Silk.NET•3y ago

we didn't settle, this is just us pursuing a prototype

dfkeenan🍍🍕•10/17/23, 9:59 PM

I don't know why I'm here.

PPerksey we didn't settle, this is just us pursuing a prototype

PerkseyOP•10/17/23, 9:59 PM

none of this has been formally agreed

Aqua•10/17/23, 9:59 PM

i'm here because i enjoyed that silk offered a lot more options than opentk
i didn't care for unsafe initially
and i'm currently trying to play devil's advocate, that is all
and i'm trying to come at this objectively, subjectively my opinions are somewhat different to what i am saying

Maxine•10/17/23, 9:59 PM

yeah but it's good enough

AAqua i'm here because i enjoyed that silk offered a lot more options than opentk i di...

PerkseyOP•10/17/23, 10:00 PM

oh yeah no it's appreciated

Maxine•10/17/23, 10:00 PM

all bureaucracy aside, i don't see any way of doing this in practicality that isn't the cast system

khitiara•10/17/23, 10:00 PM

unless we get ref struct as generics and as ref fields

khitiara•10/17/23, 10:00 PM

in which case it becomes a lot easier

Curin•10/17/23, 10:01 PM

Not to derail too much, but for the string conversions, going to strings is expected just to be explicit from a function right (need for array sizes), and then implicit from string arrays to our types?

PerkseyOP•10/17/23, 10:03 PM

implicit from string/string array, explicit to string, to string array has to be a function

PerkseyOP•10/17/23, 10:03 PM

with the ReadToString/ReadToStringArray functions as appropriate

Curin•10/17/23, 10:04 PM

okay, sounds good

Ddfkeenan🍍🍕I don't know why I'm here. 😂

PerkseyOP•10/17/23, 10:22 PM

perhaps because you want to gaslight people into thinking pineapple on pizza is okay

yeah i saw your nickname change

dfkeenan🍍🍕•10/17/23, 11:16 PM

It's been like this for ages on another server. I decided to make it the same everywhere

Curin•10/18/23, 2:58 AM

/// <summary>
/// creates a <see cref="MutMut"/> from a pointer array
/// </summary>
/// <param name="array"></param>
public static implicit operator MutMut(byte*[] array)
{
    IL.Emit.Ldarg_0();
    IL.Emit.Ldc_I4_0();
    IL.Emit.Ldelema(TypeRef.Type(typeof(byte).MakePointerType()));
    IL.Emit.Newobj(
        MethodRef.Constructor(
            TypeRef.Type(typeof(MutMut)),
            TypeRef.Type(typeof(Mut))
        )
    );
    IL.Emit.Ret();
    throw IL.Unreachable();
}

/// <summary>
/// creates a <see cref="MutMut"/> from a pointer array
/// </summary>
/// <param name="array"></param>
public static implicit operator MutMut(byte*[] array)
{
    IL.Emit.Ldarg_0();
    IL.Emit.Ldc_I4_0();
    IL.Emit.Ldelema(TypeRef.Type(typeof(byte).MakePointerType()));
    IL.Emit.Newobj(
        MethodRef.Constructor(
            TypeRef.Type(typeof(MutMut)),
            TypeRef.Type(typeof(Mut))
        )
    );
    IL.Emit.Ret();
    throw IL.Unreachable();
}

So this code block gives me the following error, but I do have a constructor (public MutMut(ref Mut @Ref)). Any ideas how to fix this?

Fody/InlineIL: Type Silk.NET.Core.MutMut has no constructor with signature (Silk.NET.Core.Mut) (in Silk.NET.Core.MutMut Silk.NET.Core.MutMut::op_Implicit(System.Byte*[]) at instruction IL_0053: call System.Void InlineIL.IL/Emit::Newobj(InlineIL.MethodRef))

khitiara•10/18/23, 2:59 AM

need to makebyreftype the mut in the constructor specification

khitiara•10/18/23, 3:00 AM

ref mut and mut are diff types

Kkhitiara need to makebyreftype the mut in the constructor specification

Curin•10/18/23, 3:03 AM

Makes sense, seems to have worked. Thank you!

