Die Frage, ab wann „Last Christmas“ von „WHAM!“ das erste Mal im Jahr gespielt wird, lässt sich gar nicht so pauschal beantworten, wie ich es bei der Q&A vielleicht versucht hatte, darzustellen. Der Hintergrund davon ist, dass die Sender gerne aus Gag oder aus einem anderen Zusammenhang den Song auch schon im März oder Juli bereits „zum ersten Mal in diesem Jahr“ spielen:

Und hier zum Vergleich, wenn man ab dem 01.10. des jeweiligen Jahres erst zählt:

I’ve created via zfs a RAIDZ, which is roughly something like a RAID5 in the traditional way. It contained three disks and as time goes by, the resulting array (roughly double the size of a single disk) was nearly full.

I got an identical vendor and size disk like the first three disks and added it to the case. But the RAIDZ-expansion-feature is at the time of this writing still not completed and far from being included in Ubuntu releases.

So I came up with a plan to get a bigger array without loosing any data and without backing up all the data to an external system (I simply didn’t have *that* many spare disks).

mdadm has a „–grow“-feature, so I had to copy everything to a mdadm-Raid, but without any additional disks.

The plan looked like this:

Initial 3 disks in RAIDZ/zfs, set one as offline

Use the offline disk and the recently added new disk to a new mdadm (RAID-5), degraded from the start.

Copy everything from RAIDZ to mdadm.

Destroy the remaining zfspool and add the remaining two disks to mdadm, growing, reshaping and resyncing all in one big step.

The only possible setback would be a drive failure on ZFS after degrading the ZFS or a drive failure in the first two disks of the new mdadm. – This would always be a full disaster. So – fingers crossed – and everything worked fine. It took a few days (copy and resyncing is slow), but it finally worked.

To make sure that I don’t mess up, I created a demo-script to test the several steps and whether my idea worked at all.

It comes in two steps, the first one works until the start of resyncing the mdadm. The second one should be started AFTER the resyncing has finished. In the small demo files I used this happend very fast, in reality this could take days.

#!/bin/bash
mkdir -p /test
cd /test # Separates Verzeichnis
cd test/
rm -f 1.disk
rm -f 2.disk
rm -f 3.disk
rm -f 4.disk
losetup -D
umount /test/mnt
mdadm --stop /dev/md0
mdadm --remove /dev/md0
rm -f /test/backupfile.mdadm

echo "##### Creating images"
dd if=/dev/zero of=1.disk bs=1M count=256
dd if=/dev/zero of=2.disk bs=1M count=256
dd if=/dev/zero of=3.disk bs=1M count=256
dd if=/dev/zero of=4.disk bs=1M count=256
DISK1=$(losetup --find --show ./1.disk)
DISK2=$(losetup --find --show ./2.disk)
DISK3=$(losetup --find --show ./3.disk)
DISK4=$(losetup --find --show ./4.disk)
parted ./1.disk mklabel gpt
parted ./2.disk mklabel gpt
parted ./3.disk mklabel gpt
parted ./4.disk mklabel gpt
parted -a optimal -- ./1.disk mkpart primary 0% 100%
parted -a optimal -- ./2.disk mkpart primary 0% 100%
parted -a optimal -- ./3.disk mkpart primary 0% 100%
parted -a optimal -- ./4.disk mkpart primary 0% 100%

echo "##### Starting zfs pool on disk 1, 2, 3"
zpool create origtank raidz ${DISK1} ${DISK2} ${DISK3}

echo "##### zpool status"
zpool status -v origtank

echo "##### Creating test file on /origtank"
dd if=/dev/zero of=/origtank/data bs=1M count=300

echo "##### Setting third disk as faulty"
zpool offline origtank ${DISK3}

echo "##### zpool status"
zpool status -v origtank

echo "##### ls -lA /origtank; df -h /origtank"
ls -lA /origtank; df -h /origtank

echo "##### Creating new md0 from disk3 and disk4"
#parted -s ./3.disk mklabel gpt
#parted -s ./4.disk mklabel gpt
#parted -s -a optimal -- ./3.disk mkpart primary 0% 100%
#parted -s -a optimal -- ./4.disk mkpart primary 0% 100%
wipefs -a ${DISK3}
wipefs -a ${DISK4}
parted -s ${DISK3} set 1 raid on 
parted -s ${DISK4} set 1 raid on 
mdadm --create /dev/md0 -f --auto md --level=5 --raid-devices=3 ${DISK3} ${DISK4} missing

echo "##### mdstat"
cat /proc/mdstat
mdadm --detail /dev/md0

echo "## # ## Formatting /dev/md0"
sleep 2
mkfs.ext4 /dev/md0

echo "##### Mount md0"
mkdir /test/mnt
mount /dev/md0 /test/mnt

echo "##### ls -lA /test/mnt; df -h /test/mnt"
ls -lA /test/mnt; df -h /test/mnt

echo "## # ## Copy data"
sleep 2
# rsync --delete -avPH /origtank/ /test/mnt
rsync -avPH /origtank/ /test/mnt

echo "##### ls -lA /test/mnt; df -h /test/mnt"
ls -lA /test/mnt; df -h /test/mnt

echo "##### Creating NEW test file on /origtank"
dd if=/dev/zero of=/origtank/dataNEW bs=1M count=30

echo "## # ## Copy NEW data"
sleep 2
# rsync --delete -avPH /origtank/ /test/mnt
rsync -avPH /origtank/ /test/mnt

echo "##### ls -lA /test/mnt; df -h /test/mnt"
ls -lA /test/mnt; df -h /test/mnt

echo "## # ## destroying pool"
sleep 2
zpool destroy origtank

echo "## # ## Adding disks to md0"
sleep 2
mdadm --add /dev/md0 ${DISK1} ${DISK2}
mdadm --grow --raid-devices=4 /dev/md0 --backup-file=/test/backupfile.mdadm
cat /proc/mdstat

After the successful resync of the mdadm, you can resize the filesystem:

resize2fs /dev/md0
cat /proc/mdstat

This only takes a few minutes (even on very huge disks), but be patient! You can see the progress by looking at mdadm –detail.

echo "##### mdstat"
cat /proc/mdstat
mdadm --detail /dev/md0

echo "##### ls -lA /test/mnt; df -h /test/mnt"
ls -lA /test/mnt; df -h /test/mnt

When I got stable-diffusion to work, I really wanted to see what the AI would do on its own, letting it describe a given picture and re-create the picture from this description. How detailed can or should a description be?

First install the „clip-interrogator“:

python3 -m pip install --upgrade pip setuptools wheel
sudo apt install -y rustc cargo
pip install clip-interrogator

And here’s the sample python file. I got an error when I didn’t reassign/clear the „pipe“.

import torch
from torch import autocast
from diffusers import StableDiffusionPipeline
from PIL import Image
from clip_interrogator import Config, Interrogator

print("### Starting Stable Diffusion Pipeline")
pipe = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4")
pipe.to("cuda")

prompt = "a dark environment, two warriors standing on a chess field with swords drawn"
# prompt = "a laptop sitting on top of a wooden table"
steps = 50
width = 512
height = 512

print("### Creating images with Stable Diffusion")
with autocast("cuda"):
  for i in range(1):
    output = pipe(prompt, width=width, height=height, num_inference_steps=steps)
    image = output["images"][0]
    file = prompt.replace(" ", "_").replace(",", "")
    image.save(f"{file}-{i}.png")

pipe = "" # Destroy?
print("### Initializing Interrogator")
ci = Interrogator(Config(clip_model_name="ViT-L-14/openai"))
file = prompt.replace(" ", "_").replace(",", "")
for i in range(1):
  print("Loading file ", i)
  image = Image.open(f"{file}-{i}.png").convert('RGB')
  print(ci.interrogate(image))

First install WSL2 as suggested by the many different websites out there. I installed it like this:

wsl --install -d Ubuntu-22.04

Open the installed Ubuntu and install the necessary packages for stable diffusion itself:

sudo apt update && sudo apt -y upgrade && sudo apt -y install git-lfs python3-pip
pip install torch --extra-index-url https://download.pytorch.org/whl/cu117
pip install diffusers transformers==4.26 scipy ftfy accelerate

Add CUDA-Support for the WSL (from this site):

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.0-1_all.deb
sudo dpkg -i cuda-keyring_1.0-1_all.deb
sudo apt-get update
sudo apt-get -y install cuda

Write yourself a small python file (e.g. as main.py) which will enable the image generation and save the images to a file with the prompt name:

import torch
from torch import autocast
from diffusers import StableDiffusionPipeline

pipe = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4")
pipe.to("cuda")

prompt = "a dark environment, two warriors standing on a chess field with swords drawn"
steps = 50
width = 512
height = 512

with autocast("cuda"):
  for i in range(4):
    output = pipe(prompt, width=width, height=height, num_inference_steps=steps)
    image = output["images"][0]
    file = prompt.replace(" ", "_").replace(",", "")
    image.save(f"{file}-{i}.png")

Eine alte Overleaf-Version lief bei mir lokal und diese wollte ich nun nach geraumer Zeit mal updaten. Installiert hatte ich die Version noch mithilfe des „docker-compose“, eine entsprechende Anleitung hatte ich hier bereits geschrieben.

Inzwischen hat sich overleaf etwas weiterentwickelt und man arbeitet mit dem Overleaf-Toolkit, welches inkompatibel zu meiner bisherigen Installation ist.

Zunächst musste ich von allem ein Backup erstellen, was aufgrund nicht-aktueller Beispiele von der Overleaf-Seite nicht trival war:

# Backup-Verzeichnis erstellen und hineinwechseln, hier als Beispiel:
mkdir -p ~/backupsharelatex/
cd ~/backupsharelatex/
# Die docker-compose.yml, welche zur Installation verwendet wurde, muss ebenfalls in dieses Verzeichnis. Danach kann das "exit 0" auskommentiert werden:
exit 0
# Backup mongodb
docker-compose exec mongo mongodump --archive="sharelatex-$(date +%Y%m%d).mongodb" --db="sharelatex"
docker cp mongo:sharelatex-$(date +%Y%m%d).mongodb .
tar cvfz ./sharelatex_data-$(date +%Y%m%d).tgz ~/sharelatex_data
docker-compose stop sharelatex

Dann führt man die Schritte aus dem Quick-Start-Guide aus. Hier muss man ggfs. noch die Einträge SHARELATEX_LISTEN_IP und SHARELATEX_PORT in der config/overleaf.rc ändern.

Am Ende (nachdem wir bin/up erfolgreich ausgeführt und wieder beendet haben) unsere alten MongoDB-Daten wieder zu importieren. Dazu brauchen wir im Backup-Verzeichnis wie bereits erwähnt die docker-compose.yml, sonst wird „mongo“ nicht gefunden.

docker cp ./sharelatex-$(date +%Y%m%d).mongodb mongo:sharelatex.mongodb
docker-compose exec mongo mongorestore --archive="sharelatex.mongodb"

Falls noch Daten im „data“-Verzeichnis eures alten sharelatex/data-Ordners liegen, müssen diese ebenfalls noch kopiert werden:

# "overleaf" ist hier das Verzeichnis welches man im "Quick-Start-Guide" angelegt hatte.
cd overleaf/data/sharelatex/data
sudo cp -r ~/sharelatex_data/data/* .

Jetzt ist alles wieder importiert und wir können das System über bin/start korrekt starten.

I ran into some trouble trying to backup and restore a InfluxDB to a new server. Here are the steps to correctly backup a database and restore it on the remote host.

On the source host, backup the InfluxDB you want while the influxdb is running:

# This will create a directory "backup" for DATABASE_NAME:
influxd backup -database DATABASE_NAME backup

Transfer the directory to the target machine and import the data (using default directories from influx):

sudo service influxdb stop # Restore works ONLY on stopped InfluxDB
sudo influxd restore -metadir /var/lib/influxdb/meta BACKUP_DIRECTORY
sudo influxd restore -database DATABASE_NAME -datadir /var/lib/influxdb/data BACKUP_DIRECTORY
sudo chown -R influxdb:influxdb /var/lib/influxdb
sudo service influxdb start

Ich finde overleaf extrem cool, wollte aber für vergleichsweise sensible Daten nicht die öffentlich angebotene Instanz nutzen, sondern diese lieber lokal einsetzen. Da overleaf seinen Quellcode frei verfügbar gestellt hat, ist das ziemlich einfach umzusetzen.

Ich habe mir hierfür eine einfache „install.sh“ erstellt, welche die notwendigen Schritte ausführt. Die Datei sollte man als nicht-priviligierter Nutzer ausführen. Falls Docker (in einer Nicht-Community-Edition) bereits installiert ist, aber nicht läuft würde dieses Skript das Docker automatisch durch die Community Edition ersetzen, also bitte ggf. den Teil aus dem Skript herausnehmen.

