shell that runs) which you're not allowed to do as a UWP-era Microsoft Store app, but they got internal permission to do that.
Its now a WinUI 3.x program, apparently, and now the Store no longer requires UWP programs, so WT now only needs to make the "ships inside of Windows release images" guys happy, which apparently is harder than making the UWP-era Store guys happy.
WinRT was windows 8. Remember the ARM-powered Surface RT had the same branding?
UWP came along in windows 10.
WinRT was introduced with windows 8, the WinRT APIs still exist in Windows 11.
Yep! I was implying it was the same timeframe as windows 8, but I see where my wording could easily have been taken literally.
I genuinely thought a lot of these sibling comments were satire at first! The acronyms, the lengthy and confused explanations of versions, frameworks etc:
> WinUI is an evolution of UWP which is an evolution of WinRT
> WinUI 3
> WinUI 3 still supports WinRT
> XAML
> Shell UI
> Avalon
> WPF
> WPF = WinUI 1.0
> Project Reunion
> UWP
> Win2D
> ATL/MFC
> Just go for MFC FTW
> wxWidgets as it is kinda MFC-y
> Or go with Qt
> MFC/Win32 + XAML Islands
I can produce similar lists on other domains in the industry, it is a matter of being part of the ecosystem.
OK but stick to the UI framework domain for a fairer comparison. (We all know tech has acronyms)
So what you got?
Replace the Microsoft specific terms of those examples with,
- kexts
- Cocoa
- UIKit
- Carbon
- SwiftUI
- MSL
- XPC
- ARC
- plist
- CoreData
- CloudKit
- SpriteKit
- SceneKit
- RealityKit
Do you think anyone that never bothered with Apple's developer ecosystem has any clue of those names are for?
And that is a very small overview.
They probably started something new and shiny (Now with AI!) and want to get rid of the old baggage without causing too much of a user revolt (all dozens of them) ;)
https://xkcd.com/927/
There are three UI frameworks in Windows, and only two actively used/developed.
All the other "countless" frameworks are iterations of one of two lines: Win32/Native (WinAPI, MFC, WinRT, WinUI3, etc) and WPF/Managed (Avalon, WinUI2-3, etc). WinUI3 exists to bridge the gap.
Just go for MFC FTW, it is in feature freeze but I will last probably for the next 20 years yet.
You could also go for wxWidgets as it is kinda MFC-y but better and cross-platform, though like MFC you can combine it with Win32 API code (almost) seamlessly.
Or go with Qt, though that doesn't use native controls.
QT uses native controls/widgets, it just polyfills when there is no good native option or if you use custom styling.
No, it implements its own functionality. As an example consider one of the most basic controls which is available pretty much everywhere (i.e. no need for polyfill), the push button: the source code[0] for QPushButton clearly implements the behavior itself, it does not rely on any native button.
Compare with wxWidgets' equivalent to QPushButton, wxButton, where there is a backend-specific header[1] and implementation[2] where 99% of the wxButton functionality is there (there is a `btncmn.cpp` under `common` that is shared across backends but that has very little code itself).
[0] https://github.com/qt/qtbase/blob/dev/src/widgets/widgets/qp...
[1] https://github.com/wxWidgets/wxWidgets/blob/master/include/w...
[2] https://github.com/wxWidgets/wxWidgets/blob/master/src/msw/b...
I don't even need to dig through the paint loop. I've written QT software, and you can do it yourself.
Add a QButton with no styling, you will have a standard button widget in macOS and Windows (or a KDE styled one in Linux). Now have an event that modifies the button in some way (adding a CSS style to it is common, but also changing the sizing manually will usually do it) and the style of the button will immediately change to a QT-style one. The difference is so obvious in Windows and macOS, that it would be laughable to claim otherwise; but you can also just ask the QT Group devs or the community, they've reiterated/experienced this behavior plenty of times.
MFC/Win32 + XAML Islands (through the Windows App SDK) is a pretty nice combination for stability and access to new features.
Maybe people can cannibalize some of the rendering code and extrapolate the controls to a better class library than they already have. Like a kind of winforms but using modern rendering APIs. I know you already can create such controls but they often end up being very verbose and just look like xaml but in C#
> Maybe people can cannibalize some of the rendering code and extrapolate the controls to a better class library than they already have.