Curin•10/18/23, 2:06 PM

@Perksey What do the ConstAny and Any types map to with the new types?

CCurin @Perksey What do the ConstAny and Any types map to with the new types?

PerkseyOP•10/18/23, 4:47 PM

they’re equivalent to the non-generic ones

Curin•10/18/23, 4:47 PM

okay that's what I thought. What are the generation steps, I need to regenerate GLFW.

PerkseyOP•10/18/23, 5:07 PM

should just be a case of running SilkTouchX with the test.json path as an argument (oh and add “—skip Windows” or whatever it’s called in test.json, as that will take forever

PPerksey should just be a case of running SilkTouchX with the test.json path as an argume...

Curin•10/18/23, 5:20 PM

okay got it running but it didn't generate anything

PerkseyOP•10/18/23, 5:20 PM

-l debug?

PerkseyOP•10/18/23, 5:21 PM

(you could also do trace but that may be excessive)

Curin•10/18/23, 5:23 PM

in either case it just looks like it is doing no work

Curin•10/18/23, 5:23 PM

PerkseyOP•10/18/23, 5:24 PM

hmmm, do you have submodules initialised

Curin•10/18/23, 5:24 PM

yup

PerkseyOP•10/18/23, 5:25 PM

maybe crank it up to trace?

Curin•10/18/23, 5:25 PM

that is on trace

PerkseyOP•10/18/23, 5:25 PM

this is so strange

PerkseyOP•10/18/23, 5:26 PM

I mean it’s probably a case of ClangSharp not running

PerkseyOP•10/18/23, 5:26 PM

but how it gets to that point is a mystery

Curin•10/18/23, 5:26 PM

okay I guess I'll go down the debug rabbit hole and see if I can figure out what's going wrong

PerkseyOP•10/18/23, 5:27 PM

unless it’s to do with the working directory. I tried my best to make it read relative to the file but may have missed a few places

Curin•10/18/23, 5:37 PM

yeah so I'm definitely missing the clang response file

PerkseyOP•10/18/23, 5:38 PM

wait is that not in the repo?

Curin•10/18/23, 5:38 PM

nope

Curin•10/18/23, 5:40 PM

I don't have any changes from yesterday, but it doesn't seem like any of that should affect this

PerkseyOP•10/18/23, 5:40 PM

pushed them up, looks like they were gitignored

Curin•10/18/23, 5:41 PM

for future reference, is there a way to generate those response files on my end?

PerkseyOP•10/18/23, 5:41 PM

they're not generated, they're hand-crafted

PerkseyOP•10/18/23, 5:42 PM

i don't know why they're called response files, but basically they include the command line arguments that are passed to ClangSharp's P/Invoke generator

Curin•10/18/23, 5:42 PM

okay

Curin•10/18/23, 5:45 PM

still not generating, doesn't look like it is finding the response file

PerkseyOP•10/18/23, 5:47 PM

are you running in the SilkTouchX dir (the same dir as the test.json)? this shouldn't be required, and if this works then that's a bug i need to fix

Curin•10/18/23, 5:51 PM

no I'm running from the VS output directory (in the artifacts folder)

PerkseyOP•10/18/23, 5:52 PM

ahh, give running in the test.json dir a go, if that works then i need to fix that because ideally it'd do everything relative to the config file in question

/// <summary> /// creates a <see cref="MutMut"/> from a pointer array /// </summary> /// <param name="array"></param> public static implicit operator MutMut(byte*[] array) { IL.Emit.Ldarg_0(); IL.Emit.Ldc_I4_0(); IL.Emit.Ldelema(TypeRef.Type(typeof(byte).MakePointerType())); IL.Emit.Newobj( MethodRef.Constructor( TypeRef.Type(typeof(MutMut)), TypeRef.Type(typeof(Mut)) ) ); IL.Emit.Ret(); throw IL.Unreachable(); }

we didn't settle, this is just us pursuing a prototype

Similar Threads

Similar Threads

Similar Threads