#!/bin/bash
EMAILADRESS="" # Set this FIRST!
if [ -z ${EMAILADRESS} ]
then
  echo "Please set a value for EMAILADRESS first. That user will be made admin."
  exit 1
fi
. /etc/lsb-release
if (systemctl -q is-active docker)
then
  echo "Docker is already running, will NOT install."
else
  echo "### Initial apt update"
  sudo apt update
  echo "### Installing prerequisites"
  sudo apt -y install apt-transport-https ca-certificates curl software-properties-common
  echo "### Adding docker GPG key"
  curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -
  echo "### Adding Repository"
  sudo add-apt-repository "deb [arch=amd64] https://download.docker.com/linux/ubuntu ${DISTRIB_CODENAME} stable"
  echo "### apt update for docker"
  sudo apt update
  echo "### Choosing docker-ce"
  sudo apt-cache policy docker-ce
  echo "### Installing docker-ce"
  sudo apt -y install docker-ce
  echo "### Checking im docker is running"
  if (systemctl -q is-active docker)
  then
    echo "### Docker is running"
  else
    echo "Docker is NOT running."
    exit 1
  fi
  echo "### Adding docker group"
  sudo groupadd docker
  newgrp docker
  sudo usermod -aG docker ${USER} ## Privilege user
fi
echo "### Installing docker-compose"
sudo apt -y install docker-compose
echo "### Pulling sharelatex-docker-image"
docker pull sharelatex/sharelatex
echo "### Pulling docker-compose.yml"
wget -nc https://raw.githubusercontent.com/overleaf/overleaf/master/docker-compose.yml
echo "### Modifying docker-compose.yml"
sed -i 's#- 80:80#- 2080:80#' docker-compose.yml # Only once, for the first "ports:"-directive
echo "### Running docker-compose"
docker-compose up -d sharelatex
docker exec sharelatex /bin/bash -c "cd /var/www/sharelatex; grunt user:create-admin --email=${EMAILADRESS}"
docker exec sharelatex wget http://mirror.ctan.org/systems/texlive/tlnet/install-tl-unx.tar.gz
docker exec sharelatex tar xvfz install-tl-unx.tar.gz
echo "#######################################################################"
echo "Execute 'cd install-tl-[0-9]*; ./install-tl'"
echo "Please choose 'y' to import settings, then choose 'I' for installation."
echo "Leave afterwards with 'exit'"
echo "#######################################################################"
docker exec -i -t sharelatex /bin/bash
docker exec sharelatex wget http://mirror.ctan.org/systems/texlive/tlnet/update-tlmgr-latest.sh -O /usr/local/texlive/2020/update-tlmgr-latest.sh
docker exec sharelatex chmod u+x /usr/local/texlive/2020/update-tlmgr-latest.sh
docker exec sharelatex /usr/local/texlive/2020/update-tlmgr-latest.sh -- --upgrade
docker exec sharelatex tlmgr update --self
docker exec sharelatex tlmgr install scheme-full
docker exec sharelatex luaotfload-tool -fu

Anschließend sollte auf dem aktuellen Host auf Port 2080 die Overleaf-Instanz erreichbar sein.

Die Installation von „scheme-full“ dauert eine ganze Weile, weil alles einzeln angestoßen wird. Bitte nicht ungeduldig werden, 1-2 Stunden je nach Rechnergeschwindigkeit sind hier völlig normal.

Falls merkwürdige Fehler auftauchen (z.B. dass die Silbentrennung von ngerman bzw. german nicht geladen werden kann), dann kann das an einer veralteten TeXLive-Installation im Docker-Container liegen. Bei mir war im vorgefertigen Docker noch TeXLive 2019 installiert, welches man manuell auf 2020 updaten kann (seit dem 07.04. muss dies ausgeführt werden, daher habe ich es in das Skript oben integriert, leider muss man durch das interaktiv „install-tl“-Skript die Fragen selbst beantworten):

docker exec -i -t sharelatex /bin/bash
wget http://mirror.ctan.org/systems/texlive/tlnet/install-tl-unx.tar.gz
tar xvfz install-tl-unx.tar.gz
cd install-tl-*
./install-tl

Die Fragen der Installation sollte man einfach so übernehmen, das „neue“ TeXLive wird dann in die Verzeichnisse des 2019er-TeXLive installiert. Im Idealfall kann man also direkt nach dem Starten von install-tlmit einem I die Installation ausführen. Das Upgrade dauert auch wieder lange, da wieder ca. 4000 Pakete installiert werden.

During a holiday I typed quite a few information in the „SNotes“-App on my Samsung Galaxy S8 phone. After my return, I couldn’t just easily copy the written text on my PC, I had to „export“ to sdoc and save them.

The trouble is, that sdoc-Files can’t be opened easily on a PC. So I figured it out on my own how to get the most of the information out of the file.

First of all you should know that every .sdoc-file is really only a .zip, so you can easily append .zip or change the .sdoc to .zip and extract the file. That way you have access to all data (like pictures or audio). The written text is included in the „text.dat“. Unfortunately, there are quite a few bytes inside the files that are unnecessary. Here’s my command I’m using to convert the included text to plain readable text. This even works with german umlauts:

xxd -p text.dat | sed 's/00//g; s/^1d12//;' | tr -d '\n' | sed 's/01010301/0a/' | xxd -r -p | sed 's/EOF$//' >text.txt

It removes all Zero-Bytes, the Header (1d12) and the seperator between the title and the body (01010301). The final „EOF“ is unnecessary as well and afterwards everything is written to text.txt, works like a charm.

Hier meine angepasste Version, die mit Ambibox (unter Windows) kommuniziert und ein Ambilight vergleichsweise einfach ermöglicht. Die LED-Anzahl und ggf. ein paar weitere Parameter müssen natürlich noch angepasst werden.

#include "FastLED.h"
#include "Arduino.h"

#define ANALOG_MODE_AVERAGE  0
#define ANALOG_MODE_LAST_LED 1
#define FASTLED_ESP8266_D1_PIN_ORDER

/**************************************
   S E T U P
   set following values to your needs
 **************************************/

#define INITIAL_LED_TEST_ENABLED true
#define INITIAL_LED_TEST_BRIGHTNESS 32  // 0..255
#define INITIAL_LED_TEST_TIME_MS 2000  // 10..

// Number of leds in your strip. set to "1" and ANALOG_OUTPUT_ENABLED to "true" to activate analog only
// As of 26/1/2017:
// 582 leaves ZERO bytes free and this
// 410 is ok
// tested with 500 leds and is fine (despite the warning)
#define MAX_LEDS 7+7+11+1

// type of your led controller, possible values, see below
//#define LED_TYPE APA102
#define LED_TYPE WS2812

// 3 wire (pwm): NEOPIXEL BTM1829 TM1812 TM1809 TM1804 TM1803 UCS1903 UCS1903B UCS1904 UCS2903 WS2812 WS2852
//               S2812B SK6812 SK6822 APA106 PL9823 WS2811 WS2813 APA104 WS2811_40 GW6205 GW6205_40 LPD1886 LPD1886_8BIT 
// 4 wire (spi): LPD8806 WS2801 WS2803 SM16716 P9813 APA102 SK9822 DOTSTAR

// For 3 wire led stripes line Neopixel/Ws2812, which have a data line, ground, and power, you just need to define DATA_PIN.
// For led chipsets that are SPI based (four wires - data, clock, ground, and power), both defines DATA_PIN and CLOCK_PIN are needed

// DATA_PIN, or DATA_PIN, CLOCK_PIN
#define LED_PINS D2        // 3 wire leds
// #define LED_PINS D3, D2  // 4 wire leds

// #define COLOR_ORDER RBG  // colororder of the stripe, set RGB in hyperion
// #define COLOR_ORDER BGR  // colororder of the stripe, set RGB in hyperion
#define COLOR_ORDER GRB  // colororder of the stripe, set RGB in hyperion

#define OFF_TIMEOUT 0    // ms to switch off after no data was received, set 0 to deactivate

// analog rgb uni color led stripe - using of hyperion smoothing is recommended
// ATTENTION  this pin config is default for atmega328 based arduinos, others might work to
//            if you have flickering analog leds this might be caused by unsynced pwm signals
//            try other pins is more or less the only thing that helps
#define ANALOG_OUTPUT_ENABLED false
#define ANALOG_MODE           ANALOG_MODE_LAST_LED  // use ANALOG_MODE_AVERAGE or ANALOG_MODE_LAST_LED
#define ANALOG_GROUND_PIN     8                     // additional ground pin to make wiring a bit easier
#define ANALOG_RED_PIN        9
#define ANALOG_GREEN_PIN      10
#define ANALOG_BLUE_PIN       11

// overall color adjustments
#define ANALOG_BRIGHTNESS_RED   255              // maximum brightness for analog 0-255
#define ANALOG_BRIGHTNESS_GREEN 255              // maximum brightness for analog 0-255
#define ANALOG_BRIGHTNESS_BLUE  255              // maximum brightness for analog 0-255

#define BRIGHTNESS 255                      // maximum brightness 0-255
#define DITHER_MODE BINARY_DITHER           // BINARY_DITHER or DISABLE_DITHER
#define COLOR_TEMPERATURE CRGB(255,255,255) // RGB value describing the color temperature
#define COLOR_CORRECTION  TypicalLEDStrip   // predefined fastled color correction
//#define COLOR_CORRECTION  CRGB(255,255,255) // or RGB value describing the color correction

// Baudrate, higher rate allows faster refresh rate and more LEDs
//#define serialRate 460800      // use 115200 for ftdi based boards
#define serialRate 115200     // use 115200 for ftdi based boards
//#define serialRate 500000         // use 115200 for ftdi based boards


/**************************************
   A D A L I G H T   C O D E
   no user changes needed
 **************************************/

// Adalight sends a "Magic Word" (defined in /etc/boblight.conf) before sending the pixel data
uint8_t prefix[] = {'A', 'd', 'a'}, hi, lo, chk, i;

unsigned long endTime;

// Define the array of leds
CRGB leds[MAX_LEDS];

// set rgb to analog led stripe
void showAnalogRGB(const CRGB& led) {
  if (ANALOG_OUTPUT_ENABLED) {
    byte r = map(led.r, 0,255,0,ANALOG_BRIGHTNESS_RED);
    byte g = map(led.g, 0,255,0,ANALOG_BRIGHTNESS_GREEN);
    byte b = map(led.b, 0,255,0,ANALOG_BRIGHTNESS_BLUE);
    analogWrite(ANALOG_RED_PIN  , r);
    analogWrite(ANALOG_GREEN_PIN, g);
    analogWrite(ANALOG_BLUE_PIN , b);
  }
}

// set color to all leds
void showColor(const CRGB& led) {
  #if MAX_LEDS > 1 || ANALOG_OUTPUT_ENABLED == false
  LEDS.showColor(led);
  #endif
  showAnalogRGB(led);
}

// switch of digital and analog leds
void switchOff() {
  #if MAX_LEDS > 1 || ANALOG_OUTPUT_ENABLED == false
  memset(leds, 0, MAX_LEDS * sizeof(struct CRGB));
  FastLED.show();
  #endif
  showAnalogRGB(leds[0]);
}

// function to check if serial data is available
// if timeout occured leds switch of, if configured
bool checkIncommingData() {
  boolean dataAvailable = true;
  while (!Serial.available()) {
    if ( OFF_TIMEOUT > 0 && endTime < millis()) {
      switchOff();
      dataAvailable = false;
      endTime = millis() + OFF_TIMEOUT;
    }
  }

  return dataAvailable;
}

// main function that setups and runs the code
void setup() {
  Serial.begin(serialRate);

  // analog output
  if (ANALOG_OUTPUT_ENABLED) {
    // additional ground pin to make wiring a bit easier
    pinMode(ANALOG_GROUND_PIN, OUTPUT);
    digitalWrite(ANALOG_GROUND_PIN, LOW);
    pinMode(ANALOG_BLUE_PIN , OUTPUT);
    pinMode(ANALOG_RED_PIN  , OUTPUT);
    pinMode(ANALOG_GREEN_PIN, OUTPUT);
  }

  int ledCount = MAX_LEDS;
  if (ANALOG_MODE == ANALOG_MODE_LAST_LED) {
    ledCount--;
  }

  #if MAX_LEDS > 1 || ANALOG_OUTPUT_ENABLED == false
    FastLED.addLeds<LED_TYPE, LED_PINS, COLOR_ORDER>(leds, ledCount);
  #endif
  
  // color adjustments
  FastLED.setBrightness ( BRIGHTNESS );
  FastLED.setTemperature( COLOR_TEMPERATURE );
  FastLED.setCorrection ( COLOR_CORRECTION );
  FastLED.setDither     ( DITHER_MODE );

  // initial RGB flash
  #if INITIAL_LED_TEST_ENABLED == true
  for (int v=0;v<INITIAL_LED_TEST_BRIGHTNESS;v++)
  {
    showColor(CRGB(v,v,v));  
    delay(INITIAL_LED_TEST_TIME_MS/2/INITIAL_LED_TEST_BRIGHTNESS);
  }
 
  for (int v=0;v<INITIAL_LED_TEST_BRIGHTNESS;v++)
  {
    showColor(CRGB(v,v,v));  
    delay(INITIAL_LED_TEST_TIME_MS/2/INITIAL_LED_TEST_BRIGHTNESS);
  }
  #endif
  showColor(CRGB(0, 0, 0));

  Serial.print("Ada\n"); // Send "Magic Word" string to host


  boolean transmissionSuccess;
  unsigned long sum_r, sum_g, sum_b;

  // loop() is avoided as even that small bit of function overhead
  // has a measurable impact on this code's overall throughput.
  for(;;) {
    // wait for first byte of Magic Word
    for (i = 0; i < sizeof prefix; ++i) {
      // If next byte is not in Magic Word, the start over
      if (!checkIncommingData() || prefix[i] != Serial.read()) {
        i = 0;
      }
    }

    // Hi, Lo, Checksum
    if (!checkIncommingData()) continue;
    hi = Serial.read();
    if (!checkIncommingData()) continue;
    lo = Serial.read();
    if (!checkIncommingData()) continue;
    chk = Serial.read();

    // if checksum does not match go back to wait
    if (chk != (hi ^ lo ^ 0x55)) continue;

    memset(leds, 0, MAX_LEDS * sizeof(struct CRGB));
    transmissionSuccess = true;
    sum_r = 0;
    sum_g = 0;
    sum_b = 0;

    int num_leds = min ( MAX_LEDS, (hi<<8) + lo + 1 );

    // read the transmission data and set LED values
    for (int idx = 0; idx < num_leds; idx++) {
      byte r, g, b;
      if (!checkIncommingData()) {
        transmissionSuccess = false;
        break;
      }
      r = Serial.read();
      if (!checkIncommingData()) {
        transmissionSuccess = false;
        break;
      }
      g = Serial.read();
      if (!checkIncommingData()) {
        transmissionSuccess = false;
        break;
      }
      b = Serial.read();
      leds[idx].r = r;
      leds[idx].g = g;
      leds[idx].b = b;
      #if ANALOG_OUTPUT_ENABLED == true && ANALOG_MODE == ANALOG_MODE_AVERAGE
          sum_r += r;
          sum_g += g;
          sum_b += b;
      #endif
    }

    // shows new values
    if (transmissionSuccess) {
      endTime = millis() + OFF_TIMEOUT;
      #if MAX_LEDS > 1 || ANALOG_OUTPUT_ENABLED == false
      FastLED.show();
      #endif

      #if ANALOG_OUTPUT_ENABLED == true
        #if ANALOG_MODE == ANALOG_MODE_LAST_LED
          showAnalogRGB(leds[MAX_LEDS-1]);
        #else
          showAnalogRGB(CRGB(sum_r/MAX_LEDS, sum_g/MAX_LEDS, sum_b/MAX_LEDS));
         #endif
      #endif
    }
  }
} // end of setup

void loop() {
  // Not used. See note in setup() function.
}

I have a dual-boot setup on my computer, the Windows 10 was already installed and is locked to the hardware („digital license“). After resizing the Windows partition, I was able to install ubuntu as a secondary operation system. I could change between system after a reboot, but I was unsatisfied that I had to reboot from Linux to Windows only for a short usage of the latter.