This just sounds like another Microsoft UI stack
I kind of wish Microsoft would just continue development of WPF. I've used it for years for various projects, and there is a learning curve but I've since enjoyed working with it. XAML, data bindings, ViewModels... all of it I actually like. But, WPF needs a few improvements to really make it perfect. I tried several of Microsoft's newer frameworks and the open source ones (Avalonia, Uno), but I either couldn't get the sample projects to even build successfully on my machine, or I never got comfortable with development workflow, and went back to what I know.
My big idea to fix WPF is to rebuild the data binding system to use the .NET compile-time code generation feature instead of run-time reflection. I think that would solve a lot of problems. For one, projects could do an actual AOT build of their applications (right now, you either need to rely on an installed .NET runtime or "publish" the project with a lot of .NET libraries included for self-extract, bloating the final file size). Code generation would probably improve performance quite a bit too, maybe open up the possibility to compile for cross-platform, introduce type safety for XAML bindings (rather than getting vague runtime binding errors), remove the need for so much class scaffolding, etc... I've thought about starting an open source project to do it myself, but seems like a pretty big task and I would essentially be starting a project to help with my other project which I already don't have enough time to work on...
Your second paragraph sounds like you're describing Avalonia. Avalonia has AOT, compile-time binding errors and cross-platform support. Maybe there have been some updates since you last tried it? I'm not very familiar with Avalonia or WPF though so maybe there's more to it than that.
[0]: https://docs.avaloniaui.net/docs/basics/data/data-binding/co...
[1]: https://github.com/kekekeks/XamlX
Thanks, yes I'll probably have to give it another try some day. I might be confusing Avalonia and Uno, but I think I first attempted it a couple years ago, and then again last year. I remember spending a whole weekend trying to get it running but wasn't having success. Also, I was a bit turned off by how heavy the development environment was. I had to download and install a tool, then that installed more build tools and packages, and then there was also a "recommended" VS Code extension. With WPF, I've gotten used to writing XAML without a designer, so I can get by with just VSCode, the C# extension, and the .NET CLI.
It would also allow for assembly trimming, which would be a huge boon if you are trying to do a self contained deployment. Right now you either do framework dependant or have like a 200MB+ deployment.
Last I evaluated it WinUI3 was a terrible developer experience. The application had to be literally installed on the system to even debug it, which means you end up with a large number of useless start menu entries, not to mention registry entries and such. Another thing was that the example programs crashed when I clicked on a button.
All I want is something simple to work with to make applications for Windows, and so far I'm still using Win32 with WTL.
> The application had to be literally installed on the system to even debug it
I think that's because you chose "packaged" application, these apps need to be installed so that capabilities are handled correctly.
To be fair, macOS has the same issue, although they won't show in Launchpad, they still can be indexed by Spotlight.
I want to be able to create self-contained GUI application that are relatively small and can be just copied and run on another computer. Installation should be an option, not a requirement. From my evaluation, WinUI3 doesn't offer that.
WinUI3 offers that, it's just not the recommended way of building+deploying.
What you are looking for is "unpackaged, framework-dependent". This will build an application like an old-school WinAPI executable, one which expects the relevant DLLs to be installed on the host system (or distributed in the same directory).
Size of hello world app that includes relevant DLLs is over 150 MB. Which is almost like including a web engine.
What I'm looking for is old-school AOT compiled executable where UI framework dependency is statically compiled and trimmed.
I know that these days a lot of game and application developers just render a custom UI to a 2d or 3d rendering device. I've been looking for a UI framework like that as I don't want to write my own, but so far most of them have been either web based or require their own mini-language and compilation step.
You mean just self-drawn controls or also immediate mode GUI?
I haven't tried it, but there's ImGui.NET (or a bit easier ImGui.Forms) immediate mode GUI. Downside is no accessibility.
Avalonia uses Skia to draw controls, although it primarily uses XAML to define UI.
The issue with self-drawn controls is that creating such framework is a lot of work, so smaller projects often have issues with things like accessibility.
EDIT: Just did hello world test with ImGui.Forms. Self-contained AOT compiled result was 7 MB.
