1. The motivation with all of this was to allow Vector3<T> to use int for storage while still return

The motivation with all of this was to allow Vector3<T> to use int for storage while still returning float for methods like Length(), which is a rare but useful use-case. The extension methods for IFloatingPoint<T> allow you to not have to specify the return type
ok
so from options 1-4 which is your personal preference?
yes for BigInteger, Complex only implements INumberBase, i can broaden the scope of T if desired
noted
noted
noted
i don't understand what you meant by this
noted
noted
noted
i already implemented it
noted
noted

Kai•10/21/23, 8:43 PM

I'll look once home.
Only thing I want to point out is that Silk maths is the only lib were we didn't really pay attention in the working group tbh. We've revised and been unhappy with many things so do also take everything there with a grain of salt. Overall it's solid of course.

MaxineOP•10/21/23, 8:43 PM

"no more work than we were prepared to do for 2.0" does that mean you're open to implementing it, or don't want to do it because it's not worth it?

Perksey•10/21/23, 8:43 PM

yes open to implementing, vectorization is actually called out in the approved 3.0 SDP

MaxineOP•10/21/23, 8:44 PM

we can copy tanner's existing code for some operators, others will have to be done from scratch

MaxineOP•10/21/23, 8:44 PM

maybe could coordinate to share code with them?

Kai•10/21/23, 8:45 PM

Only from perkseys comments - do consider thinking about separating maths and geometry. Graphic programmers often put those together, but a vector and a point are very different.
Also, my stance has always been that we should allow the user to do things even if we think it's not such a good idea, ie VectorN<sbyte>

MaxineOP•10/21/23, 8:45 PM

Perksey•10/21/23, 8:45 PM

it's worth noting @tannergooding often attends working group meetings, and will probably want any discussions we have here to feed into whatever system.numerics does down the line

Perksey•10/21/23, 8:45 PM

so code sharing can probably be discussed there as well. i mean it's all .NET Foundation code really, so reuse is trivial.
but it's nice if there's prior art to inform what questions need to be asked and answered for the .NET API Review

MaxineOP•10/21/23, 8:46 PM

no one said i couldn't do this

MaxineOP•10/21/23, 8:46 PM

(for real, why isn't this public???)

Perksey•10/21/23, 8:48 PM

i mean, isn't that trivial to implement

Perksey•10/21/23, 8:48 PM

it's public in NumberFormatInfo, just not TChar

MaxineOP•10/21/23, 8:49 PM

that would require encoding, which allocates in a supposedly allocation-less API, so we wanna cache, so we wanna use a weak table...

khitiara•10/21/23, 8:49 PM

if doing utf8spanformattable remember utf8.trywrite for efficiency and interpolation and also consider implementing the other formattable-related interfaces

Perksey•10/21/23, 8:49 PM

        [MethodImpl(MethodImplOptions.AggressiveInlining)]
        internal ReadOnlySpan<TChar> NumberGroupSeparatorTChar<TChar>() where TChar : unmanaged, IUtfChar<TChar>
        {
            Debug.Assert(typeof(TChar) == typeof(char) || typeof(TChar) == typeof(byte));
            return typeof(TChar) == typeof(char) ?
                MemoryMarshal.Cast<char, TChar>(_numberGroupSeparator) :
                MemoryMarshal.Cast<byte, TChar>(_numberGroupSeparatorUtf8 ??= Encoding.UTF8.GetBytes(_numberGroupSeparator));
        }

        [MethodImpl(MethodImplOptions.AggressiveInlining)]
        internal ReadOnlySpan<TChar> NumberGroupSeparatorTChar<TChar>() where TChar : unmanaged, IUtfChar<TChar>
        {
            Debug.Assert(typeof(TChar) == typeof(char) || typeof(TChar) == typeof(byte));
            return typeof(TChar) == typeof(char) ?
                MemoryMarshal.Cast<char, TChar>(_numberGroupSeparator) :
                MemoryMarshal.Cast<byte, TChar>(_numberGroupSeparatorUtf8 ??= Encoding.UTF8.GetBytes(_numberGroupSeparator));
        }

MaxineOP•10/21/23, 8:49 PM

but the group separator format can be reset, so we need to make sure to check if it's up to date when we pull from the weak table, and also it needs to be thread-safe...

Kkhitiara if doing utf8spanformattable remember utf8.trywrite for efficiency and interpola...

MaxineOP•10/21/23, 8:50 PM

did you read my code

MaxineOP•10/21/23, 8:50 PM

i am using Utf8.TryWrite

khitiara•10/21/23, 8:50 PM

haven't had time

but good

MMaxine 1. The motivation with all of this was to allow Vector3<T> to use int for storag...