So I installed VirtualBox and instead of creating a new virtual Windows 10 which would require a different license, I figured out how to use the raw partition with the activated windows 10 within VirtualBox.

VirtualBox does support access to the raw disk, but there are some ceavats that need to be considered.

First, you’ll have to figure out which partition are EFI and belong the Windows 10 as well. Using gparted, you’ll quickly find the partitons which are labeled like this:

EFI system partition (fat32)
Microsoft reserved partition (unknown)
Basic data partiton (ntfs)

Using these partition information, you can now create a „raw vmdk“ file which can be used by VirtualBox. If the mentioned partitions are for example /dev/sda1, /dev/sda2 and /dev/sda3, issue this command (compensate for different device names, if necessary):

sudo VBoxManage internalcommands createrawvmdk -filename win10native.vmdk -rawdisk /dev/sda -partitions 1,2,3 -relative

I have named the raw file here „win10native.vmdk“, it will create a secondary file with „-pt.vdmk“ at the end, we won’t be using it here, but please leave it where it is. Now change the ownership of these two files to the user you want to use VirtualBox with (change „user“ and „group“ accordingly):

sudo chown -R user:group win10native*

Your current user must have disk access to continue. Beware, this literally means that this user is essentially given root rights as he can access all data on any disc. (Change „user“ accordingly)

sudo usermod -aG disk,vboxusers user

Enable „EFI“-Support in your System Settings of your VirtualBox.

If you would boot now your new virtual machine with the win10native.vmdk attached, Windows 10 can be booted, but as it detects different hardware, it will complain that it is not activated. To circumvent this, we have to read your system-uuid and write it to the virtual machine. The following command will print you the ID:

sudo dmidecode -s system-uuid

Modify your VM to change the system-uuid to the value you have just received (correct the XXX accordingly):

VBoxManage modifyvm Win10native --hardwareuuid XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX

If you boot now, you’ll most certainly only see a shell from „GRUB 2.02“ or something similar, as grub can’t find the Windows bootloader, this information is missing inside the UEFI from VirtualBox (and grub can’t find its root partition). To test your windows installation and activation, simply type these commands (you might have to change the „(hd0,gpt1)“-part – in fact, as the „root“ variable was already pointing to the correct device, I didn’t need to change it at all) and you can boot into windows. Remember that you only have the US-keyboard-layout.

insmod chain
set root=(hd0,gpt1)
chainloader /EFI/Microsoft/Boot/bootmgfw.efi
boot 

And here’s the result with an activated Windows 10 within Virtual Box. I can still boot native into Windows 10 without any further trouble. Don’t forget to install the VirtualBox Guest Addins for more comfort.

To circumvent the grub shell, you’ll have to enter the UEFI of VirtualBox. Boot the VM and immediatly (!) hit the F12-Button to enter the settings.

Add Boot Option and select the entry beginning with „SYSTEM“, this will let you traverse the directory path. Navigate to EFI/Microsoft/Boot/bootmgfw.efi and press ENTER. Add something at „Input the description“, press „F10“ and acknowledge with „Y“ then commit your changes.

In Boot -> Change Boot Order, select the EFI-Entry together with your newly created entry, highlight your entry and press „+“ until it is on the very top.

As the last step, „Reset system“ and it will boot Windows 10 directly. The changes made here a NOT permanent and will be lost the next time you boot (hey, at least you don’t have to type all those command on a us-layout-keyboard). This is a known issue by VirtualBox.

Natürlich können die Bootpartitionen und die verschlüsselte Partition völlig andere Namen haben.

cryptsetup luksOpen /dev/nvme0n1p6 nvme0n1p6_crypt
mount /dev/mapper/nvme0n1p6_crypt /mnt
# boot und boot/efi können nun unter /mnt/etc/fstab ausgelesen werden
mount /dev/nvme0n1p5 /mnt/boot
mount /dev/nvme0n1p1 /mnt/boot/efi
cp /proc/mounts /mnt/etc/mtab
mount -o bind /dev /mnt/dev
mount -o bind /sys /mnt/sys
mount -t proc /proc /mnt/proc
chroot /mnt /bin/bash
update-initramfs -k all -c
grub-install
update-grub

Nach und nach dies hier ausführen. Die Befehle am Ende, die mit einem # gekennzeichnet sind, müssen innerhalb der chroot-Umgebung ausgeführt werden.

/dev/nvme1n1p6 ist das mit cryptsetup bereits verschlüsselte root-Dateisystem und muss ggfs. hier angepasst werden. Dasselbe gilt für die Boot- und EFI-Partitionen.

#!/bin/bash
cryptsetup luksOpen /dev/nvme1n1p6 nvme1n1p6_crypt
mount /dev/mapper/nvme1n1p6_crypt /mnt

mount /dev/nvme1n1p5 /mnt/boot
mount /dev/nvme0n1p1 /mnt/boot/efi
cp /proc/mounts /mnt/etc/mtab
mount -o bind /dev /mnt/dev
mount -o bind /sys /mnt/sys
mount -t proc /proc /mnt/proc
chroot /mnt /bin/bash
# update-initramfs -k all -c
# mount -t efivarfs none /sys/firmware/efi/efivars
# grub-install --target=x86_64-efi --efi-directory=/boot/efi/ --bootloader-id=grub
# update-grub

I wanted my Rasperry Pi 2, which was running LibreELEC, to additionally scan for specific network devices. I didn’t want to change the LibreELEC, as I was very satisfied with it. So here’s how I managed to cross-compile on my linux box the tool `arp-scan`.

1.) Get the LibreELEC.tv-Source and compile an image.

Make sure that this works before continuing with anything else. I managed to compile the 9.0-devel version.

Instructions are given on the LibreELEC-Wiki

2.) Create the arp-scan directory and the package.mk

mkdir -p packages/tools/arp-scan
cat >packages/tools/arp-scan/package.mk <<EOF
################################################################################
#      This file is part of LibreELEC - https://libreelec.tv
#      Copyright (C) 2018-present Team LibreELEC
#
#  LibreELEC is free software: you can redistribute it and/or modify
#  it under the terms of the GNU General Public License as published by
#  the Free Software Foundation, either version 2 of the License, or
#  (at your option) any later version.
#
#  LibreELEC is distributed in the hope that it will be useful,
#  but WITHOUT ANY WARRANTY; without even the implied warranty of
#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
#  GNU General Public License for more details.
#
#  You should have received a copy of the GNU General Public License
#  along with LibreELEC.  If not, see <http://www.gnu.org/licenses/>.
################################################################################
 
PKG_NAME="arp-scan"
PKG_VERSION="1.9"
PKG_ARCH="arm"
PKG_LICENSE="GPL"
PKG_SITE="http://www.nta-monitor.com/wiki/index.php/Arp-scan_Installation_Guide"
PKG_URL="https://github.com/royhills/arp-scan/releases/download/$PKG_VERSION/arp-scan-$PKG_VERSION.tar.gz"
PKG_DEPENDS_TARGET="toolchain zlib libpcap"
PKG_SECTION="tools"
PKG_SHORTDESC="arp-scan"
PKG_LONGDESC="The arp-scan utility"
PKG_TOOLCHAIN="auto"
 
post_patch() {
  rm $PKG_BUILD/configure
  cd $PKG_BUILD
  autoconf
}
EOF

3.) Add needed patches

To get a patch-File from a github-commit, you can just add `.patch` to the url and you’ll get a correct patchable diff.

mkdir -p packages/tools/arp-scan/patches

I needed the change for the cross compilation, but the author hadn’t released a new version yet, so I had to download and include the patch:

wget https://github.com/royhills/arp-scan/commit/f74edaf821f49652b7649ff6113fc6685d2c952a.patch -O packages/tools/arp-scan/patches/Assume-long-long-int-format-is-lld-if-cross-compiling.patch

Sadly this won’t help with the release file, as the configure script already exists. So we have to delete the configure-File to force the package-System to run „autoconf“. That’s the reason for the `post_patch()`-Section in the package.mk.

4.) Add ourself to the dependency tree

sed -i 's/PKG_DEPENDS_TARGET="toolchain connman netbase ethtool openssh"/PKG_DEPENDS_TARGET="toolchain connman netbase ethtool openssh arp-scan"/' ./packages/virtual/network/package.mk

5.) Copy resulting file to host
In this case, we have to copy the libpcap.a and the arp-scan binary, like this:

scp build.LibreELEC-RPi2.arm-9.0-devel/libpcap-1.7.4/.armv7ve-libreelec-linux-gnueabi/libpcap.a targethost:
scp build.LibreELEC-RPi2.arm-9.0-devel/arp-scan-1.9/.armv7ve-libreelec-linux-gnueabi/arp-scan targethost:

I wanted to create a headless VirtualBox on a remote server. I did this before but didn’t remember the correct sequence, so I thought I’d better save the work for the next time and write my steps down:

VMNAME="newvm"
mkdir -p "/opt/VirtualBox VMs/${VMNAME}"
cd "/opt/VirtualBox VMs/${VMNAME}"
# You may have to change the next line:
wget -nc http://releases.ubuntu.com/18.04/ubuntu-18.04-desktop-amd64.iso
# Create a disk size that would fit your operation system. Here I chose 32768 MB = 32 GB.
VBoxManage createhd --filename "${VMNAME}".vdi --size 32768
# Create the VM itself. The available ostypes can be listed by issuing 'VBoxManage list ostypes'
VBoxManage createvm --name "${VMNAME}" --ostype "Ubuntu_64" --register
# Add a SATA controller with the dynamic disk attached.
VBoxManage storagectl "${VMNAME}" --name "SATA Controller ${VMNAME}" --add sata --controller IntelAHCI
VBoxManage storageattach "${VMNAME}" --storagectl "SATA Controller ${VMNAME}" --port 0 --device 0 --type hdd --medium ${VMNAME}.vdi
# Add an IDE controller with a DVD drive attached, and the install ISO inserted into the drive:
VBoxManage storagectl "${VMNAME}" --name "IDE Controller ${VMNAME}" --add ide
VBoxManage storageattach "${VMNAME}" --storagectl "IDE Controller ${VMNAME}" --port 0 --device 0 --type dvddrive --medium ubuntu-18.04-live-server-amd64.iso
# Memory and graphics settings
VBoxManage modifyvm "${VMNAME}" --memory 4096 --vram 128
# If you want the network to be nat'ed:
VBoxManage modifyvm "${VMNAME}" --nic1 nat
 
# Disable audio
VBoxManage modifyvm "${VMNAME}" --audio none
 
# Enable ssh-access
VBoxManage modifyvm "${VMNAME}" --natpf1 "guestssh,tcp,127.0.0.1,60022,,22"
# If you ever want to remove this:
# VBoxManage modifyvm "${VMNAME}" --natpf1 delete "guestssh"
 
# Enable VRDE / You need this especially for the installation process
VBoxManage modifyvm "${VMNAME}" --vrde on
VBoxManage modifyvm "${VMNAME}" --vrdeaddress 127.0.0.1
 
# Extensionpack to access via RDP etc.
wget -nc https://download.virtualbox.org/virtualbox/5.2.12/Oracle_VM_VirtualBox_Extension_Pack-5.2.12.vbox-extpack
VBoxManage extpack install Oracle_VM_VirtualBox_Extension_Pack-5.2.12.vbox-extpack
 
 
# Bootup:
VBoxHeadless -s "${VMNAME}"
# It should be available via RDP on local port 3389.
 
# Shutdown (three methods, best to worst):
# 1.) Inside of the virtualbox
# 2.) virtual powerbutton
VBoxManage controlvm "${VMNAME}" acpipowerbutton
# 3.) virtual power off
VBoxManage controlvm "${VMNAME}" poweroff
 
# If the installation is complete, shutdown the VirtualBox and eject the iso:
VBoxManage storageattach "${VMNAME}" --storagectl "IDE Controller ${VMNAME}" --port 0 --device 0 --type dvddrive --medium none

If you want to secure your virtual machine even more (after logging into another openvpn, for example), you can use ufw, but remember that the connection comes from the VirtualBox Interface:

sudo ufw allow from 10.0.2.0/24 to any

Inspired by these posts on the internet:

• https://www.perkin.org.uk/posts/create-virtualbox-vm-from-the-command-line.html
• https://forums.virtualbox.org/viewtopic.php?f=6&t=65975
• https://askubuntu.com/questions/42482/how-to-safely-shutdown-guest-os-in-virtualbox-using-command-line
• http://arunnsblog.com/2010/07/20/nat-with-port-forwarding-for-virtual-box/
• https://www.reddit.com/r/virtualbox/comments/5lze8p/remove_sound_card_with_vboxmanage/

Updating from 14.04.5 LTS to 16.04.3 LTS can be tricky in some details. The most parts will run fine during the update, but for mysql there simply isn’t a migration path from 5.5 to 5.7 directly.