Within game development it is predominantly immediate mode as it more closely resembles the main game loop. With other applications I'm not sure, some I know just use ImGui and run an immediate mode UI, others would be retained mode.
In my case I'm looking for a C++ UI library, not .Net. Also not a fan of immediate mode GUIs either, so would want something more traditional.
I don't want to write my own, so will probably try Qt again, but might switch to a custom library in time.
I did try to develop an unpackaged test application but I gave up trying to implement it and just went with Win32 instead as I wanted make something rather than messing around with a UI framework.
These days if I were to switch from Win32 I might try some custom rendered framework which a lot of apps seem to use, or Qt.
Many here point out, and rightly so, what an abominable mess UI frameworks have been on Windows for quite a while.
That being said, I'd point out that the situation has been (and still is) unfortunately far from satisfying on the cross-platform and open-source side as well: GTK has also been a mess for quite long; Qt has - aside from the bloat - an increasingly repulsive licensing model for professional use (too bad the hopes from the early Nokia adoption days got smashed by MS mole Elop and the later owners of Qt), …
… and while there are some nice solutions for specific niches (e.g. Dear Imgui), the overall situation in general-purpose cross-platform native UI frameworks looks quite bleak and unsatisfying imho. There's a huge gaping opening for a permissive-licensed open-source cross-platform UI/Widgets framework that compiles to native, has a good set of widgets and gives access to Vulkan contexts for 3D rendering.
People like to hate on Electron and React Native (because web bad!!!!) but tend to fail to realize there’s basically no alternative if you want to be platform flexible.
Microsoft could’ve made a real impact in this area with a great advantage to themselves, but both efforts I’m aware of were so half hearted they could never have worked out.
I hope this leads to having a native vertical taskbar, which has been absent in W11 despite being a taskbar feature dating back as early as Windows 98.
Third-party tools have tried to reimplement it but it's either been by bastardizing the native W11 horizontal taskbar to be vertical (eg: Windhawk) or just restoring the old W10 taskbar code (eg: StartAllBack).
How will making the UI framework open source lead to taskbar changes? For third party contributions in this area, they need to open source the taskbar, not the UI framework.
Nit-pick: Windows 95, actually. The vertical taskbar was an option in its very first version.
> Nit-pick: Windows 95, actually. The vertical taskbar was an option in its very first version.
Yes, but this was when computer were less capable. Now, you are not able to choose colors or fonts, the taskbar is fixed and the programs need at least 2 seconds to start. Terrific evolution.
The taskbar is a feature of explorer.exe.
The news being discussed is not about explorer being open sourced.
Is the Windows team even using WinUI for the native Win11 desktop UI? ;)
The Start Menu is apparently a React Native app, so I'm going to hazard a guess and just assume WinUI is built on top of React, and that the Start Menu at least is thus indeed built with WinUI. But it's also clear that some other parts aren't, so who knows what's what. I'm sure there are folks who spent time reverse engineering it all though who do.
The start menu is not a React Native app, but it's actually even worse. Only the recommended section (which is basically recently used files - plus probably advertisements in some scenarios) is. The rest of the start menu is WinUI, to my knowledge.
I cannot stand the latency using a local app. Same with rendering views of local file systems. Frontend reactivity as the expense of responsive performance is the problem with modern user interfaces in my opinion.
Like I’m searching for an installed app. I don’t need news articles about that and never expected a file system ui to be a web portal.
WinUI is its own thing. The React Native stuff just shows that even Windows developers don't want to use WinUI.
> The Start Menu is apparently a React Native app
What's the source for this?
https://news.ycombinator.com/item?id=44124688
So the source is a twitter meme?
Featuring a screencap of an MS engineer's talk from some React conference, yes.
The talk in question showcases a widget in the start menu that's using react native. Apart from the meme itself, I have not found indication of the start menu being a react native app. Every time this comes up it seems to always lead back to that meme.
That'd be correct. See also the top comment in that thread that (evidently after I originally read that thread) also explained exactly this, and a sibling comment in this thread tree that also did so (more than two hours after I posted my original comment, meaning I was unable to amend or delete it).
the talk is linked as well: https://news.ycombinator.com/item?id=44125217
It's confusing what exactly 'WinUI' is, but does Explorer looks WinUI-ish. Parts of it at least.
