Prioritize Modernization investments

[pchaurasia14]

We kindly ask you for your valuable input on the priorities for modernizing investments that the WPF team should focus on, as outlined in the roadmap document for 2023.

Please use thumbs up (no other emoji responses will be considered to avoid duplication) on options below which you would like us to prioritize.

[pchaurasia14]

1. Windows 11 Theming

Goal	Description	Rationale
Windows 11 Theming	Bringing Windows 11 look and feel for majority of WPF controls. Support for Win11 features such as snap layout, rounded corners for controls and newer color schemes would bring enhanced experience for WPF applications. We will iterate on full scope of the work with WPF community.	For all consumer applications that are built on WPF running on Win11, this feature would ensure that applications can take advantage of modern design elements and behaviors.

[maxima120]

That's disappointing

There is a very well known wpf controls suite that costs $900. It has 60 themes that you can apply on out of the box microsoft controls easily. If you can't afford that may be you shouldnt be voting for features, skewing the actual developers representation.

[kevwkev]

Any chance we could get a modern Office 365 Ribbon control and Simplified Ribbon style?

[NeverMorewd]

Well, It seems that Windows 11 Theming win the election.So have we started yet?

[pchaurasia14]

@NeverMorewd - Soon enough. There are some things pending on the internal partner dependencies. Once sorted, we should be able to make progress on this front.

[PabloGallegosH]

Hello good day Every one. Excelent job. Starter now this plan?

[pchaurasia14]

2. Newer Controls

Goal	Description	Rationale
Newer controls	WPF FolderBrowserDialog - Introducing native support for FolderBrowserDialog for WPF	This has been a top ask from the community since .NET Core 3. This feature would reduce dependency on WinForms and other third-party alternatives.

[uecasm]

I've always found a folder browser to be completely useless, because it doesn't allow you to type in a path. I always use a plain OpenFileDialog with a dummy filename+filter instead; works much better. There's other new controls I would like, but that one definitely isn't one of them.

[aquinn39]

I've always found a folder browser to be completely useless, because it doesn't allow you to type in a path. I always use a plain OpenFileDialog with a dummy filename+filter instead; works much better. There's other new controls I would like, but that one definitely isn't one of them.

That's the old FolderBrowser. The newer one is much better - it looks and behaves more like a file browser including letting users type in a path.

[Kurpanik]

With ookii. dialogs there ist already a WPF solution, which solves the problem of the missing dialogue very quickly. For the other two options, however, there is far less possibility of implementation.

[Alexgoon]

Thank you for considering implementing this feature. If you decide to move forward with it, I would like to suggest that you also consider the possibility of adding support for custom theming to the dialog window. Doing so would provide users with a more cohesive and personalized experience, as they would be able to customize the look and feel of the dialog to match the rest of their application.

Probably, supporting custom theming will be available without additional investments on your side, however, I wanted to make sure that you are keeping this in mind as you move forward with the implementation.

[harikrishnan-p-v]

Thanks for implementing this I have been using "Microsoft-WindowsAPICodePack-shell" by rpastric,contre,dahall. Introducing native support for FolderBrowserDialog for WPF is a much awaited feature.

[pchaurasia14]

3. Nullability annotations

Goal	Description	Rationale
Nullability annotations	Enable nullability annotations in WPF	This increases the quality of the code base, as well as the quality for all WPF apps consuming it and reduces time spent debugging ArgumentNullExceptions and NullReferenceExceptions. Rest of the dotnet (eg. winforms) repo is already moving in the direction and this goal would bring WPF to latest standards as well. We welcome any community contributions in this area.

[pchaurasia14]

@danmoseley @dotMorten - Great idea, thanks for suggestions. I just created new issue and we will populate the plan there in a few weeks. It is great if there is interest from community to contribute, we would be thrilled to collaborate on that and we will make sure to make time for PR reviews and project driving.

[zspitz]

This seems to me the lowest hanging fruit:

1. The need can't be filled by third-party libraries (unlike file dialogs).
2. It requires the least amount of domain internals knowledge (unlike updates to theming).
3. Each individual change has a specific, relatively local effect, thus making it easier for the community to collaborate.

[SteveDunn]

I'd be happy to work on some of these annotations. @zspitz sums it up perfectly; I don't have much knowledge of the internals, but would be happy to skip around the code to determine nullability of existing contracts.

[kant2002]

https://github.com/dotnet/winforms/pulls?q=is%3Apr+is%3Aopen+nullability Here basically one man show for supporting nullable annotations. It was done in small steps. Basically you need to show support for some contributor and process will happens by itself. Probably it can be done in WinForms because they have tests already in place.

[BlameTwo]

Is it compatible with higher DX version?

[cliyr]

+1

[LightBender]

+1

[nistech]

+1

[NeverMorewd]

+1

[lindexi]

Can we get something from D3D9On12 ?

[wstaelens]

#3546
dotnet/runtime#78374 (comment)
± 7 - 8 years and still not fixed. 🤞

[batzen]

@wstaelens could you please stop mentioning the XPS issues in every conversation that comes up?
This discussion is a vote for something totally unrelated to XPS.