Before anything else, backup your data first!

mysqldump --lock-all-tables -u root -p --all-databases > backup.sql

First you’ll have to upgrade vom 5.5 to 5.6, this happens pretty straightforward via the provided packages:

sudo apt update
sudo apt install mysql-server-5.6 mysql-client-5.6 mysql-server-core-5.6 mysql-client-core-5.6

The trouble is, that the packages for 5.7 are not available on Ubuntu 14.04.5 LTS. If you want to transfer the data from a host running 14.04.5 to another host running on 16.04.3, you’ll have to upgrade to 5.7 first. If you want to upgrade the host currently running your mysqld, you can just continue with do-release-upgrade, the upgrade to mysql 5.7 handles the database structure upgrade just fine.

To upgrade to mysql 5.7, you’ll have to leave the supported upgrade path from ubuntu 14.04 and insert your own apt-source from mysql.

Download the latest mysql-apt-config_w.x.y-z_all.deb from https://dev.mysql.com/downloads/repo/apt/, mine was mysql-apt-config_0.8.9-1_all.deb.

sudo gdebi mysql-apt-config_0.8.9-1_all.deb
sudo apt update
sudo apt install mysql-server
mysql_upgrade -u root -p
sudo service mysql restart

If you get errors like „Couldn’t execute ‚SHOW VARIABLES LIKE ‚gtid\_mode“: Native table ‚performance_schema‘.’session_variables‘ has the wrong structure“, you forgot to restart your mysql-server.

This answer from askubuntu.com helped me work out the details.

Raspberry Pis sind berüchtigt dafür, dass sie die Inhalte verwendeter SD-Karten gerne mal beschädigen. Häufig ist der Grund dahinter, dass der Strom getrennt wurde, während der Pi versuchte, auf die SD-Karte schreibend zuzugreifen.

Sofern man nur eine einfache Anwendung hat, welche keine lokalen Daten schreiben muss, kann man die SD-Karte auch direkt „readonly“ mounten, und damit den schreibenden Zugriff komplett unterbinden.

Es ist unerheblich, ob man auf dem Raspberry Pi ein klassisches Raspbian oder ein minimal-Image nutzt. Allerdings wird es sehr viel einfacher, wenn man auf eine grafische Benutzeroberfläche verzichten kann.

Ich habe bei zwei Anwendungsfällen hierzu bereits Erfolg gehabt. Bei dem einen wurde der eingesetzte Raspberry Pi als Steuereinheit für ein Flipdot im Chaostreff Dortmund eingesetzt. Bei dem anderen sollte der Raspbery Pi ausschließlich im Kiosk-Modus eine (lokal gehostete) Webseite anzeigen. Für den letzeren Fall griff ich auf Chromium zurück, welcher sich gut über Kommandozeilenparameter anpassen ließ.

Zunächst müssen wir das System vorbereiten und ein wenig verkleinern. Sofern ein klassisches Raspbian verwendet wird, sollte man vorher mit sudo su - auf den root-Benutzer wechseln, um die nachfolgenden Befehle ausführen zu können. Um möglichst kompatibel in den Befehlen zu bleiben, führe ich das sudo nicht immer mit aus, da bei minimal-Linuxen sudo gar nicht erst als Paket eingebunden ist und man sowieso nur als root arbeiten kann.

Als erstes sollte sichergestellt sein, dass die gesamte SD-Karte auch verwendet wird. In raspi-config kann man dies einstellen.

apt-get update
apt-get -y upgrade
reboot

Der reboot ist notwendig, damit neuere Kernel bereits geladen sind.

Nachfolgend werden alle grafischen Bentuzerelemente entfernt sowie der cron-Daemon, welcher sonst log-Einträge verursachen würde.

apt-get remove --purge wolfram-engine triggerhappy anacron logrotate dphys-swapfile xserver-common lightdm
insserv -r x11-common
apt-get autoremove --purge

Log-Messages werden noch im Speicher gehalten und können mit logread ausgelesen werden, nachdem man den „busybox“-syslogd verwendet:

apt-get install busybox-syslogd; dpkg --purge rsyslog

Die Einträge fastboot noswap ro fügen wir nun der cmdline.txt an, hierdurch wird das Dateisystem als nicht-schreibbar („ro“ = „readonly“) gesetzt und der Swap (die Auslagerungsdatei) abgeschaltet. Ressourcenintensive Speicherfresser können somit nicht mehr eingesetzt werden!

echo -n "fastboot noswap ro" >>/boot/cmdline.txt

Einige typische Dateien und Verzeichnisse müssen nun direkt auf das tmpfs (das ist die Ramdisk im Speicher des Raspberry Pis) verlinkt werden.

rm -rf /var/lib/dhcp/ /var/run /var/spool /var/lock /etc/resolv.conf
ln -s /tmp /var/lib/dhcp
ln -s /tmp /var/run
ln -s /tmp /var/spool
ln -s /tmp /var/lock
touch /tmp/dhcpcd.resolv.conf; ln -s /tmp/dhcpcd.resolv.conf /etc/resolv.conf

Die dhcpcd.pid muss ebenfalls auf das tmpfs zeigen, hierfür muss (zumindest auf dem Raspberry Pi 3) eine Datei angepasst werden.

sed -i /etc/systemd/system/dhcpcd5 's#PIDFile=/run/dhcpcd.pid#PIDFile=/var/run/dhcpcd.pid#'

Falls mit random-seed gearbeitet wird, muss dies ebenfalls ins tmpfs. Der Befehl ist ungefährlich, falls dies nicht der Fall ist und kann trotzdem ausgeführt werden:

rm /var/lib/systemd/random-seed
ln -s /tmp/random-seed /var/lib/systemd/random-seed

Das random-seed wird allerdings bei jedem Neustart benötigt, daher müssen wir die Datei mithilfe von systemd noch anlegen, wenn der Dienst gestartet wird.

sed -i /lib/systemd/system/systemd-random-seed.service 's#ExecStart=/lib/systemd/systemd-random-seed load#ExecStartPre=/bin/echo "" >/tmp/random-seed\nExecStart=/lib/systemd/systemd-random-seed load#'
systemctl daemon-reload

Abschließend müssen wir noch ein paar startup-Skripte loswerden:

insserv -r bootlogs; insserv -r console-setup

Da die SD-Karte üblicherweise mit „/dev/mmcblk0p1“ eingebunden wird, können wir mit einem kleinem Skript noch die /etc/fstab verändern, so dass „/boot“ und „/“ beim nächsten starten nur noch als Nur-Lesbar gemountet werden.

sed -i /etc/fstab 's#\(.*mmcblk.*defaults\)\(.*\)#\1,ro\2#'
cat <<EOT >>/etc/fstab
 
tmpfs           /tmp            tmpfs   nosuid,nodev         0       0
tmpfs           /var/log        tmpfs   nosuid,nodev         0       0
tmpfs           /var/tmp        tmpfs   nosuid,nodev         0       0
EOT

Um beim Einloggen auf den Raspberry Pi trotzdem Änderungen durchführen zu können, reichen zwei Kommandos ro und rw, welche man mit in die bash.bashrc hinzufügt. Falls man auf einem nicht-Multiuser-System unterwegs ist, muss man die „sudo“ natürlich noch vorher entfernen:

cat <<EOT >>/etc/bash.bashrc
# set variable identifying the filesystem you work in (used in the prompt below)
set_bash_prompt(){
    fs_mode=$(mount | sed -n -e "s/^\/dev\/.* on \/ .*(\(r[w|o]\).*/\1/p")
    PS1='\[\033[01;32m\]\u@\h${fs_mode:+($fs_mode)}\[\033[00m\]:\[\033[01;34m\]\w\[\033[00m\]\$ '
}
 
alias ro='sudo mount -o remount,ro / ; sudo mount -o remount,ro /boot'
alias rw='sudo mount -o remount,rw / ; sudo mount -o remount,rw /boot'
 
# setup fancy prompt"
PROMPT_COMMAND=set_bash_prompt
EOT

Und beim Ausloggen noch automatisiert zurück auf read-only wechseln. Dies speichert auch die history für den aktuellen Benutzer ab.

cat <<EOT >>/etc/bash.bash_logout
mount -o remount,rw /
history -a
mount -o remount,ro /
mount -o remount,ro /boot
EOT

Getting the xmr-stak-cpu to work on an ubuntu 14.04:

* Compile gcc from scratch and install to /home/user/inst (script)

#!/bin/bash
VER=7.2.0
cd
WORKDIR=${PWD}/work/gcc-${VER}
INSTDIR=${PWD}/inst
mkdir -p ${WORKDIR}
mkdir -p ${INSTDIR}
cd ${WORKDIR}
if [ ! -e gcc-${VER}.tar.gz ]; then
	wget -O gcc-${VER}.tar.gz ftp://ftp.gwdg.de/pub/misc/gcc/releases/gcc-${VER}/gcc-${VER}.tar.gz
fi
if [ ! -e gcc-${VER} ]; then
	tar xvfz gcc-${VER}.tar.gz
fi
cd gcc-${VER}
./contrib/download_prerequisites
# 64bit only
./configure --disable-multilib --prefix=${INSTDIR} && make -j $(grep -c ^processor /proc/cpuinfo) && make install

* Compile cmake from scratch

cd
INSTDIR=${PWD}/inst
mkdir -p work
cd work
mkdir -p cmake
cd cmake
wget https://cmake.org/files/v3.9/cmake-3.9.3.tar.gz
tar xvfz cmake-3.9.3.tar.gz
cd cmake-3.9.3
CC=${INSTDIR}/bin/gcc CXX=${INSTDIR}/bin/g++ PATH=${INSTDIR}/bin:$PATH ./configure --prefix=${INSTDIR} && LD_LIBRARY_PATH="$HOME/inst/lib64" make -j $(grep -c ^processor /proc/cpuinfo) && LD_LIBRARY_PATH="$HOME/inst/lib64" make install

* Download the xmr-stak-cpu files and compile

cd
INSTDIR=${PWD}/inst
mkdir -p $HOME/work/xmr
cd $HOME/work/xmr
git clone https://github.com/fireice-uk/xmr-stak-cpu
cd xmr-stak-cpu
LD_LIBRARY_PATH="$HOME/inst/lib64" CC=${INSTDIR}/bin/gcc CXX=${INSTDIR}/bin/g++ PATH=${INSTDIR}/bin:$PATH cmake . -DCMAKE_INSTALL_PREFIX=$HOME/xmr-stak-cpu -DMICROHTTPD_REQUIRED=OFF && LD_LIBRARY_PATH="$HOME/inst/lib64" make && LD_LIBRARY_PATH="$HOME/inst/lib64" make install

This will compile the file with the new gcc and the new cmake. The result will be installed to $HOME/xmr-stak-cpu, it’s only the executable file and the config.txt.

To run, you must point your LD_LIBRARY_PATH to the correct directory, otherwise you’ll get strange errors:

LD_LIBRARY_PATH="$HOME/inst/lib64" ./xmr-stak-cpu

Ich würde gerne meine Honeywell Heizkörperregler HR-20 Rondostat zeitgesteuert nicht nur passiv auslesen (das geht auch schon mit der nativen Firmware), sondern auch über einen ESP8266 steuern. Das wurde schon von schlauen Leuten vorgedacht, und daher gibt es mit OpenHR20 auch eine neue Firmware, welche man mittels JTAG-Interface auf das HR20 flashen kann.

Ich habe nur kein JTAG-Programmierer zur Hand, aber dafür einen Arduino Uno. Ich habe den Ansatz eines JTAGduino gefunden. Die notwendigen Lötarbeiten waren schnell erledigt, ich habe die nachfolgende Steckerbelegung (Quelle) berücksichtigt, um den JTAGduino zu versorgen. Um eine durchgehende Stiftleiste am Arduino Uno verwenden zu können, habe ich die gesamte Kommunikation auf die Analogen Ports A0 bis A4 gelegt.