Explorer uses XAML Islands. Parts of it are WinUI, while the rest is still Win32.
For Windows UIs I've been getting into Win32/GDI/DirectDraw/etc.
Tools like CsWin32 and modern C# (ref returns) make working with these APIs a lot more approachable today. It used to be the case that you had to create a nasty C++ project to do any of this. Now you can just list the methods you need access to in your nativemethods.txt file and the codegen takes care of the rest.
Win32 is a lot lower level than other things you'd typically consider to be a "UI framework", but the important tradeoff is that it is also a lot harder for Microsoft to remove or screw with in any meaningful way. I cannot come up with something that has been more stable than these APIs. The web doesn't even come close if we are looking at the same timescales.
You still need C++ in many places, because of the COM rulez attitude within Windows team.
Windows Runtime Components was a lost opportunity to level up the play field for .NET.
As such, if you want to do something like a shell extension, or context menu extension, it is C++ as always, or having your little C++ stub that calls out into a .NET process.
Yeah this entire discussion about these high level framework would be cooler to discuss the lower level APIs. As far as I know, the entirety of Windows rendering stack is built on either GDI or DirectX? With Win32 even being built on GDI?
It would be much cooler to discuss building a ground-up Windows UI stack as close to metal as possible which I guess would necessarily be using DirectX...
> Win32 is a lot lower level than other things you'd typically consider to be a "UI framework",
From a user's perspective, Win32 is high level. No other Windows toolkit is able to draw buttons or scrollbars properly.
I think windows needs a community effort to create an actually good framework for native development on windows. Unfortunately I just dont think such a community is big enough.
What I miss about UI development on Windows is the commitment to help me create applications that seamless blend in, and look like something that Microsoft themselves had created. The introduction of web technologies has allowed the visual language of the Windows experience to fragment. Not just because Microsoft hasn't kept older applications up to date (like the Control Panel), but also because the modern tooling isn't providing the libraries to match their style.
From memory, this first started during Vista, where 3rd party devs just weren't able to match the translucency of the new Aero UI. And the lack of articles being published by the MSDN group with titles like "Here's how you add this new feature to your app" with examples in C, C# and Visual Basic.
Seasons may come and go, but one thing will never change.
Windows and an absolutely baffling array of UI frameworks with various pitfalls, uncertain futures, and no clear winners.
(honorable mention to WinForms though.)
My analogy is every Microsoft UI framework was almost completed in the sense of someone being almost pregnant.
A framework that has just one show-stopper missing feature or problem is... unusable. You can't embark on a large, complex application development journey if you even suspect that you'll be painted into a corner.
For example, many of WPF-derived frameworks had atrocious performance, with fundamental mistakes in their design that made them incompatible with list virtualization. It wasn't until they had to eat their own dogfood and use WPF for Visual Studio that they started fixing these issues.
Win UI 3 meanwhile dropped all support for HDR, wide-gamut, etc... going backwards to SDR sRGB only in an era where all mobile phone manufacturers were starting to standardise on OLED HDR displays. The logic behind this decision? Microsoft wanted a UI framework that is "mobile compatible"!
I just have to wonder.. why after decades can Microsoft not get this right? I’d love to hear insider stories about what’s going on here.
From watching the community calls, long after I stopped caring, management doesn't seem to care to actually hire people that have Windows development background, many times they would ignore community questions or don't get where they were coming from.
And I still give them points for trying, a rarity among the tech giants.
As someone reading the comments here and never made a real Windows app outside of a visual basic hello world a pretty long time ago. Why doesn't Microsoft just stop making these? They already own GitHub and vscode so why not just admit that electron/typescript is the Windows UI framework now?
Because I don’t want to run even more browsers simultaneously than I already am.
Because Teams is slow again
I wonder how much longer Microsoft stays committed to Windows as a whole. Windows is less than 10% of the company, users are migrating to phones, tablets, and Chromebooks (all of which can run Office), and with .NET on Linux, Windows servers are making less sense. It's a shrinking market.
It is not really about Windows anymore.
But Windows is where most people run Office, including all the back-end implications like Exchange and all the "Power" stuff.