[wstaelens]

@batzen fixed in .net 9 preview 7 ;-)

[h3xds1nz]

And it was a runtime issue as you sure have noticed.

[wstaelens]

@h3xds1nz xps is a subset of xaml and indeed the culprit was located in the compression libraries. But improvements are coming towards XPS which potentially might improve XAML/WPF also: #4000 (comment)

[SkinnySackboy]

If you're looking at DX improvements perhaps the WinForms airspace issue could be resolved, as it's something that still causes us headaches?

[pchaurasia14]

@SkinnySackboy - DX improvements are currently not planned for this year. We will revisit the plan for future releases.

[SkinnySackboy]

Ok thanks.

[johnottone]

If you're looking at DX improvements perhaps the WinForms airspace issue could be resolved, as it's something that still causes us headaches?

@SkinnySackboy, It's funny I ran across this post not long ago. Not saying it is going to solve any problems you maybe having or anything, but you never know, it may spark an idea or something, that may lead to you discovering a solution...

jonathanantoine - WPF 4.5 – Part 8: No more airspace problems - Integrating WPF with win32: finally, I can use the WebBrowser control

LOTS of USEFUL KNOWLEDGE

Check out this control and get some ideas from it perhaps:
Airspace Fixer

and also there is this which youve probably read, but still relevant:

WPF Interoperation

Changing Layout Dynamically in Windows Forms

also, see TableLayoutPanel, FlowLayoutPanel, Docking controls,

[StefanKoell]

While I understand the desire for new controls, I think the priorities and the focus should be at a fundamental level. Areas which are not (and will not be) covered by 3rd parties, like

• more efficient and faster rendering (new DirectX version, better call batching, etc)
• improved interop (airspace issues)
• compiled bindings
• native SVG support

To name just a few. There have been lots of issues/discussions over the year but no progress at all.

[292390450]

Indeed, its rendering performance now seems to be a bit slow compared with avalonia，More updates such as reflective performance and virtualized ItemsRepeater make wpf look faster

[pchaurasia14]

@StefanKoell - I agree that areas that are not covered by 3rd party solutions should be our focus. However, some of those areas are lacking adequate test coverage which puts us at a disadvantage since we cannot ascertain if the newer changes will cause unexpected regressions.
Which is why our plan is to first open source the tests and then work on bigger initiatives.

[pvillads]

1. Theming
2. Nullability support

[pchaurasia14]

@pvillads - Thanks for the feedback. Appreciate if you could cast your vote via thumbs up on the above items.

[rbqm]

Whatever you do, please do not break backward compatibility with .NET Framework. For mission critical WPF applications with release cycles measure in years it is impossible to migrate from .NET Framework to .NET X because of the difference in lifecycle policy between .NET Framework and .NET X. The new .NET X lifecycle policy where version X receives security updates for a limited amount of time only is not suitable for such applications, unless there is a way to upgrade applications installed in the field from .NET X to .NET X+1 without having to release a new version of the application itself. I hope that Microsoft realizes that there needs to be a future version of .NET X that has the same lifecycle policy as .NET Framework. When/if that time comes, I want to be able to port my application from .NET Framework to .NET X without having to rewrite it.

[jimfoye]

@Teddy55Codes How big was it?

[Teddy55Codes]

I think it was some where around 60MB bigger

[jimfoye]

So let's say your app is 10MB - that's 70MB, you're installing the entire .NET runtime on the computer in order to run your application. Meanwhile, other folks are doing the same, you have to wonder how many .NET runtimes a computer is going to end up having installed on it? It's good to have the option, of course, but many of us want to do things "the old way" where the runtime is installed on the computer and any app compiled for that runtime will work. Microsoft does not seem to see things that way, though.

[Teddy55Codes]

I do see your perspective. However, .NET is no longer a Windows only technology. In my opinion that is actually very good. Things like the GAC where always a head scratcher for me. I happily take a bit of a larger app size if it means i have to deal with less Windows bs.

[jimfoye]

I do see your perspective. However, .NET is no longer a Windows only technology.

Yeah but we're talking about WPF apps here. In that context, it is a Windows only technology. (And always will be, apparently).

[pchaurasia14]

Thank you very much for your insights folks and for your help modernization investments prioritization!

As of 2/3/2023, we have current voting status:

• Windows 11 Theming - 225 votes
• Newer Controls - 155 votes
• Nullability Support - 99 votes 

Based on the results, we are inclined to channel our focus on delivering on the Windows 11 theming and newer controls support as top 2 items.
As for Nullability support -- we are thankful for community interest in helping with the work. We are actively finalizing the plan (draft plan is already available here) to guide community contributions in the space. We plan to make PR reviews for Nullability support our top priority to unblock the community effort. 

Thank you! And as usual, let us know what you think or if we missed anything.