Die ersten Tests mit den von JTAGduino mitgelieferten Skripten verliefen erfolgreich, wenn ich das JTAG-Interface vom Arduino an den HR20 angeschlossen habe und mein Arduino an meinen Linux-Rechner per USB verbunden ist, kommt folgende Ausgabe:

user@host:~/work/JTAGduino$ ./host_test.py 
if_ver_major = 0; if_ver_minor = 1
fw_ver_major = 0; fw_ver_minor = 1
set_serial_speed(115200) = 0
clear pin TDI rsp = 1
set pin TDI rsp = 0
get pin TDO rsp = 0; val = 0
jtag_clock(1,1): rsp = 1; tdo = 0
jtag_clock(1,0): rsp = 0; tdo = 1
jtag_sequence([1,1,1,1], [1, 1, 0, 1]), rsp = 0; tdo_seq = [1, 0, 0, 0]
jtag_sequence([1] * 255, [1] * 255), rsp = 0; tdo_seq = [1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]
device.set_jtag_speed(1) = 255
jtag_sequence([1] * 255, [1] * 255), rsp = 127; tdo_seq = []
user@host:~/work/JTAGduino$ ./readid.py 
Traceback (most recent call last):
  File "./readid.py", line 81, in 
    if rsp != 0: raise Exception('error: set_jtag_speed returned %d' % rsp)
Exception: error: set_jtag_speed returned 191
user@host:~/work/JTAGduino$ ./readid.py 
#Taps = 1
Device identifier registers (beginning from closest to TDO):
0x6940503F

Die Fehlermeldung habe ich absichtlich drin gelassen, damit man auch andere den Grund für den Fehler finden können. Man muss vor JEDEM Ansprechen des JTAGduino diesen manuell auf durch einen Druck auf den Reset-Button am Arduino zurücksetzen.

Die ID 0x6940503F ist bereits die vom HR20, wir können also grundsätzlich per JTAG bereits mit ihm kommunizieren.

Als nächstes muss ich noch herausfinden, wie avrdude mit dem so gebauten „Programmer“ sprechen kann, bisher habe ich nur ermitteln können, dass es eine mögliche Liste von Programmern in avrdude vordefiniert gibt: (Der Pfad zur Arduino-Installation muss natürlich angepasst werden)

user@host:~/work/JTAGduino$ PATH=$PATH:/arduino-1.8.1/hardware/tools/avr/bin avrdude -c dslfkj -C /arduino-1.8.1/hardware/tools/avr/etc/avrdude.conf 2<&1 | grep -i jtag
  atmelice         = Atmel-ICE (ARM/AVR) in JTAG mode (avrdude: jtag3_open_common(): JTAGICE3/EDBG port names must start with "usb")
  dragon_jtag      = Atmel AVR Dragon in JTAG mode
  jtag1            = Atmel JTAG ICE (mkI)
  jtag1slow        = Atmel JTAG ICE (mkI)
  jtag2            = Atmel JTAG ICE mkII
  jtag2avr32       = Atmel JTAG ICE mkII im AVR32 mode
  jtag2dw          = Atmel JTAG ICE mkII in debugWire mode
  jtag2fast        = Atmel JTAG ICE mkII
  jtag2isp         = Atmel JTAG ICE mkII in ISP mode
  jtag2pdi         = Atmel JTAG ICE mkII PDI mode
  jtag2slow        = Atmel JTAG ICE mkII
  jtag3            = Atmel AVR JTAGICE3 in JTAG mode
  jtag3dw          = Atmel AVR JTAGICE3 in debugWIRE mode
  jtag3isp         = Atmel AVR JTAGICE3 in ISP mode
  jtag3pdi         = Atmel AVR JTAGICE3 in PDI mode
  jtagkey          = Amontec JTAGKey, JTAGKey-Tiny and JTAGKey2
  jtagmkI          = Atmel JTAG ICE (mkI)
  jtagmkII         = Atmel JTAG ICE mkII
  jtagmkII_avr32   = Atmel JTAG ICE mkII im AVR32 mode
  o-link           = O-Link, OpenJTAG from www.100ask.net (avrdude: Error: no libftdi or libusb support. Install libftdi1/libusb-1.0 or libftdi/libusb and run configure/make again.)
  xil              = Xilinx JTAG cable (tested - avrdude: can't claim device "/dev/ttyACM0": Inappropriate ioctl for device)
  xplainedpro      = Atmel AVR XplainedPro in JTAG mode

Den nachfolgenden Befehl nutze ich zum Test, ob der jeweilige Programmer-Typ funktioniert, er sollte bei erfolgreicher Kommunikation nur Backups von fuses anlegen, also führt noch keine schreibenden Operationen aus:

user@host:~/work/JTAGduino$ PATH=$PATH:/arduino-1.8.1/hardware/tools/avr/bin avrdude -p m169 -c xil -P /dev/ttyACM0 -C /arduino-1.8.1/hardware/tools/avr/etc/avrdude.conf -U lfuse:r:backup/unknown/`date  "+%F_%T"`/lfuse.hex:h -U hfuse:r:backup/unknown/`date  "+%F_%T"`/hfuse.hex:h -U efuse:r:backup/unknown/`date  "+%F_%T"`/efuse.hex:h

So wie es aussieht, ist die Library nicht komplett, und für meinen Anwendungszweck vielleicht gar nicht geeignet. Ich bin dadurch über eine JTAG-Library gestoßen, welche man direkt über die Arduino-IDE nachinstallieren kann: https://github.com/mrjimenez/JTAG

Nach dem Upload des JTAGTest-Programms (wird als Beispiel in der ArduinoIDE bereitgestellt, wenn man die Bibliothek installiert hat) auf dem Arduino kann man das python-Programm „xsvf“ nutzen, um erste Tests durchzuführen:

user@host:~/work/JTAG/extras/python$ ./xsvf ../xsvf/all/idcode_simpler.xsvf
File: /work/JTAG/extras/xsvf/all/idcode_simpler.xsvf
Ready to send 18 bytes.
Sent:       18 bytes,        0 remaining
IMPORTANT: ********
IMPORTANT: Success!
IMPORTANT: ********
IMPORTANT: Last TDO: FF FF FF FF/32 bits
IMPORTANT: Processed 5 instructions.
IMPORTANT: Checksum:  0xC2/18.
IMPORTANT: Sum: 0x0000003E/18.
Quit: No error (0).
  Expected checksum:  0xC2/18.
  Expected sum: 0x0000003E/18.
Elapsed time: 0.02 seconds.

Mit der neuen JTAG-Library alle möglichen Programmiermodi des avrdude durchgegangen:

atmelice         = Atmel-ICE (ARM/AVR) in JTAG mode (avrdude: jtag3_open_common(): JTAGICE3/EDBG port names must start with "usb")
  dragon_jtag      = Atmel AVR Dragon in JTAG mode (no response)
  jtag1            = Atmel JTAG ICE (mkI) (avrdude: jtagmkI_open(): failed to synchronize to ICE)
  jtag1slow        = Atmel JTAG ICE (mkI) (avrdude: jtagmkI_open(): failed to synchronize to ICE)
  jtag2            = Atmel JTAG ICE mkII (no response)
  jtag2avr32       = Atmel JTAG ICE mkII im AVR32 mode (avrdude: jtagmkII_recv_frame(): timeout     avrdude: jtagmkII_getsync(): sign-on command: status -1)
  jtag2dw          = Atmel JTAG ICE mkII in debugWire mode
  jtag2fast        = Atmel JTAG ICE mkII
  jtag2isp         = Atmel JTAG ICE mkII in ISP mode
  jtag2pdi         = Atmel JTAG ICE mkII PDI mode
  jtag2slow        = Atmel JTAG ICE mkII
  jtag3            = Atmel AVR JTAGICE3 in JTAG mode
  jtag3dw          = Atmel AVR JTAGICE3 in debugWIRE mode
  jtag3isp         = Atmel AVR JTAGICE3 in ISP mode
  jtag3pdi         = Atmel AVR JTAGICE3 in PDI mode
  jtagkey          = Amontec JTAGKey, JTAGKey-Tiny and JTAGKey2
  jtagmkI          = Atmel JTAG ICE (mkI)
  jtagmkII         = Atmel JTAG ICE mkII
  jtagmkII_avr32   = Atmel JTAG ICE mkII im AVR32 mode
  o-link           = O-Link, OpenJTAG from www.100ask.net (avrdude: Error: no libftdi or libusb support. Install libftdi1/libusb-1.0 or libftdi/libusb and run configure/make again.)
  xil              = Xilinx JTAG cable (avrdude: can't claim device "/dev/ttyACM0": Inappropriate ioctl for device)
  xplainedpro      = Atmel AVR XplainedPro in JTAG mode (avrdude: jtag3_open_common(): JTAGICE3/EDBG port names must start with "usb")

Ich habe einen der Mikrofon-Sensoren ergattert, die man für Arduino bzw. Raspberry Pis günstig erwerben kann. Meiner war in einem „37 in 1“-Sensor-Kit enthalten. Schließt man diesen direkt mit A0 an den Arduino an, erhält man allerdings keine sinnvollen Werte, insbesondere weichen diese kaum von 2.5V Eingang ab. Die prinzipielle Schaltung lässt aber auf „D0“ das High bzw. Low erkennen, der Sensor selbst funktioniert also.

Um sinnvoll damit arbeiten zu können, musste ich zunächst einmal den Ausgang entkoppeln, dafür habe ich einen kleinen Kondensator (4.7 uF) angeschlossen. Das aus dem A0-Anschluss dieses Mikrofon-Breakouts erhaltene Signal ist nur zu schwach, muss also deutlich verstärkt werden. Ich habe einen MAX406 gefunden, der herumlag. Mit 470 Ohm als Vorwiderstand und 100 kOhm als Widerstand zwischen IN- und OUT sollte eine etwa 25-fache-Verstärkung herauskommen.

Das Signal aus dem Mikrofon komm ebenfalls an IN-, an IN+ habe ich GND angeschlossen.

Mit einem Piezo-Lautsprecher angeschlossen an GND und PIN9 vom Arduino Uno und das verstärkte Eingangssignal an A0 kann man nun gute das „Knackern“ hören, was vorher nur ein unangenehmes Piepsen war. Im seriellen Output kann man gut nachvollziehen, dass von der Schwallwelle nur eine Seite der Welle verstärkt wird, die andere ist praktisch immer Null und somit kommt auch noch nicht wirklich etwas verständliches heraus. Immerhin sind einfach Testtöne (Sinus-Wellen) darüber schon zu erkennen.

Ein möglicher nächster Schritt wäre nun:

Über einen (einstellbaren) Spannungsteiler die Referenzspannung, die an IN+ gegeben wird, zu verkleinern. Ferner muss das Eingangssignal durch einen 1:1-Spannungsteiler geschickt werden, um dann bei „keinem“ Signal etwa 50% Ausgangssignal zu liefern und dies dann beim Arduino wieder herauszurechnen. So würde das Signal vollständig verstärkt werden.

Den letzten „Developer’s Pack“ von WinPCap herunterladen:

http://www.winpcap.org/devel.htm

Auspacken und die Location merken, wo die Dateien nun liegen.

Wrapper herunterladen:

https://github.com/PcapDotNet/Pcap.Net/tree/master

Am einfachsten hierzu das aktuelle Repository als ZIP-Datei herunterladen und entpacken:

https://github.com/PcapDotNet/Pcap.Net/archive/master.zip

In dem ausgepackten Pfad diese Solution öffnen:

Pcap.Net-master\PcapDotNet\src\PcapDotNet.sln

• Auf „Release“ umstellen
• Auf „x64“ umstellen
• PcapDotNet.Core -> Rechtsklick -> Eigenschaften
  • VC++ Directories
    • Include Directories: Den Pfad vom WpdPack\Include hinzufügen
    • Library Directories: Den Pfad von WpdPack\Lib\x64 hinzufügen
• Alles kompilieren (Das Target Framework muss identisch mit dem im nächsten Schritt verwendeten sein)

Testweise etwas kompilieren:

Pcap.Net-master\PcapDotNet.DevelopersPack\src\PcapDotNet.DevelopersPack.sln öffnen

Bei „ObtainingTheDeviceList“ muss man die References anpassen, die Pfade müssen auf die „Relase“-Zweige der oben erstellten DLLs zeigen.

Das Resultat sollte ein Kommandozeilenfenster mit „1. rpcap://\Device…“ sein.

Recently I created a complex LEGO model with Lego Digital Designer and wanted to import the resulting model into Blender. My export files were missing the LEGO Mindstorms EV3-components that I had added.

The reason for this behaviour are the missing parts in the ldraw.xml-File, which is in the same directory as the ldd.exe, the directory C:\Program Files (x86)\LEGO Company\LEGO Digital Designer held this file for me.

The ldraw.xml is essentially a mapping file, which maps all the LEGO-internal structures to the more open „LDRAW“-File format. The EV3-parts weren’t listed and even in updated ldraw.xml-Files, which I could find on the web, these were never added.

So I created the missing mappings on my own and the result works pretty good. Not all the colours are correct, but I couldn’t find an easy way to correct these.

Here are the necessary steps to successfully export LEGO EV3-parts from Lego Digital Designer and import these with correct colours e.g. into Blender:

1. Copy ldraw.xml file to your location of ldd.exe, overwrite the old ldraw.xml. (Unzip the ldraw.zip to ldraw.xml first)
2. Export your model from Lego Digital Designer to „LDraw-Files (*.ldr)“-format.
3. Open the file in notepad++, search for the following part numbers (they are located at the end of the line) and change the BEGINNING of the line from „1 71“ to „1 15“.
   • Example:
     • 1 71 -90.175277709960937 -176.40634155273437 -275.99981689453125 0 1.0000001192092896 0 0 0 1.0000001192092896 1.0000001192092896 0 0 95658.dat
     • 1 15 -90.175277709960937 -176.40634155273437 -275.99981689453125 0 1.0000001192092896 0 0 0 1.0000001192092896 1.0000001192092896 0 0 95658.dat
   • Do this for all lines with these part numbers (not all might be in your file, only if you have added the corresponding EV3-bricks):
     • 95650.DAT
     • 95658.DAT
     • 95648.DAT
     • 99455.DAT
4. Import the corrected file in LeoCAD and export into „Wavefront (.obj)“
5. Open Blender and follow these instructions to get a first decent result:
   1. File -> Import -> Wavefront (.obj)
   2. Select -> Select all by Type -> Mesh
   3. Press „s“ (SCALE) and enter a value (blind, there is no dialog box), e.g. „0.1“ and hit return. Or you resize with your mouse after pressing „s“.
   4. Right-click and hold on the camera in blender, and move it around to where you want it to be.
   5. Check your current view by hitting the „0“ on your numpad. (This will change to the camera view)
   6. Right-Click during camera view on the border of the camera (it will turn orange), and then right-click and hold outside the camera view to tilt for your needs.
   7. Add some lights via Add -> Lamp -> Area  (or Sun) and position them.
   8. Render by pressing F12
   9. In the render-menu on the right hand side you can change the size of the resulting image, e.g. „Preset: HDTV 1080p“.
   10. By pressing F3 the rendered image is stored to your disc. (Current used path is at the bottom of the render-menu, defaults to /tmp.)