It is a lot more than 10% of Microsoft revenue really (just not directly).
I agree though that they do not really care about the platform anymore. In fact, they care less and less every day about everything that is not AI on Azure.
I would guess Fortune 500 still runs on desktop windows? (Don’t know but this is just my guess).
Yep. Most any regular company is majority windows. It’s not exciting but why would they bother throwing that influence away.
Yup! Boring technologies are great.
Yes, but for how long? At some point, IT realizes everything's in the browser, and general-purpose computing is a security risk.
It will be a long time before Microsoft Exchange, Active Directory, and Excel are not important enough to enterprises to justify the Microsoft subscription. And Visual Studio is still a huge part of the Microsoft capture strategy.
Microsoft is working on separating all of the above from Windows without losing too much control but it is going to take a while.
> Microsoft Exchange, Active Directory, and Excel
Exchange should be SaaS for 99% of businesses. There's a web version of Excel that's probably OK-ish. I'm not sure what Active Directory looks like in a world without Windows.
Not every company is a startup that only needs gmail and Figma, though. The ecosystem of LoB apps and device controller software is ginormous, and not going away any time soon.
Meanwhile a single developer makes performant native UIs in his first native program by actually learning how the OS and winapi work
https://m.youtube.com/watch?v=bUOOaXf9qIM
Microsoft has a long history of releasing numerous UI frameworks: VB, MFC, WTL, Silverlight, WPF, WinForms, and others. Yet despite this abundance, many of the core components Microsoft used in its own applications were never available to developers. They rarely ate their own dog food, and desktop UI development relied on third party components. For the past two decades, native desktop UIs have steadily declined in favor of web-based components, so it's unclear what the real benefit of another native framework would be today.
I won’t benefit from this. At the same time, I cannot see a single bad thing about it, so I’m surprised about all the negative energy.
The "bad thing" is that it's effectively getting abandoned, open sourcing it won't make any difference.
It's not like external contributions will suddenly turn it into something usable, and they'll just have a skeleton crew maintaining it, like they do WinForms and WPF.
People are tired of Microsoft and their ever growing graveyard of ill thought out, half-baked, "native" UI frameworks.
Native UI is effectively dead outside of Apple’s platforms, and even there it’s hanging on for dear life. HTML, CSS and JavaScript won the cross platform toolkit battle.
Sadly yes. And all the platforms are to blame. Microsoft and their 1000 half-working frameworks made writing a wrapper that was any better than wxWidgets impossible.
But also Apple "totally not deprecating" AppKit and pushing everyone to the mess that is SwiftUI, Gnome breaking backwards compatibility as a sport, and Qt messing around with QML, meant "native UI" became quicksand.
Even going HTML, CSS and JavaScript wouldn't be too bad if the browser engines provided by the OSes were any good, but it took Microsoft giving up and switching to rebranded Chromium as a browser for Windows to provide a usable one in WebView 2.
WebKitGTK is also terrible compared to the macOS version of WebKit, which hurts projects like Wails and Tauri. So everyone bundles a freaking copy of Chromium with their applications.
I should have studied mechanical engineering.
On the Apple side of things, AppKit and UIKit work as well as they always did, and they’ve been less pushy about SwiftUI lately probably because they realized that the old toolkits aren’t going away any time soon.
For Qt, the hard-coupling of C++ or Python for Widgets and Quick being JS-centric haven’t done it any favors. C++ and Python are fine, but not everybody wants to write either, and most people interested in writing JS are going to gravitate towards front end web stacks over anything else.
I think that for a cross platform desktop UI toolkit to see any degree of long term success, a high degree of bindability is non-optional even if it’s most capable when used with its native language. The toolkit needs to meet developers where they are, and that means being usable in the language of their choice.
The most important thing the web standards get right is their insistence on never ever breaking backwards compatibility. HTML, CSS, and JS accumulate a lot of cruft, but they do move forward into the future without leaving anyone behind.
Flash was a web standard, albeit a closed one. Good luck opening a site with it today.
The web has continually added and removed features. It is absolutely not perfectly backward compatible. It’s not even close.
Flash was never a web standard, what are you talking about? It was a commercial browser plugin developed by Macromedia and later Adobe.
