RI am currently trying to understand what Silktouch.PInvokeNativeContextOverride does 
Pyeah osx-universal isn't a thing
Pthat's a imgui bug
PPRRight, well how do they load it then, since it looks to be just a simple DllImport: https://github.com/mellinoe/ImGui.NET/blob/master/src/ImGui.NET/Generated/ImGuiNative.gen.cs
PNativeLibrary is [sometimes] smart enough to look in the runtimes folder
Pand DllImport just uses NativeLibrary
RAh, and since Silk overrides it, somehing goes wrong?
Poh no
PPInvokeNativeContextOverride is a shim to force Silk.NET to use DllImport too on some platforms
Pwouldn't impact imgui whatsoever
P(it's only a SilkTouch thing)
RRight, so then why does it not work when using Silk?
Pdoes it work when not using silk?
Pit's very likely imgui.net's just broken
ROk Imgui.NET is broken...
RTheir sample program worked, because it is set to copy cimgui to root of the output folder: https://github.com/mellinoe/ImGui.NET/blob/master/src/ImGui.NET.SampleProgram.XNA/ImGui.NET.SampleProgram.XNA.csproj#L18
Limgui on silk works fine with mac for me
Lthe screenshot was veldrid, but it worked on my own custom GL 4.1 code too
Platest version of ImGui.NET?
Lnah the one that the controller uses, which is 1.86 i believe
Pah. you in theory should be able to up to date relatively seamlessly if you did want to try it up
RThat older version of Imgui.Net only works on Intel Macs, not M1.
RThis is a hack, but since DllImport does not actually load the library if it is already in memory, we can get latest Imgui.NET to work by just loading it first with Silks LoadNativeLibrary before calling
ImGuiNET.ImGui.CreateContext();.RIm sure this can break in many ways 
So maybe best to wait until Imgui.NET is actually working before updating it.
So maybe best to wait until Imgui.NET is actually working before updating it.
Kthat's actually surprisingly robust
Pfinally fixed most of the input backend differences between sdl and glfw
Pcoming to a 2.15 near you
Pwent through and pressed every button on my keyboard to compare the output of inputtest
Poh, and no this isn't an april fools joke
Pthis has actually been fixed
K
amazing work as alwaysPthank you
Phttps://github.com/dotnet/Silk.NET/pull/875 is ready for review
Pgiven that most of the regenerated bindings were tackled in the directstorage PR, this one is pretty tiny by comparison!
KWhat you build for #867 is really cool! This will also help with some issues surrounding platforms without a File system
Paha thank you, yeah hopefully it’s one of those things that power users can use to tweak things to their liking
Li have a silk windowing Veldrid imgui controller, but not sure if i should upstream it here or to veldrid repo
Pprobably upstream, I think Silk.NET.Windowing.Extensions.Veldrid.ImGui is scope creeping a bit
Lgotcha
KYouTubeGoogle Chrome Developers
Jake challenges Ada (from Samsung Internet https://developer.samsung.com/internet) to a game of "what key was pressed" while diving into how keyboard events work, and the best way to handle user input.
The original tweet → https://goo.gle/36QUegg
Wordledge → https://goo.gle/3qIPtMJ
KeyboardEvent → https://goo.gle/3tLnk9X
Dvorak layout → http...
The original tweet → https://goo.gle/36QUegg
Wordledge → https://goo.gle/3qIPtMJ
KeyboardEvent → https://goo.gle/3tLnk9X
Dvorak layout → http...
P@kaijellinghaus had a meeting with Tom Pappas (executive director of DNF) this evening and pushed the server request again
Phopefully we’ll actually be able to get a jupyterhub server setup
Pand from there, start working on try silk.net I guess
Lpog
P(Except don’t, because I’ve already said I’ll hold down the fort on 2.X alone so you can work on 3.0)
TI see Issue #821 is still open, didn't this get fixed by PR #828 (the GLFW callbacks not being released thing)?
PNope, that was just one issue
Pthe underlying issue we diagnosed in that issue was that SilkTouch doesn’t allocate each individual native name a GC slot