When trying to access my recordings on my NAS, kodi seems to stop working for a short while when the NAS realizes that it is being accessed and starts spinning up the discs. I figured that it would be nice for the NAS to spin up earlier, if the screen saver of a „connected KODI“ is deactivated.

After enabling the JSON-RPC-TCP 9090-Port on Kodi (In System/Settings/Network/Services activate „Allow programs on this system to control Kodi for localhost access only“ and „Allow programs on other systems to control Kodi for access from other computers as well“), this script should be run at boottime on the NAS. Make sure to replace the accessed discs with your /dev/sdX-entries on your system. Also replace HOSTNAME with the host which as Kodi running, if the discs are in use on the same host, replace it with „localhost“.

The command must be run as root.

onidlespinupdiscs.sh:

#!/bin/bash
# Must be run as root
while true
do
  netcat -d HOSTNAME 9090 | (
    cnt=0
    line=
    while read -N 1 c; do
      line="$line$c"
      if [ "$c" = "{" ]; then
        cnt=$((cnt+1))
      elif [ "$c" = "}" ]; then
        cnt=$((cnt-1))
        if [ $cnt -eq 0 ]; then
          printf "%s\n" "$line"
          line=
        fi
      fi
    done
  ) | while read line
  do
    match=$(echo ${line} | grep -c "GUI.OnScreensaverDeactivated")
    if [ $match -eq 1 ]; then
      dd if=/dev/sdb bs=4096 count=1 of=/dev/null iflag=direct >/dev/null 2>&1 &
      dd if=/dev/sdc bs=4096 count=1 of=/dev/null iflag=direct >/dev/null 2>&1 &
    fi
  done
  sleep 60
done

Zunächst einmal braucht man für die Programmierung der ATMEGA328P-PU einen „normalen“ Arduino. Ich habe mir bei ebay für wenig Geld aus Hong Kong einen Arduino-Klon besorgt. Es handelte sich dabei um einen „Pro.mini.328P“ („New Pro Mini atmega328 Board 5V 16M Arduino Compatible Nano NEW“). Das Gerät war ohne USB-Header ausgestattet, also musste ich zunächst noch einen USB-Serial-Wandler hierfür nutzen. Diesen hatte ich schon, er stammte ebenfalls aus Hong Kong.

Nachdem RX vom USB-Wandler zum TX vom mini Pro verbunden wurden, musste noch TX<->RX, 5V<->VCC und GND<->GND verbunden werden. Die „Blink“-Demo sollte laufen.

Danach die Blink-Demo in der Arduino-Programmierumgebung laden und testweise auf den Ardunio laden. Blinkt die Diode nun im (neu) vorgegebenen Takt, hat alles geklappt. Die Voraussetzung für die eigentliche Programmierung ist jetzt gegeben.

Nun kann man sich grob weiter an dem Artikel der make-Redaktion entlanghangeln: http://www.heise.de/make/artikel/Arduino-Uno-als-In-System-Programmer-2769246.html

Übrigens: Um das AVRDUDE unter Windows zu nutzen, lädt man sich das Open Source-Programm von hier: http://download.savannah.gnu.org/releases/avrdude/avrdude-6.1-mingw32.zip

Als nächsten Schritt muss der Pro.mini.328P zu einem ISP-Programmer gemacht werden. Hierzu lädt man in der Arduino-IDE unter „Beispiele“ den Sketch „Ardunio ISP“ aus und lädt diesen hoch.

Die Verkabelung muss danach wie folgt aussehen (den 10kOhm-Widerstand am Pin 1 kann man auch weglassen).

Der Quarz muss wirklich gut verbunden sein, sonst erhält man fortlaufend nur Fehlermeldungen der Art:

avrdude: stk500_recv(): programmer is not responding
avrdude: stk500_getsync() attempt 1 of 10: not in sync: resp=0x36
avrdude: stk500_recv(): programmer is not responding
avrdude: stk500_getsync() attempt 2 of 10: not in sync: resp=0x36
avrdude: stk500_recv(): programmer is not responding
avrdude: stk500_getsync() attempt 3 of 10: not in sync: resp=0x36
avrdude: stk500_recv(): programmer is not responding
avrdude: stk500_getsync() attempt 4 of 10: not in sync: resp=0x36
avrdude: stk500_recv(): programmer is not responding
avrdude: stk500_getsync() attempt 5 of 10: not in sync: resp=0x36
avrdude: stk500_recv(): programmer is not responding
avrdude: stk500_getsync() attempt 6 of 10: not in sync: resp=0x36
avrdude: stk500_recv(): programmer is not responding
avrdude: stk500_getsync() attempt 7 of 10: not in sync: resp=0x36
avrdude: stk500_recv(): programmer is not responding
avrdude: stk500_getsync() attempt 8 of 10: not in sync: resp=0x36
avrdude: stk500_recv(): programmer is not responding
avrdude: stk500_getsync() attempt 9 of 10: not in sync: resp=0x36
avrdude: stk500_recv(): programmer is not responding
avrdude: stk500_getsync() attempt 10 of 10: not in sync: resp=0x36

Und man sollte beim als ISP verwendeten Arduino nicht vergessen, zwischen GND und RESET einen 10 Micro-Farad-Kondensator einzusetzen, da dieser sonst immer wieder zurückgesetzt wird.

In der Arduino-IDE den ZIEL-Arduino auswählen (Hier: ATMEGA328P-PU = Arduino Uno) und als Programmer auf „ArduinoISP“ umstellen. Dann kann man mit „Bootloader brennen“ hoffentlich das blinken des Arduinos sehen und nach wenigen Sekunden ist der Bootloader fertig gebrannt. Da ich hiernach eine Menge unnötiger Zeit mit einer Fehlersuche vertan hatte, hier noch der wichtige Tipp, dass man den Quell-Arduino (den man als ISP einsetzt) am Ende, also nach fertigem Brennen noch einmal in den RESET schickt. Aus irgendeinem Grund hatte bei mir der ATMEGA328P sonst in einer Breadboard-Anordnung keinerlei Anstalten gemacht, das Bootloader-Blinken zu Beginn von sich zu geben, noch irgendein Programm via RX/TX anzunehmen.

Noch ungelöstes Problem: Nur nach einem manuellen (!) Reset akzeptiert der ATMEGA328P ein neues Programm aus der Arduino-IDE via RX/TX.

Nachdem ich den Minimal-Arduino auf dem Breadboard aufgebaut hatte, wollte ich diesen gerne über eine CR2032-Batterie mit Strom versorgen, und das möglichst lange.

Zum Glück gibt es schon einige Artikel im Internet, die sich mit dem Stromverbrauch des ATMEGA328P auseinandergesetzt haben. Allerdings erwarten hier viele einen Interrupt von Extern, was in meinem Anwendungsfall (halbwegs regelmässig einen Sensor auslesen und die Information abspeichern oder verschicken) nicht ganz praktisch war. Vor allem, weil ich den passenden Uhrenquarz nicht verfügbar habe…

Hier also (nachvollziehbar) die einzelnen Schritte, die ich bei meinem ATMEGA328P unternommen habe, um ihn den hohen Stromverbrauch streitig zu machen.

Step 1: Default blink program with 5s delay, 16 MHz ext. Osz., 3V CR2032-battery

void setup() {
  // initialize digital pin 13 as an output.
  pinMode(13, OUTPUT);
}
 
void loop() {
  digitalWrite(13, HIGH);   // turn the LED on (HIGH is the voltage level)
  delay(5000);              // wait for a second
  digitalWrite(13, LOW);    // turn the LED off by making the voltage LOW
  delay(5000);              // wait for a second
}

Result:

• LED On: 6.8mA
• LED Off: 5.8mA

Step 2: All other pins as input

void setup() {
  //To reduce power, setup all pins as inputs with no pullups
  for(int x = 1 ; x < 18 ; x++){
    pinMode(x, INPUT);
    digitalWrite(x, LOW);
  }
  // initialize digital pin 13 as an output.
  pinMode(13, OUTPUT);
}
 
void loop() {
  digitalWrite(13, HIGH);   // turn the LED on (HIGH is the voltage level)
  delay(5000);              // wait for a second
  digitalWrite(13, LOW);    // turn the LED off by making the voltage LOW
  delay(5000);              // wait for a second
}

Result:

• LED On: 6.45mA
• LED Off: 5.65mA

Step 3: Power Down during sleep

I’ll admit that this next step is a bit confusing, as a lot of things are introduced.

#include <avr/sleep.h> //Needed for sleep_mode
#include <avr/wdt.h>
 
volatile boolean f_wdt=1;
 
void setup() {
  //To reduce power, setup all pins as inputs with no pullups
  for(int x = 1 ; x < 18 ; x++){
    pinMode(x, INPUT);
    digitalWrite(x, LOW);
  }
  // initialize digital pin 13 as an output.
  pinMode(13, OUTPUT);
  setup_watchdog(8);
}
 
byte state=0;
 
void loop() {
    if (f_wdt==1) {  // wait for timed out watchdog / flag is set when a watchdog timeout occurs
    f_wdt=0;       // reset flag
    switch (state){
    case 0:
      digitalWrite(13, HIGH);   // turn the LED on (HIGH is the voltage level)
      state = 1;
      setup_watchdog(0);
      break;
    case 1:
      digitalWrite(13, LOW);    // turn the LED off by making the voltage LOW
      state = 0;
      setup_watchdog(8);
      break;
    }
    system_sleep();
  }
}
 
//****************************************************************  
// set system into the sleep state 
// system wakes up when watchdog is timed out
void system_sleep() {
  // power_adc_disable()
  set_sleep_mode(SLEEP_MODE_PWR_DOWN); // sleep mode is set here
  sleep_enable();
  sleep_mode();                        // System sleeps here
  sleep_disable();                     // System continues execution here when watchdog timed out 
  // power_adc_enable()
}
 
//****************************************************************
// 0=16ms, 1=32ms,2=64ms,3=128ms,4=250ms,5=500ms
// 6=1 sec,7=2 sec, 8=4 sec, 9= 8sec
void setup_watchdog(int ii) {
 
  byte bb;
  if (ii > 9 ) ii=9;
  bb=ii & 7;
  if (ii > 7) bb|= (1<<5);
  bb|= (1<<WDCE);
 
  MCUSR &= ~(1<<WDRF);
  // start timed sequence
  WDTCSR |= (1<<WDCE) | (1<<WDE);
  // set new watchdog timeout value
  WDTCSR = bb;
  WDTCSR |= _BV(WDIE);
}
//****************************************************************  
// Watchdog Interrupt Service / is executed when  watchdog timed out
ISR(WDT_vect) {
  f_wdt=1;  // set global flag
}

Result:

• LED On: 1.07mA
• LED Off: 0.02mA (206uA)

As long as we depend on the internal oscillator, we won’t get any further then 206uA, as the watchdog is still enabled and drawing current. Time to get one of those nice 32kHz-oscillators.

Here are some more links related to this topic, some of those helped me in my process (some in german):

• http://www.mikrocontroller.net/articles/Sleep_Mode
• http://interface.khm.de/index.php/lab/interfaces-advanced/sleep_watchdog_battery/

Abrufbare Sensoren Auflisten

Im verwendeten „Blank Activity“ ist bereits ein Textfeld enthalten, welches derzeit nur ein „Hello world!“ von sich gibt. Dieses wollen wir erstmal als Ausgabefeld für unsere ersten Tests behalten.

Zunächst möchte ich wissen, welche Sensoren in dem verwendeten Android-Gerät enthalten sind. Hierzu fügt man zunächst die folgende Prozedur in „MainActivity“ hinzu:

public void listSensors()  {
    // Find the text view
    TextView  t;
    t=(TextView)findViewById(R.id.textView);
    t.setText("");
 
    // Init SensorManager
    SensorManager sensorManager = (SensorManager)getSystemService(Context.SENSOR_SERVICE);
    // Get the list of sensors
    List listSensor = sensorManager.getSensorList(Sensor.TYPE_ALL);
    List listSensorType = new ArrayList();
    // Append all sensor names to the text view
    for(int i=0; i<listSensor.size(); i++){
        listSensorType.add(listSensor.get(i).getName());
        t.append(listSensor.get(i).getName());
        t.append("\n");
    }
}

Die IDE wird einige Begriff in ROT kennzeichnen, weil wir die notwendigen Imports nicht durchgeführt haben. Das lässt sich aber sehr bequem erledigen. Maus auf eines der roten Wörter positionieren und Alt+Eingabe drücken.

Zum Schluss beschwert sich die IDE noch darüber, dass „R.id.textView“ nicht gefunden werden kann. Dazu wechselt man in die activity_main.xml, wählt das „textView“-Element aus und vergibt in den Properties unter „id“ den Namen „textView“.