[maxima120]

Why do we go to a third party controls supplier like devexpress

How much did you pay for. Net? Nothing? And you expect free 1hr support response and a fix tomorrow? What do you smoke? I want it too

[wstaelens]

Why do we go to a third party controls supplier like devexpress

How much did you pay for. Net? Nothing? And you expect free 1hr support response and a fix tomorrow? What do you smoke? I want it too

Don't worry about that, we don't smoke and pay our Microsoft bills each month :-)
I didn't say we expect such a fast support but if we pay for it and it happens so fast, we can only enjoy working with such a third party controls supplier in that way.

I look forward to seeing the promises turned into action in the next weeks.

[johnottone]

@pchaurasia14

@johnottone - ThemeBuilder sounds like nice idea. How about logging an issue and see how we can bring it to fruition.

I will log it under issues I guess, but wouldnt it be better in the Discussions forum under the "Ideas" category, or is that where Ideas actually go to die? LOL

[Alexgoon]

@batzen
As a member of the DevExpress WPF team, I would like to let you know that we appreciate your input. Large third-party component suites, such as DevExpress assemblies, can sometimes cause startup performance issues due to the lack of native images in the system. However, we have found that using Ngen, MultiCore JIT, or ReadyToRun can often help address these issues. In fact, using MultiCore JIT and ReadyToRun together can yield even better results in .NET/.NET Core. You can find more information on these techniques in our blog post here: 9 Tips to Reduce WPF App Startup Time.

We are always striving to improve our products, and reducing startup time is one of our top priorities. To that end, we are currently working on a new optimized theme for our upcoming release that we believe will further enhance performance.

Thank you again for your feedback. If you decide to give our components another try, please let me know, and I will personally review your requirements and provide you with general guidelines to help you get started as smoothly as possible.

[batzen]

@Alexgoon feel free to contact me on Twitter. Your DMs seem to be closed there.

[huoyaoyuan]

Fundamental overhauls like porting C++/CLI to C# should be supported better. They can block a series of further follow-ups. Perhaps BCL experts can help in this area.

[jimfoye]

Nothing significant like that is going to happen.

[RoyLaiYi]

The air space issue is very serious, there’s no good solution to resolve it perfectly for many years.

[wgebczyk]

Maybe admit failure of Maui, winui, etc and provide non Win rederers like Avalonia. You did that with Mono "fork" then you can do WPF 2.0 cross platform.

I know, this wish require some strong visionaries dragging it regardless all obstacles that maintainers. Maybe hire someone that would drag it till happy end?

[jimfoye]

I have switched over to the "New Outlook" and...it is a progressive web app! Does not seem like a vote of confidence in native Windows desktop development.

[wstaelens]

I have switched over to the "New Outlook" and...it is a progressive web app! Does not seem like a vote of confidence in native Windows desktop development.

don't switch to the "new outlook", doesn't support COM or anything. Breaks a lot of applications.
Switch to it if you want many support calls 😞

[johnottone]

As far as Theming WPF is concerned, I know it was voted on and was put on the roadmap, but if there are more important things that need to be addressed, or resources have been or need to be shifted, well, it happens a lot in my world, so I'm used to it.

Maybe admit failure of Maui, winui, etc and provide non Win rederers like Avalonia. You did that with Mono "fork" then you can do WPF 2.0 cross platform.

I know, this wish require some strong visionaries dragging it regardless all obstacles that maintainers. Maybe hire someone that would drag it till happy end?

I have to strongly disagree with "admitting failure". If that were the answer, then perhaps there wouldn't be WPF today at all. Yes, it took time, but if it was just given up on, then you wouldn't be asking for a WPF2.0.
I personally like to stay on the cutting edge, leveraging new tools and frameworks of the time, because as these tools emerge, other advances are made. And be clear that sometimes it's not .NET that's behind, it may be the underlying technology is just not there yet. A couple years back I had the need to start using Xamarin, and it was ....ehhhh......
So, I was excited when MAUI was announced, especially with Blazor, because it filled many a void and I gladly replaced our Xamarin apps with MAUI apps, and getting ready to implement 3 more entirely new apps with MAUI Admittedly, this held (and still holds) some uncertainties.

For you specifically, it seems that the lack of FULL NativeAOT WASM, due to certain dependencies on the IL, you may be, or may have been deterred. But the fact that it can be embedded in virtually all frameworks; WPF, WinForms, Maui, etc., and the fact that NOW, you can compile that portion into the Web Assembly itself, it's still very promising, IMHO... I think it still increases the storage, and I acknowledge that it may have kept some devs from leveraging the newer technologies (NET Core, MAUI, WinUI). But I'm optimistic and think that it's a path worth following. Showing my age here, but It kind of reminds me of how WPF started and the uncertainties that existed when I started using it, even before its RC1, back when my company was subscribed to the MSDN (on cdroms...lol) and MS would send us the CTPs. I still have that old medical app demo that MS released with the flipping tabs and live heart rate monitors and such. And still have the very first apps I wrote in WPF (Thank you JMAC). It literally changed my whole world.