Nick Porcino meshula

boost removal is an ongoing, low noise effort.

At this point, if you build without python, you can build the usd libraries treating boost as header only, without actually building boost. one or two of the command line tools use boost for command line options.

Porting those to (my suggestion) CLI11 would remove the last non-python boost dependency.

to be redundant and clear: you can manually cherrypick the boost headers

Crusty	Rusty	Source	Home
cat	bat	cargo install bat	https://github.com/sharkdp/bat
cp	fcp	cargo install fcp	https://github.com/Svetlitski/fcp
diff	git-delta	cargo install git-delta	https://github.com/dandavison/delta
find	fd	cargo install fd	https://github.com/sharkdp/fd
grep	rg	cargo install ripgrep	https://github.com/BurntSushi/ripgrep
ls	lsd	cargo install lsd	https://github.com/Peltoche/lsd
tree	lsd --tree	cargo install lsd	https://github.com/Peltoche/lsd

on Windows, pip largely works for PySide2, but PySide2 is poorly supported and mostly broken under conda. Mixing and matching pip and conda is a bad practice.
in a non-studio environment, Qt is problematic outside of the GPL3 environment.
usdview is great, but if Python, or Qt, or PySide2 provide challenges, the onboarding process to USD involves a large effort
no low-dependency usdview is available
boost is a barrier to adoption due to well documented versioning issues (e.g., it is in vfxplatform as a configuration difficulty) and an onboarding-unfriendly build system
tbb is a barrier outside of the DCC environment, as it is Yet Another threading system that may be at odds with either a developer's own tasking system, or a system native one such as Apple's Grand Central Dispatch. It is very easy to induces oversubscription problems with TBB in practice.
lack of standard cookbooks. there are multiple community efforts, but no acknowledged best practice sources
Python is highly adopted in Fil

Character Animation

[https://github.com/ethz-adrl/control-toolbox] Control Toolbox - An Open-Source C++ Library for Robotics, Optimal and Model Predictive Control
[https://github.com/davepagurek/DynamicKelvinlets] Dynamic Kelvinets to add secondary animation
[http://domedb.perception.cs.cmu.edu/monototalcapture.html] Yassar's group's monocular mocap
[https://arxiv.org/abs/1812.11103v3] Learning to Walk via Deep Reinforcement Learning
[https://arxiv.org/abs/1810.02363v4] Recurrent Transition Networks for Character Locomotion
[http://www.andreasaristidou.com/publications/papers/CUEDF-INFENG,%20TR-632.pdf] FABRIK
[https://github.com/dgriff777/a3c_continuous] A3 Continuous Character Controller

https://gitlab.com/theswiftdev/cfactorial/-/blob/master/Package.swift

https://theswiftdev.com/how-to-call-c-code-from-swift/

https://developer.apple.com/documentation/swift/imported_c_and_objective-c_apis/importing_objective-c_into_swift

https://github.com/apple/swift-package-manager/blob/main/Documentation/Usage.md#defining-dependencies

https://rderik.com/blog/making-a-c-library-available-in-swift-using-the-swift-package/


	class Marker {
	}

	class Element {
	var layers = [CGImage?]() // this will become an ImageIndex that refers to a jpg, tile, etc
	var markers = [Int]()
	var start_frame = Int(0)
	var end_frame_ex = Int(0)
	var in_frame = Int(0)

	# This is not original code, it's from https://github.com/ziglang/zig-pypi
	# this copy's purpose is to add comments in order to understand how it works

	import os
	import hashlib
	import urllib.request
	from email.message import EmailMessage
	from wheel.wheelfile import WheelFile, get_zipinfo_datetime
	from zipfile import ZipInfo, ZIP_DEFLATED
	import libarchive # from libarchive-c

	#include <windows.h>
	#include <stdio.h>
	#define DIRECTINPUT_VERSION 0x0800
	#include <dinput.h>
	#pragma comment(lib, "dinput8.lib")
	#pragma comment(lib, "dxguid.lib")


	class DiJoyStick {
	public:

	#include <iostream>
	#include <assert.h>
	#include <vector>

	template<class T>
	struct Func{};

	// std::func replacement without dynamic memory allocation (captures are limited to 48 bytes)
	// this could be extended to support captures >48 bytes... of course, with a bit more of logic
	// and some state.

	#!/bin/bash

	# This script, executed at the root of a git repository, deletes traces of every old file in this repository, index + blob on all branches
	# It can take 10-30 minutes to run and will print regular warning stating than some references are unchanged
	# time ./clear_git_repositor.bash >cleaning.log

	# We need several passes to clean files renamed multiple times (git log --find-renames prevents its deletion for each renaming)
	# MAXIMUM_PASSES should be more than the maximum number of renamings/movings for any file, if not then we might keep some traces of former files
	MAXIMUM_PASSES=10 # Maximum number of passes