Nun müssen wir die Prozedur listSensors() nur noch im onCreate() aufrufen und schon werden beim Start der App alle Sensoren aufgelistet.

The currently new Intel NUCs are great hardware, but they need some special treatment for them to work with ubuntu und KODi. So here’s my experience and what steps I had to take to make this functional.

Hardware

• Intel NUC5i5RYH
• 64GB SanDisk Solid State Disk 2.5″ (6.4cm) SATA 6Gb/s MLC asynchron
• 2 * 2GB Crucial ValueRAM DDR3-1600 SO-DIMM CL11 Single

Update BIOS

This is fairly important, make sure that you update your BIOS first. I tried to install the newest version which was available from the Intel website, but I got an error, stating that the choosen BIOS cannot be installed. So I first upgraded to the previous versions step by step, until I reached the newest BIOS. Prior to 0247, the IR receiver won’t work, so updating the BIOS is fairly necessary to get a HTPC to work.

Initial Install with Ubuntu 15.04

You won’t have any luck with Ubuntu 14.04 or 14.10, as the graphic drivers in the installer don’t work with the HD6000 from the broadwell processor. So download at least Ubuntu 15.04 as an ISO-Image and store it to your disk.

Download the Universal USB Installer, choose the previously downloaded image and create a bootable USB-Stick. Afterwards, plug it into your Intel NUC and give it power. Be sure to have a keyboard, mouse, Network and HDMI/DisplayPort-Display attached to your device. Hit „F10“ and choose the USB-stick.

Choose „Install Ubuntu“ and use all default values that are presented to you except for the login – make sure that this is automatic and that you don’t have to enter a password. Of course, you can change your username. At the end, your NUC is rebooted, be sure to remove your USB-stick.

Things to do after the first boot of 15.04

There are some lists out there in the internet, but there are two things you should consider when using the NUC as a HTPC:

• Disable/Uninstall/Remove Amazon
• Disable the screen saver (and the „Lock“-Feature)
• Remove libsane (because it’s responsible for the crash if you get a „System program problem detected“ after login.

Remote Control

Ubuntu 15.04 is using rc_core, no longer lirc. Installing ir-keytable will make a new directory /etc/rc_keymaps:

sudo apt-get install ir-keytable
cd /etc/rc_keymaps

In my case, I use a Logitech Harmony One, which is configure to four devices:

• TV (for showing the content)
• Receiver (for Volume)
• Intel NUC (this is ONLY needed to turn it on and off)
• MCE (EVERY OTHER button should be mapped to this device)

It should look something like this. I guess you get the idea, even if some text is in german:

For your buttons from the remote to work, you must change some of the keys to the shortcuts from KODi. Create a new file /etc/rc_keymaps/rc-rc6-mce and put the following inside of it:

0x800f0400 KEY_0
0x800f0401 KEY_1
0x800f0402 KEY_2
0x800f0403 KEY_3
0x800f0404 KEY_4
0x800f0405 KEY_5
0x800f0406 KEY_6
0x800f0407 KEY_7
0x800f0408 KEY_8
0x800f0409 KEY_9
0x800f040a KEY_Q
0x800f040b KEY_ENTER
0x800f040c KEY_SLEEP
0x800f040d KEY_MEDIA
0x800f040e KEY_MUTE
0x800f040f KEY_I
0x800f0410 KEY_VOLUMEUP
0x800f0411 KEY_VOLUMEDOWN
0x800f0412 KEY_CHANNELUP
0x800f0413 KEY_CHANNELDOWN
0x800f0414 KEY_F
0x800f0415 KEY_R
0x800f0416 KEY_P
0x800f0417 KEY_RECORD
0x800f0418 KEY_SPACE
0x800f0419 KEY_X
0x800f041a KEY_DOT
0x800f041b KEY_COMMA
0x800f041c KEY_NUMERIC_POUND
0x800f041d KEY_NUMERIC_STAR
0x800f041e KEY_UP
0x800f041f KEY_DOWN
0x800f0420 KEY_LEFT
0x800f0421 KEY_RIGHT
0x800f0422 KEY_ENTER
0x800f0423 KEY_BACKSPACE
0x800f0424 KEY_ESC
0x800f0425 KEY_TUNER
0x800f0426 KEY_C
0x800f0427 KEY_ZOOM
0x800f0432 KEY_MODE
0x800f0433 KEY_PRESENTATION
0x800f0434 KEY_EJECTCD
0x800f043a KEY_BRIGHTNESSUP
0x800f0446 KEY_TV
0x800f0447 KEY_AUDIO
0x800f0448 KEY_PVR
0x800f0449 KEY_CAMERA
0x800f044a KEY_VIDEO
0x800f044c KEY_LANGUAGE
0x800f044d KEY_TITLE
0x800f044e KEY_PRINT
0x800f0450 KEY_RADIO
0x800f045a KEY_SUBTITLE
0x800f045b KEY_Q
0x800f045c KEY_GREEN
0x800f045d KEY_YELLOW
0x800f045e KEY_TAB
0x800f0465 KEY_POWER2
0x800f046e KEY_PLAYPAUSE
0x800f046f KEY_PLAYER
0x800f0480 KEY_BRIGHTNESSDOWN
0x800f0481 KEY_SPACE

Afterwards, load these keycodes and test them:

root@nuc:/etc/rc_keymaps# ir-keytable -w rc-rc6-mce
Wrote 63 keycode(s) to driver
root@nuc:/etc/rc_keymaps# ir-keytable -t
Testing events. Please, press CTRL-C to abort.

Install KODi

Now this step is pretty straightforward:

sudo apt-get install software-properties-common
sudo add-apt-repository ppa:team-xbmc/ppa
sudo apt-get update
sudo apt-get install kodi

Autostarting KODi after boot

We’ll make sure that KODi is started no matter which user starts the desktop. We can achieve this by adding a file to /etc/xdg/autostart:

/etc/xdg/autostart/kodi.desktop:

[Desktop Entry]
Type=Application
Name=Kodi
Exec=/opt/kodi.sh

As sometimes kodi drops to desktop (i.e. crashes), we want to make sure that it’s automatically restarted. That’s why we create another file, /opt/kodi.sh:

/opt/kodi.sh:

#!/bin/bash
killall pulseaudio
while true
do
AE_SINK=ALSA /usr/bin/kodi --standalone
done

Make sure that this file is executable:

root@nuc:/opt# chmod a+x /opt/kodi.sh

Get rid of PULSEAUDIO

Pulseaudio did not work well with my receiver, so I chose to disable it. As the default login user (NOT root), execute this command:

nuc@nuc:~$ echo "autospawn = no" > $HOME/.config/pulse/client.conf

Configure KODi

Now that we have a working KODi (simply reboot to make sure that this is true), we have to change some settings for it to work properly with the Intel NUC.

• Language / Location / Keyboard (if necessary)
• Audio Device (HDMI)
• Change further settings in ~/.kodi/userdata/advancedsettings.xml (if you did this before and know what you’re doing)

After a while, I got another device as a mediaplayer, called ODROID C1. This one has a much stronger cpu and has a gigabit ethernet interface. I can’t use OpenELEC on this one, as ARMv7 isn’t supported by now. So I stick with the default Ubuntu 14.04, which has its advantages, as it’s a complete Linux OS.

To monitor the temperature of this ODROID as I did before with my Raspberry Pis, I had to change the scripts a bit to get it working, my Ubuntu 14.04 uses upstart:

root@odwz:# ps -p1 | grep systemd && echo systemd || echo upstart
upstart

Of course, the prerequesite is socat:

sudo apt-get install socat

/opt/tempservice/t.sh

#!/bin/bash
cat /sys/devices/virtual/thermal/thermal_zone0/temp

/opt/tempservice/socat-service.sh

#!/bin/bash
socat -T 1 -d -d tcp-l:9888,reuseaddr,fork,crlf system:"/opt/tempservice/t.sh"

/etc/init/socat.conf

description "Socat service for temperature monitoring"
 
start on runlevel [2345]
 
stop on runlevel [016]
 
exec /opt/tempservice/socat-service.sh >>/var/log/socatservice.log 2>&1

sudo initctl reload-configuration
sudo initctl start socat

If you have a catchall on a domain – like anginf.de – you will probably get spam on some adresses which you would like to exclude from the general catchall. This can be achieved by doing some simple steps.

First you have to create a file /etc/postfix/recipient_block with all those email adresses you want rejected:

spam@anginf.de REJECT
morespam@anginf.de REJECT

To be useable for postfix, you have to postmap this file.

root@host:~# postmap /etc/postfix/recipient_block

In your /etc/postfx/main.cf, add a new line to your smtpd_recipient_restrictions:

hash:/etc/postfix/recipient_block

Now you only have to reload postfix and all mails to those mentioned in /etc/postfix/recipient_block will be rejected:

root@host:~# service postfix reload

1. sed -i 's/debug_enable\ =\ 0/debug_enable\ =\ 1/' /etc/odroid_remote.conf

2. odroid_remote /etc/odroid_remote.conf

3. tail -f /var/log/kern.log

4. Now press the buttons on your IR remote. You will get an error code which changes in the first letters. The last four letters are the factory code.
5. Replace the ZZZZ in the following with the four letters command and execute the command.

   sed -i 's/factory_code.*/factory_code\ =\ 0xZZZZ0001/' /etc/odroid_remote.conf

6. odroid_remote /etc/odroid_remote.conf

7. tail -f /var/log/kern.log

8. The custom error code will be replaced by „scancode is ...., invalid key is 0x....„
9. The LAST two letters of the „scancode“ are the needed keycodes. Enter it under key_begin and repeat_key_begin in your /etc/odroid_remote.conf

10. odroid_remote /etc/odroid_remote.conf

11. If you’re satisfied with the result, disable the debug_enable flag again:

12. sed -i 's/debug_enable\ =\ 1/debug_enable\ =\ 0/' /etc/odroid_remote.conf

13. And one final last time, run the odroid_remote-program:

14. odroid_remote /etc/odroid_remote.conf

Here’s my sample /etc/odroid_remote.conf, I configured it using a YAMAHA RV520 remote control:

factory_code = 0x1f000001
work_mode = 0
repeat_enable = 1
repeat_delay = 40
repeat_peroid = 39
release_delay = 121
debug_enable = 1
 
key_begin
        0x45 2   # Key "1"
        0x41 3   # Key "2"
        0x5e 4   # Key "3"
        0x49 5   # Key "4"
        0x4d 6   # Key "5"
        0x51 7   # Key "6"
        0x58 8   # Key "7"
        0x5c 9   # Key "8"
        0x5d 10  # Key "9"
#       0x00 11  # Key "0"
        0x1f 25  # Remote ">" (Play), Kodi "Play", Key "P"
        0x13 33  # Remote ">>" (FFWD), Kodi "Fast Forward", Key "F"
        0x15 19  # Remote "<<" (RWND), Kodi "Rewind", Key "R"
#       0x00 57  # Remote "||" (Pause), Kodi "Pause/play", Key "<SPACEBAR>"
        0x14 14  # Remote STOP, Kodi STOP, Key "X"
        0x1c 105  # Remote Left, Kodi Left, Key Left
        0x1d 106  # Remote Right, Kodi Right, Key Right
        0x1e 103  # Remote Up, Kodi Up, Key Up
        0x1a 108  # Remote Down, Kodi Down, Key Down
        0x4a 28  # Remote Select, Kodi Select, Key "<ENTER>"
        0x1b 14  # Remote Return, Kodi Back, Key "<BACKSPACE>"
        0x1b 1   # Remote Exit, Kodi Previous Menu/Home Screen, Key "<ESC>"
        0x40 71  # Remote Home, Kodi Top Menu, Key "<HOME>"
key_end
 
repeat_key_begin
        0x45 2   # Key "1"
        0x41 3   # Key "2"
        0x5e 4   # Key "3"
        0x49 5   # Key "4"
        0x4d 6   # Key "5"
        0x51 7   # Key "6"
        0x58 8   # Key "7"
        0x5c 9   # Key "8"
        0x5d 10  # Key "9"
#       0x00 11  # Key "0"
        0x1f 25  # Remote ">" (Play), Kodi "Play", Key "P"
        0x13 33  # Remote ">>" (FFWD), Kodi "Fast Forward", Key "F"
        0x15 19  # Remote "<<" (RWND), Kodi "Rewind", Key "R"
#       0x00 57  # Remote "||" (Pause), Kodi "Pause/play", Key "<SPACEBAR>"
        0x14 14  # Remote STOP, Kodi STOP, Key "X"
        0x1c 105  # Remote Left, Kodi Left, Key Left
        0x1d 106  # Remote Right, Kodi Right, Key Right
        0x1e 103  # Remote Up, Kodi Up, Key Up
        0x1a 108  # Remote Down, Kodi Down, Key Down
        0x4a 28  # Remote Select, Kodi Select, Key "<ENTER>"
        0x1b 14  # Remote Return, Kodi Back, Key "<BACKSPACE>"
        0x1b 1   # Remote Exit, Kodi Previous Menu/Home Screen, Key "<ESC>"
        0x40 71  # Remote Home, Kodi Top Menu, Key "<HOME>"
repeat_key_end

And here are the keycodes you’ll need for each key:

Helpful links for this article:

• Keyboard Controls for KODI

Auf einem Samba-Laufwerk, welches auf einem Linux-Rechner liegt und für die Windows-Computer freigegeben wurde, liegen sehr viele Dateien vor. Statt aus dem Windows Explorer heraus zu suchen (was aufgrund des Netzwerks und des Samba-Dienstes selbst ziemlich langsam ist), wird durch das Erstellen eines Ordners eine Suche definiert.

