Very true indeed. And we only need to take a look at how these newfangled interfaces are being received by the users to tell the whole story. Just because it works fine on Mobile it doesn't mean it's the way to go for actual working machines as well.dpJudas wrote: Windows 8's infamous metro UI is WPF/XAML, a complete redesign of how those things works. Likewise, Qt 5's new toy Qt Quick and the resulting desire for KDE5 to build itself on top of it changed what is easy or hard to build. When you look at websites with HTML 5, technically you can do almost anything that a desktop UI can do, but you almost never see it in practice because it requires too much work. By the same token, WPF or Qt Quick seem to suffer from the fact that while an application can do things far more freeform now, they also have to do a lot more work for stuff that used to be easy and quick to do before. As a result most applications don't bother to do half of what older apps did, usually defining themselves out of the problem by being more "modern". Less is more, supposedly.
TL;DR: not every technology rewrite is an upgrade. Sometimes the old toolkit actually was better.
It's all about visual design these days and these newfangled tools clearly have their focus there. But what's ironic is that getting a good UI out of those tools is exceedingly hard. Even when working on IOS I generally steer clear of Apple's Interface Builder and choose to construct the UI manually. The results are always far, far superior. So this becomes a genuine problem if some UI framework totally depends on these visual editors and cannot work without them. Windows/Metro and XAML are probably the worst example. While it can be programmed directly, it is a lot more complicated than doing it on iOS or Android. One more nail in the coffin of this misbegotten piece of crap.
And regarding games and the API they target, I think the choice between Vulkan and DX12 is a lot more balanced than between DX10 and OpenGL3. Say what you want, but the initial GL3 specs were a disaster. The system was saddled with problems that took years to fix and by the time they finally did get fixed (which was as late as GL 4.4!), anyone had lost interest because XP and Vista were no longer relevant platforms.
With Vulkan it's very different: Go DX12 and you lose all customers still on Win 7 or 8.1, but with Vulkan you can still support them. And the latest Steam survey still shows those at roughly 35%, i.e. more than one third. Aside from that, DX12 has no deciding advantage over Vulkan. In the end this will mostly be decided by developers' preference, unless they feel compelled to target the Windows Store - which is more suited for casual games because the real gamers probably won't have to deal with it anyway.