Kai•10/21/23, 8:50 PM

Kinda sniped by 1. here. I always wanted to have the ability to have ops have the ability to have a storage precision and each op have a specified precision.
Iirc there's lots of gains to have.

khitiara•10/21/23, 8:50 PM

I thought my being busy as hell was over and then nope heck me I guess

MaxineOP•10/21/23, 8:51 PM

worst part is having to manually call TryWriteInterpolatedStringHandler because interpolated strings can't natively do what the API needs them to

khitiara•10/21/23, 8:51 PM

wait really?

MaxineOP•10/21/23, 8:51 PM

oh yeah: and i have to allocate a new string(format)new string(format) because the string handler can't take a span<char>

Kai•10/21/23, 8:51 PM

Ie (wrong types for clarity) VectorN.Dot<i8, f8>(a, b)VectorN.Dot<i8, f8>(a, b)

KKai Only from perkseys comments - do consider thinking about separating maths and ge...

Perksey•10/21/23, 8:51 PM

this stance does fundamentally change the proposal, and as much as i have agreed in the past, given that we have strict rules codified in the BCL now in the form of the generic maths interfaces, it makes sense to follow them given they're the stewards of the ecosystem

MMaxine oh yeah: and i have to allocate a `new string(format)` because the string handle...

MaxineOP•10/21/23, 8:52 PM

(could rewrite the interpolated string handler, but...)

Kai•10/21/23, 8:52 PM

Where a/b are VectorN<i8> and the output is f8

MMaxine did you read my code 😭

khitiara•10/21/23, 8:53 PM

also where is this code i dont see it in the proposal and dont see any other recent links to check

MaxineOP•10/21/23, 8:53 PM

bottom of proposal

MaxineOP•10/21/23, 8:53 PM

https://github.com/uwx/SilkGenericVectorPoC

GitHub

GitHub - uwx/SilkGenericVectorPoC

Contribute to uwx/SilkGenericVectorPoC development by creating an account on GitHub.

PPerksey this stance does fundamentally change the proposal, and as much as i have agreed...

Kai•10/21/23, 8:53 PM

Idk, where does the BCL say Length of an sbyte vector makes no sense?

khitiara•10/21/23, 8:53 PM

the method declaration is in the proposal not implementation - ah ok

KKai Idk, where does the BCL say Length of an sbyte vector makes no sense?

MaxineOP•10/21/23, 8:54 PM

the proposal already provides that

MaxineOP•10/21/23, 8:54 PM

i'll leave it up there, rest is up to the WG

MaxineOP•10/21/23, 8:54 PM

do i have to write the API specification for all like 8 types?

Kai•10/21/23, 8:54 PM

Sorry I haven't looked yet, but from what's said here I gather that it returns float always?

MaxineOP•10/21/23, 8:54 PM

no, it returns whatever type you want

MaxineOP•10/21/23, 8:55 PM

vec.Length<TFloat>()

MMaxine do i have to write the API specification for all like 8 types?

Perksey•10/21/23, 8:55 PM

point out the differences between the types, but other than that no.

MMaxine vec.Length<TFloat>()

Kai•10/21/23, 8:55 PM

TFloat = sbyte?

Perksey•10/21/23, 8:55 PM

technically you don't even have to write the type specs which is something i'm abusing in the SilkTouchX proposal to leave things open for the team to change

KKai TFloat = sbyte?

MaxineOP•10/21/23, 8:55 PM

that would be a not very useful length method

MaxineOP•10/21/23, 8:55 PM

does sbyte implement IRootFunctions

MaxineOP•10/21/23, 8:55 PM

if not, not possible

Kai•10/21/23, 8:56 PM

Fair, Length squared is totally possible though?

PPerksey technically you don't even have to write the type specs which is something i'm a...

MaxineOP•10/21/23, 8:56 PM

i wrote them because you asked!

KKai Fair, Length squared is totally possible though?

MaxineOP•10/21/23, 8:56 PM

lengthsquared is not constrained

Kai•10/21/23, 8:56 PM

Yes it'll likely overflow and isn't super common, but eh

[MethodImpl(MethodImplOptions.AggressiveInlining)] internal ReadOnlySpan<TChar> NumberGroupSeparatorTChar<TChar>() where TChar : unmanaged, IUtfChar<TChar> { Debug.Assert(typeof(TChar) == typeof(char) || typeof(TChar) == typeof(byte)); return typeof(TChar) == typeof(char) ? MemoryMarshal.Cast<char, TChar>(_numberGroupSeparator) : MemoryMarshal.Cast<byte, TChar>(_numberGroupSeparatorUtf8 ??= Encoding.UTF8.GetBytes(_numberGroupSeparator)); }

1. The motivation with all of this was to allow Vector3<T> to use int for storage while still return

Similar Threads

Similar Threads

Similar Threads