Comment by smackeyacky
4 days ago
All these changes, yet still no satisfactory UI framework.
Still no migration path from WinForms to something modern.
I love .NET but if you're taking on an older desktop app that is still using VB.NET and WinForms it feels like Microsoft have decided to let you circle the drain with no supported way forward into MAUI or whatever stupid XAML UI they decided is the solution these days.
On a server, .NET is fantastic and it's entirely possible to build and debug your system on Linux with no loss of productivity, but the desktop stuff is bad and getting worse.
I've tried each iteration of UI paradigm they've tried since WinRT and never really had any significant problems with any of them. WinRT, UWP, WinUI, MAUI...
But then they aren't even willing to invest the time to dogfood their own products and fully replace the windows UI. Really doesn't inspire confidence.
I suspect they also made a bad bet by going so hard on blazor without putting more support behind WASM itself. Now that's stalled out with WASM's own failure to reach critical mass.
Microsoft has zero vision in the UI space. It’s crazy that there isn’t there a single, obvious solution for developing Windows applications in 2025.
I agree that it seems they’ve made some bad bets and now are bogged down supporting half a dozen different frameworks with different limitations and tradeoffs.
They keep trying to reinvent the wheel, but it doesn’t seem like they’ve ever really meaningfully improved on WPF.
At least there is Avalonia for an open source, cross platform WPF alternative. It seems like the sanest way of making desktop applications with C# currently.
That would be a first, not having significant problems.
As someone that was deeply invested into WinRT since Windows 8, and went back into distributed systems and Web after all the disappointment regarding how managemetn handled it.
Everyone on the Windows development ecosystem has had enough from the multiple reboots and tool changes, UWP never being as good as WPF, WinUI iterations even worse, the team now went radio silent, the last community call was a disaster,...
Sorry, to be clear I meant no significant problems with the tech.
Sure, it was opinionated tech, but that would’ve been ok if they had stuck with it like C# itself and fleshed it out more, which only would’ve happened had they implemented the entire OS in it. The fact that they couldn’t is in my mind the exact reason the tech failed.
1 reply →
You know, I agree but thinking about it, is it really a fair ask? What's the analog in Java, Go, Rust, etc?
Is Swing the competition? It feels like mobile and web stacks have really sucked the air out of the room.
Assuming the web is not an option for whatever reason.
I did some Win32 interop recently and found it to be refreshing. It's definitely lower level and more difficult, but it will outlive all of these other paths.
Using tools like cswin32 makes this a lot more tolerable development experience from managed code.