thomaswilley

the basics (zero to bootable)

# create disk
$ qemu-img create -f qcow2 ~/QEMU/ubuntu-desktop-18.04.qcow2 10G

# setup
$ qemu-system-x86_64 -m 2048 -vga virtio -show-cursor -usb -device usb-tablet -enable-kvm -cdrom ~/QEMU/ubuntu-18.04.2-desktop-amd64.iso -drive file=~/QEMU/ubuntu-desktop-18.04.qcow2,if=virtio -accel hvf -cpu host

# run (unmount CDROM)
$ qemu-system-x86_64 -m 2048 -vga virtio -show-cursor -usb -device usb-tablet -enable-kvm -drive file=~/QEMU/ubuntu-desktop-18.04.qcow2,if=virtio -accel hvf -cpu host

thomaswilley

alias ffshow='firefox "data:text/html;base64,$(base64 -w 0 <&0)"'

# usage: curl google.com | ffshow

thomaswilley

# record desktop to .mkv which is safely ended via ctrl+c
ffmpeg -y -f gdigrab -i desktop -r 10 -vcodec libx264 -pix_fmt yuv420p output.mkv
# clip some time (in this case 22s) from the start of the video and transcode to .mp4 (a format not safe to end w/ctrl+c)
ffmpeg -ss 00:00:22.0 -i output.mkv output.mp4
# create a palette
ffmpeg -y -i output.mp4 -vf fps=10,scale=768:-1:flags=lanczos,palettegen palette.png
# create the final gif
ffmpeg -i output.mp4 -i palette.png -filter_complex "fps=10,scale=1024:-1:flags=lanczos[x];[x][1:v]paletteuse" output.gif

thomaswilley

#!/bin/bash
# brute force scanner for sonos devices on a network
# (just looks for open port 1400: e.g., http://ip:1400/support/review)

echo Scanning network for Sonos devices...
for i in $(seq 1 255);
do
  IP=192.168.1.$i
  PORT=1400
  nc -n -z -G 1 $IP $PORT 2>&1 | awk '{print $3}'

thomaswilley

# Reference guide from Google: https://www.tensorflow.org/install/install_sources
# assumes python3 and pip3 installed (/usr/local/python3.6), along with homebrew
# python pkgs path: `python3 -m site` (/usr/local/lib/python3.6/site-packages)
# --copt=-mavx --copt=-mavx2 --copt=-mfma --copt=-msse4.1 --copt=-msse4.2 
# ^ these are some potential CPU optimizations but just allowing the ./configurewizard to do its thing and use -march=native
# also seems to work fine for just CPU optim.

$ brew install bazel # v.0.9.0
$ pip3 install numpy six wheel
$ git clone https://github.com/tensorflow/tensorflow.git

thomaswilley

DIR=`pwd` awk '{ gsub("\\$DIR",ENVIRON["DIR"]); print $0 }' infile > outfile

thomaswilley

lsof -i +c 0 | awk '{ print $1 }' | uniq

thomaswilley

# Get Apple Health data as Pandas DataFrame
# ===
# pre-reqs: python3, lxml, pandas
# to get started:
#   export and mail yourself your data following steps within the Health app on iPhone
#   download and unzip contents of exported zip file; find path to export.xml and set path_to_exportxml below

import pandas as pd
import xml.etree.ElementTree
import datetime

thomaswilley

grep -w install /var/log/dpkg.log | awk '{ if ($1 == "2013-08-22") { print $4 }}' | while read -r line; do sudo apt-get -y purge $line && sudo apt-get -y autoremove; done

thomaswilley

def csvfile_to_array(csvfilepath, delimiter=',', ignore_header=False, set_required_num_cols=(True, 4)):
    results = []
    with open(csvfilepath, 'r') as csvfile:
        ctareader = csv.reader(csvfile, delimiter=delimiter)
        if ignore_header:
            next(ctareader, None)
        for row in ctareader:
            if set_required_num_cols[0]:
                # hackishly doubly check for None and for cols in one pass
                if len([v for v in row if v]) == set_required_num_cols[1]: