23 Replies
./
so it should be as simple as removing the raw winapi and blitting into a js canvas
oh yeah that's what i'm hoping, just checking that green thing is what i should expect
if you see suspicious blocky artifacts, ping me
they appear on the v128 and scalar rasterizers i think, thanks to a net8 codegen issue
you can press R to switch rasterizer impls
👍 currently working on adding a simple OpenGL based texture renderer for the rasteriser output
well not right this second, right now I’m currently drinking a cuppa outside
lol it isn't looking good
i thought my browser tab was frozen but turns out it was working
i'll let you know how many frames per minute i'm getting shortly, just need to figure out why I'm getting a PlatformNotSupportedException somewhere
sounds nicely concerning
lol
there was also an AVE but that could be a me problem
i'll push this up and if you'd like you can take over
i assume you haven't pushed yet?
can't see a fork at least
I did, it’s the latest branch I pushed up in the main repo
main silk.net repo
ah, found it
this is my experience with trying to compile hello world to wasm at the moment:
@Perksey so have you tried debugging wasm without blazor?
i tried debugging once and hated it
but that was with blazor
i literally cannot find any docs on it, and i've tried 5 different things
but VS is just not even trying
i think i give up for the moment
not only can i not attach a debugger to mono
running AOT is not actual AOT for wasm apparently
yeah AOT isn't supported for Blazor atm
i'm not even using blazor 
i think there are knobs you can tweak to get it configured 🤔
perhaps
my brain is frying
still seems to load .dlls
did you figure out the PlatformNotSupportedException?
i seem to be getting them
alright, i give up
i'll come back to this in net8
beyond sick of dealing with dotnet stupids for the last 6 hours
it was just some unguarded intrinsics that i added software fallbacks for
iirc the _MM_TRANSPOSE4_PS was the main one
almost fixed _MM_TRANSPOSE4_PS upstream
and done
my browser demo still broken thoughi really don't know where the AVE was coming from, it was within readBackBuffer which is weird
kinda smells like something being off with the memory allocation
i am le clueless
apparently there is a
WasmEnableSIMD
but setting it to true crashes mono AOT compiler :/
will try it when net8 releases