True, and at one point it had an installed base of like 90% of all browsers, and was incredibly common on all kinds of websites.
I said it was a closed standard, and I stand by that.
> The most important thing the web standards get right is their insistence on never ever breaking backwards compatibility. HTML, CSS, and JS accumulate a lot of cruft, but they do move forward into the future without leaving anyone behind.
This is the comment you originally responded to. Flash never had anything to do with web standards, which do indeed strive for backwards compatibility, it's why that classic space jam website still works.
The comment was not that the "web" as a whole strives for backwards compatibility. If that were the case we would also be running ActiveX controls and Java web applets.
They only won because developers stopped giving a shit about anything except their own ease of work. There's no such thing as a good UI built with web tech, so anyone who cares about the user's experience will use a native toolkit despite the difficulties. But very few do, turns out.
That's extremely uncharitable. I think slack and discord have pretty great UIs, and that's all "web tech." Figma just IPO'd at a $58 billion valuation - is that not proof their UI is good? If it wasn't good, no one would use it. VS code became the preeminent text editor and IDE over the last decade - all web tech.
Not sure about others but VS Code and Discord are slow and eating too much RAM for what they do.
Does that matter? The UI is great. Hordes of people agree.
Doing alright on Android as well, and third party on Windows for those willing to go with Qt,VCL, FireMonkey, Avalonia,...
Microsoft is the bad one here, unfortunately.
Too many reboots since WinRT was introduced in Windows 8, many battle scars and lost wars, only people without WinRT experience can think anything positive about WinUI.
Microsoft probably has no interest in WinUI anymore and intends to replace it with WebView, which is why they're making it 'open source.' Microsoft's open source is literally a graveyard.
Open source the mess and hope someone else picks up the tab.
I hope nobody contributes to this, MSFT is a multi billion dollar company and they can afford to fix it themselves.
If only they'd open source Windows Explorer and the taskbar/start menu, rather than resisting peoples attempts to customise them through other hackery.
Confused, the Win 11 UI framework, Electron, is already open source.
They need to feed it to all the LLMs to get help keeping it from falling apart.
They could go back to Win32 + WinForms and everything would be fine.
So, they gonna abandon it soon?
It was abandoned from birth.
First time?
Could microsoft still build windows today?
The answer to this question depends on the knowledge and quality of engineers working on the kernel and the overall Executive. These continue to evolve with more advanced technologies, like VBS or the future usermode endpoints for EDR and possibly anti-cheat, pushing those out of the kernel, which presumably requires kernel work.
David Cutler is still there but working on other stuff, last I understood.
That does lead to the question of 'would they do it the same way and/or follow the NT OS/2 spec' to get a functionally identical Windows today.
Ironically, from his latest interview, he mentioned working on Linux on Azure racks that are actually repurposed XBoxes.
Cutler working on a UNIX-like system? Did hell just froze over and I missed it?
See the last 5 minutes from his interview, done last year on Dave's garage YouTube channel.
Even for Windows-only GUI software, it is much safer and sane to use cross-platform frameworks such as wxWidgets and Qt Widgets.
And if you're on .NET, something like Avalonia.
What "UI framework"? Windows is a Frankenstein's monster of different UI elements.
I haven't used Windows for a long time but I'm sure they still have the moricons.dll of Windows 3.1 somewhere.
Rendering engine + set of native components + APIs to make your own components.
Windows definitely shot themselves in a foot with building multiple renderers while building them on top of each other.
If abandonware of Win11 UI means Win12 then maybe I like that direction :)
On the other hand it it is smpossibility that MS is working on something better, right ? They probably plan some worse UI prison...
ASCI was [business] enforced by some EU project, POSIX - the same and probably there are more examples of such things so we just need some sanity __OS with UI__ project from someone...
"We laid off most of the team despite record profits and need free labor to maintain what remains"
Actual - and rather different - title (as borne out by reading the article):
Microsoft is taking steps to open-sourcing Windows 11 user interface framework
open-sourcing it so they can get free labor.
winui3 was abandoned the moment it was conceived.
I feel like this should be called "open outsourcing"
I really hope they do and the rendering engine is decently decoupled, I'll give a try building a framework on top of it.