Hierzu wird ein neuer Ordner „_search“ verwendet. Alle unter diesem Ordner erstellten Unterordner lassen das Linux-System automatisch lokal eine Suche nach dem Namen des Ordners ausführen. Die gefundenen Dateien werden per Softlink im Verzeichnis bereitgestellt. Für Windows erscheint es dann so, als ob die Dateien direkt unterhalb des Unterordners liegen.

Am besten in /etc/rc.local das Skript aufrufen, oder direkt einen neuen Dienst erstellen.

inotify.sh

#!/bin/bash
# This example uses ".jpg" files
SEARCHPATH="/var/smb/"
mkdir -p ${SEARCHPATH}/_search
inotifywait -mq -e moved_to -e create --format %f ${SEARCHPATH}/_search | while read FILE
do
	# Nur auf Verzeichnisse reagieren, nicht auf Dateien
	if [ -d "${SEARCHPATH}/_search/${FILE}" ]
	then
		if [ `expr "${FILE}" : ".*Neuer Ordner"` -ne 0 ] # Auch Abwandlungen hiervon
		then
			continue
		fi
		rm -f "${SEARCHPATH}/_search/${FILE}/"* # Falls umbenannt wurde
		rm -f "${SEARCHPATH}/_search/${FILE}/.ready"
		SEARCH=`echo "${FILE}" | sed s/\ /*/g`
		find ${SEARCHPATH}/ -wholename '${SEARCHPATH}/_search' -prune -o -type f -iname "*${SEARCH}*.jpg" -printf "ln --backup=t --suffix=\".bak\" -s \"%p\" \"${SEARCHPATH}/_search/${FILE}/\"\n" | sh
		cd "${SEARCHPATH}/_search/${FILE}"
		# Duplicates have wrong extensions. TODO: Do this for any filename?
		for bkp in "${SEARCHPATH}/_search/${FILE}/"*.~*; do mv "${bkp}" "${bkp/%\~/.jpg}"; done >;/dev/null 2>&1
		# Mark for Windows user when we've finished. (Only necessary if local system is really slow or data is huge)
		touch "${SEARCHPATH}/_search/${FILE}/.ready"
	fi
done

I had to move a LOT of data from an old linux server to a new one. As both systems had four disks configured as a software RAID-5, I couldn’t connect all drives to one computer. I was simply out of ports to plug the drives into. 🙂

I wanted to use rsync, that way I could make sure that all the permissions etc. will stay intact. This was important for me, as a part of the data was from the BackupPC-directory, and BackupPC uses a special user and uses a LOT of hardlinks.

Simply copying everything with rsync wasn’t a success, the underlying SSH worked well, but the speed was by far too slow (~20MB/s). The „old“ machine wasn’t simply powerful enough to handle RAID-5, SSH and rsync with maximum performance at the same time.

Using a differenct cyper (-c arcfour) helped a bit, but the speed was still too slow (~40MB/s). As both computers were in the same secured LAN, I figured I didn’t need the additional protection and could use the rsync itself.

In the config file and the commands below you must change „hostname“ to the name of the „server“ and the „sharename“ accordingly. If you’re using a different IP range, change this as well.

On the server (here: the old machine) I created a file „rsyncd.conf“ with the following contents:

use chroot = true
hosts allow = 192.168.0.0/24
log file = /var/log/rsyncd.log
log format = %h %o %f %l %b
 
[sharename]
comment = sharename
path = /
read only = no
list = yes
uid = root
gid = root

And run this command on the server:

rsync --config=rsyncd.conf --daemon

On the client (here: the new, receiving machine) run this command:

rsync --delete -avPH hostname::sharename .

I had a big website with a lot of images embedded inside, which were dynamically added by users. The pictures were all of the same size, so I thought it would be easy to create a video out of it. Here are the steps to reproduce it:

Step 0 – Prerequisites

• You’ll need python and http://code.google.com/p/youtube-upload/downloads/list to upload the video directly from the server. If you want to upload yourself, ignore this line and step 5.
• md5sum
• avconv – on ubuntu, this is available through sudo apt-get install libav-tools

Step 1 – Extract the images

Save the website you’re about to process to a local file, you only need the html part. (This only works with embedded pictures, for static pictures, try downloading the website with wget -m). Call the following script with the filename of the saved HTML-file.

#!/bin/bash
file=${1}
mkdir -p result
echo "Processing ${file}."
cat ${file} | sed -e 's/>/>\n/g; s/<img/\n<img/g; s/^$//g' | grep 'data:image/jpeg;base64' | sed -e 's/<img src="data:image\/jpeg;base64,//; s/" .*//' >out
count=0
while read line
do
let "count+=1"
echo ${line} | base64 -d >$(printf "result/%05d.jpg\n" ${count})
done <out

Step 2 – Images should be unique

We must delete all the files that are twice in our results.

#!/bin/bash
declare -a list=( ./* );
declare -a sums;
cnt=${#list[@]}
 
echo "creating md5sum list"
for ((x = 0; x < $cnt -1; x++))
do
    sums[$x]=`md5sum ${list[$x]} | cut -d ' ' -f 1`
done
 
echo "doing compare"
for ((x = 0; x < $cnt -1; x++))
do
  for ((y = x+1; y < $cnt; y++))
  do
    if [ "${sums[$x]}" == "${sums[$y]}" ];then
      if [ ${list[$x]} != ${list[$y]} ];then
        #remove '#' in next line to enable
        echo "Delete file ${list[$y]}" # && rm -f ${list[$y]}
      fi
    fi
  done
done

Step 3 – Renumber

The avconv-Tool can arrange all the images, but they must start with 00000.jpg and strongly increasing numbers. Therefore, we renumber all the jpg-Files accordingly.

#!/bin/bash
find . -type f -iname '*.jpg' | sort -n >filelist
seq=0
while read line
do
        mv ${line} $(printf "%05d.jpg\n" ${seq})
        let "seq+=1"
done <filelist

Step 4 – Create the video

We want to output a video with 25 fps and want to simulate that the images are a video with 3 fps. If you want slower changes in the images, increase the „-r 3“ to your wishes. If you want it quicker, you can decrease it down to „-r 1“.

avconv -f image2 -r 3 -i %05d.jpg -r 25 out.mpg

Step 5 – Upload to youtube

Make sure that you have the correct username and a channel created. This part is quite tricky, if you keep getting the 403-Error, make sure that you’ve logged in into youtube and you don’t get any Captchas.

python youtube_upload.py --email=USERLOGIN --password=PASSWORD --title="The title" --description="The description" --category=Comedy --keywords="keywords" out.mpg

The Software GNU mailman is a great mailinglist-tool, which can be installed on a small server, ensuring independence from big companies like google & co.

In my setup, I’m using apache2 together with suexec and postfix. Unfortunately, mailman doesn’t work easy with suexec. I’m also using froxlor, which makes it easy to setup new domains, emails etc. for customers or family. Mailman isn’t integrated in froxlor (there is an outdated custom module, though), so I had to setup something on my own.

Please remember to update the following placeholders in the scripts below:

• your.ip.add.ress
• list.yourdomain.tld
• admin@account

Here’s what I have done:

# Download and extract mailman
mkdir -p ~/work/mailman
cd ~/work/mailman
wget http://ftp.gnu.org/gnu/mailman/mailman-2.1.17.tgz
tar xvfz mailman-2.1.17.tgz
# Add user and group
sudo addgroup mailman
sudo adduser mailman --ingroup mailman
# Secure stuff: no directory, no Login
sudo sed -i 's/mailman:x:\([0-9]*\):\([0-9]*\):Mailman,,,:\/home\/mailman:\/bin\/bash/mailman:x:\1:\2:Mailman,,,:\/nonexistent:\/bin\/false/' /etc/passwd
# We need to identify which is our docroot, we're creating the install dir there.
# This is most certainly "/var/www"
docrootpart=$(sudo /usr/lib/apache2/suexec -V 2>&1 | grep AP_DOC_ROOT | sed 's/.*"\(.*\)"/\1/')
myprefix=${docrootpart}/mailman
sudo mkdir -p $myprefix
cd $myprefix
sudo chgrp mailman .
sudo chmod a+rx,g+ws .
cd ~/work/mailman/mailman-2.1.17
# configure, make and install
./configure --prefix=$myprefix --with-username=mailman --with-groupname=mailman --with-cgi-gid=mailman --with-mail-gid=mailman
make
sudo make install
cd $myprefix
sudo bash -c 'bin/check_perms -f'
# Using information from http://wiki.list.org/pages/viewpage.action?pageId=4030646
sudo chown -R mailman:mailman $myprefix/cgi-bin*
sudo chmod g-w $myprefix/cgi-bin
sudo chmod g-s $myprefix/cgi-bin/*

After this, you now must create the mailman.conf-file at /etc/apache2/conf.d/mailman.conf:

Remember to check your docroot, you might have to change the /var/www below at all occurences. If you’re using SSL, change the 80 to 443 and add all the SSLEngine On etc. lines from your local apache configuration, which pretty much includes everything starting with SSL.

<VirtualHost your.ip.add.ress:80>
        ServerName      list.yourdomain.tld
        ServerAlias     list.yourdomain.tld
        SuexecUserGroup mailman mailman
        DocumentRoot    "/var/www/mailman"
        Alias /pipermail/ /var/www/mailman/archives/public/
        Alias /icons/ /var/www/mailman/icons/
        ScriptAlias /admin /var/www/mailman/cgi-bin/admin
        ScriptAlias /admindb /var/www/mailman/cgi-bin/admindb
        ScriptAlias /confirm /var/www/mailman/cgi-bin/confirm
        ScriptAlias /create /var/www/mailman/cgi-bin/create
        ScriptAlias /edithtml /var/www/mailman/cgi-bin/edithtml
        ScriptAlias /listinfo /var/www/mailman/cgi-bin/listinfo
        ScriptAlias /options /var/www/mailman/cgi-bin/options
        ScriptAlias /private /var/www/mailman/cgi-bin/private
        ScriptAlias /rmlist /var/www/mailman/cgi-bin/rmlist
        ScriptAlias /roster /var/www/mailman/cgi-bin/roster
        ScriptAlias /subscribe /var/www/mailman/cgi-bin/subscribe
        ScriptAlias /mailman/ /var/www/mailman/cgi-bin/
        ScriptAlias / /var/www/mailman/cgi-bin/listinfo
<Directory "/var/www/mailman/archives/public/">
        AddDefaultCharset off
</Directory>
</VirtualHost>

Now restart the apache webserver:

sudo service apache2 restart

Mailman is now available under http://list.yourdomain.tld/mailman/create (shouldn’t throw any errors anymore). Do NOT create a mailing list just yet. If you’re still getting errors like 500 (Internal Server Error), check your log files, especially your suexec.log, probably located at /var/log/apache2/suexec.log

# Change postfix configuration
# http://www.list.org/mailman-install/node12.html
sudo bash -c 'echo "recipient_delimiter = +" >>/etc/postfix/main.cf'
sudo bash -c 'echo "unknown_local_recipient_reject_code = 550" >>/etc/postfix/main.cf'
# Change mailman configuration
# http://www.list.org/mailman-install/postfix-integration.html
cd $myprefix/Mailman
sudo bash -c 'echo "MTA = '\''Postfix'\''" >>mm_cfg.py'
cd $myprefix
# Generate aliases
sudo bin/genaliases
sudo chown mailman:mailman data/aliases*
sudo chmod g+w data/aliases*
sudo sed -i 's/alias_maps\ =\ \$alias_database/alias_maps\ =\ \$alias_database,hash:\/var\/www\/mailman\/data\/aliases/' /etc/postfix/main.cf

Mailman is now configured, let’s create our first mailling list, this must be the „mailman“-list, this is only an internal mailing list.

## Configure mailman internal
# maillinglist "mailman"
cd $myprefix
sudo bin/newlist mailman
# Configure siteliste (use default)
sudo bin/config_list -i data/sitelist.cfg mailman
# Enable cron - make sure to change the MAILTO-Information to your email!
# http://www.list.org/mailman-install/node41.html
cd $myprefix/cron
sudo bash -c 'echo "MAILTO=admin@account" >crontab.ok'
sudo bash -c 'cat crontab.in >>crontab.ok'
sudo crontab -u mailman crontab.ok
# Add virtual host (here: list.yourdomain.tld)
cd $myprefix/Mailman
sudo bash -c 'echo "add_virtualhost('\''list.yourdomain.tld'\'')" >>mm_cfg.py'
# Now as we're super-secure, we're adding the https-Information
# Make sure that this is supported by your webserver and fully configured, otherwise skip the next two lines!
sudo bash -c 'echo "DEFAULT_URL_PATTERN = '\''https://%s/mailman/'\''" >>mm_cfg.py'
sudo bash -c 'echo "PUBLIC_ARCHIVE_URL = '\''https://%(hostname)s/pipermail/%(listname)s'\''" >>mm_cfg.py'
# Default-Host
sudo bash -c 'echo "DEFAULT_EMAIL_HOST = '\''list.yourdomain.tld'\''" >>mm_cfg.py'
sudo bash -c 'echo "DEFAULT_URL_HOST = '\''list.yourdomain.tld'\''" >>mm_cfg.py'
# for ubuntu/debian, we're adding the startup-script and enable mailman to run after reboot
# http://www.list.org/mailman-install/node42.html
sudo cp $myprefix/scripts/mailman /etc/init.d/mailman
sudo update-rc.d mailman defaults
sudo /etc/init.d/mailman start
# Set asswords
sudo $myprefix/bin/mmsitepass