Harald Steinlechner

https://pastebin.com/0xurPWgW

- text might work on iris gpu without multisampling - Slim applications in GL actually have multisampled backbuffers (samplecount = 2) according to RenderDoc -> this makes problems with outline...

so this is the last conclusion for the moment - it is not possible to make it work on ``` 0: vendor: "Intel" 0: renderer: "Intel(R) Iris(R) Xe Graphics" 0:...

strange, i have never seen glfw to fail except for headless scenarios. do vulkaninfo or glxinfo/glxgears work?

nope, not goes forever but for several minutes. it seems to happen within custom pickler registration.  https://github.com/pro3d-space/PRo3D/issues/228 PRo3D, autodeploy branch can be used to reproduce the problem. EDIT: also...

so in the end, adaptify seems to silently require f# 6.0 at runtime, and our introspection code is not robust to problems like this by printing thousands of errors till...

just a quick update on problem 1: we further invastigated and rewrote suave internals in order to use the new `socket.receiveAsync(buffer)` (available in netstandard2 and net471) instead of the eventargs...

cool. hope you can find out problem 1 :)

@krauthaufen i think the reference is for this one: ``` let observer = async { do! Async.SwitchToNewThread()` ``` And indeed you're right, this observer thread (hidden behind task) increases likelyhood...

some news. the accpet performance is most likely the same problem. in this case suave's client tasks steal work pool threads from the system, making it stoned for new clients....