RuQing Xu xrq-phys

This guide is intended to provide a OSS-based solution for enabling NTFS read+write access on modern macOS. In the meantime, I see some commercial and semi-commercial solutions for NTFS on mac. Some of them are excellent, others might have a debatable reputation. Discretion is stronly adviced there.

本说明旨在介绍一个在现代macOS平台（包括M1/M2等新架构和Intel老架构mac电脑）上可用的NTFS读写模式的挂载方案。该方案拷贝自 takeit/INSTALL.md，当时在macOS Sierra上可用使用但作者述说需要关闭SIP。本文经测试该方案实有不影响SIP的用法。

NTFS for macOS的商业或半商业的解决方案较为鱼龙混杂，亦存在部分组织误导用户认为「只有少数解决方案能支持最新系统架构」。实际上，macFUSE至今仍被维护，ntfs-3g也仍然可用（编译很快）。希望大家仔细甄别。

1. Install macFUSE:

What

Compile NumPy on Apple Silicon M1 Chip

How

Create virtual env at, say export VENV=$HOME/.local:

python3 -m venv $VENV
. $VENV/bin/activate

This guide is a summation of current situation as of Apr. 26, 2020, on how to build and use native Aarch64 (ARM64) Visual Studio Code on devices like Surface Pro X.

	/**
	* This Source Code Form is subject to the terms of the Mozilla Public
	* License, v. 2.0. If a copy of the MPL was not distributed with this
	* file, You can obtain one at https://mozilla.org/MPL/2.0/.
	*/
	#pragma once
	#include <complex>
	#include <algorithm>

	// Real softplus.

	import Foundation
	import MLCompute
	import PlaygroundSupport

	let iPage = PlaygroundPage.current
	iPage.needsIndefiniteExecution = true

	/*
	* Apple says MLCMatMulLayer does a ``batch matrix multiplication''
	* but didn't make clear its meaning.

	module BatchDef

	macro def_numed_fun(funname, num)
	numed_funname = Symbol(string(funname, num))
	@show numed_funname
	return quote
	$(esc(numed_funname))() = $num
	end
	end

	# here idx has 3 entries corresponding to e.g. "ik,jk->ij".
	contract!(A::Array{T},
	B::Array{T},
	C::Array{T},
	idx) where {T<:Dual} = contract!(T, sizeof(T)/sizeof(tovalue(T)), # tovalue unveils base type.
	A, 0, B, 0, C, 0, idx)

	contract!(Type::Dual{Tg, T, ND}, topst, # top-level stride
	A::Array, sftA, # arrays here are all at their top-level (not dispatched)
	B::Array, sftB,

	#include <iostream>
	#include <blis/blis.h>
	#include <boost/numeric/ublas/matrix.hpp>
	#include <boost/numeric/ublas/io.hpp>

	int main(const int argc, const char *argv[])
	{
	// these two doesn't conflict.
	using namespace std;
	using namespace boost::numeric::ublas;

	"""
	Tensordot.jl - Minimal Tensordot Implementation

	Minimal tensordot for supporting Zygote's automatic differentiation.
	This bunch of code is also compatible with FluxML's Tracker module.
	"""
	module Tensordot

	using Zygote: @adjoint
	using LinearAlgebra

	#!/bin/bash

	# 1. Installation location of the Wolfram Engine
	WOLFRAMROOT=/usr/local/Wolfram/WolframEngine/12.1
	# 2. MathLink path. Consider saving it for reasons in 4.
	MLINKROOT=$WOLFRAMROOT/SystemFiles/Links/MathLink/DeveloperKit/Linux-x86-64/CompilerAdditions

	# 3. Compile
	gcc -m64 -o ~/.TeXmacs/bin/tm_mathematica.bin \
	/usr/share/TeXmacs/plugins/mathematica/src.lazy/tm_mathematica.c \

RuQing Xu xrq-phys

1. Install macFUSE:

References