I wish all platforms gave access to their rendering engine similar to DOM on the web, imo SwiftUI/WinUI (or WPF, but they are very similar) are not that good.
Haven't built anything native on Linux, though, no idea how good those are.
What do you mean by access? APIs to program against or fully open sourcing the rendering engine? Because you can mix SwiftUI with a few different rendering frameworks at varying abstraction levels that it itself renders to (AppKit, UIKit, Core Graphics, Metal etc.)
Basically I want an API available to build my own SwiftUI. Definitely not on the Core Graphics level :)
But good point, I actually think AppKit might be a good abstraction level. I'll play with it a bit and see if I can abstract it behind a good component model.
What's wrong with Skia? Chrome, Firefox, and OpenOffice all use it, and it works on Windows, Linux, MacOS, and Android.
It is a ton of C++ for what is essentially something that an OS like Windows/MacOS/Android/iOS or the browser would provide anyway. Apps that use it ship with a substantial minimum amount of bloat, e.g. Flutter for web.
Nothing wrong with it, just want something a bit higher level and ideally with at least some native components/styles.
Cool, I always wanted to have a non-movable toolbar in all my software, I miss 1997.
It’s kind of wild how much less flexible and powerful software has become, while (it seems) not becoming much less buggy either.
I suppose this coincides with software development becoming much cheaper.
For a second there I read it as "Windows 3.11's UI" and it got me really nostalgic. I wonder if it's been long enough that it's ready to be hip again now.
I want to spam them with pull requests every day to remove onedrive. (sarcasm)
What we wanted: Win7 UI open source
What we got: Win11 dumpster fire, free for everyone to fix
I would be fine with really any version as open source or at least one that was free-as-in-free-beer to make it possible to maintain virtual machines running old software without having to rely on dodgy downloaded versions... I'd even pay Microsoft something reasonable if they put them up on GOG or some similar site for a few $.
Yep. Win7, especially the UI, really was the peak. It was also more stable and faster than 10/11.
Having had to setup some Windows systems for a client of mine, all I can say it that they've long abandoned all hope of ever shipping a somewhat coherent UI. It's just not going to happen. Lipstick upon lipstick on frankensteined pigs just aren't going to work. At some point they'll have to start from clean slate or else we'll see 1990s ui remnants leaking through into the 2040s. It's just not going to happen until they start bleeding market share on the desktop. iirc it took desktop linux ~10 years from 1 to 2% and 1 year from 3 to 4% so there's movement in the market.
DHH pushing desktop Linux for developers, Pewdiepie making videos about Linux and GrapheneOS, there's definitely an accelerationist argument to be made here.
So will we be able to have more than 11 programs on the taskbar without them being compacted?
Or a 2 row taskbar?
So I can easily switch between my 40 windows open? What is good for productivity?
And hopefully the customizable quick start bar... I lost hope in Windows but have to use it.
Ironic that they're opening up all the new crappy stuff, but not the good old stuff like Win32.
Maybe they're scared of WINE?
Maybe there's no one there anymore who understands the source code? Guys like Russinovich are mostly focussing on Azure nowadays.
Some of the old stuff was purchased any may never be open sourced.
I highly doubt Win32 was purchased from a third party.
I highly doubt they're certain that every line is original.
Dumb question from somebody who doesn’t do gui stuff: is this like a Window Manager, or more like GTK or QT or whatever?
It’s a UI framework like QtWidgets but closer to QML
I get there are old people and stuff, but why should anyone consider Windows these days?
Desktop/Laptop market share:
Windows: 72% Mac: 15% Linux: 5% ChromeOS: 2%
Windows share has been on a downward trend for at least 15 years. But it's a gorilla.
Lots of games don't run on Mac or Linux
Others, like adobe software, only run on Mac or windows
Windows computers are generally cheaper than macs for the same hardware
Plenty of stuff annoys me about windows, but for now it's still what I end up going with every time I need a new computer
How was it implemented? C#? C++?
The source is already available (albeit not buildable). See https://github.com/microsoft/microsoft-ui-xaml
That's not it
Classic Microsoft.
C++
Does this mean that we will be able to get the Quicklaunch toolbar back?
Another Windows UI. My goodness. 14 was enough.
I'm sure it'll be really user friendly(!)