Thanks a lot, man!

Mautic Installation Bash Script for Debian

2025-11-14T11:57:00.000-08:00

Here's a bash script to install Mautic on Debian.

Name the Mautic deployment package to "mautic.zip" and place it in the same directory as the installation script.

Health checks are built-in, so re-running the script won't clobber existing installed items. The database user configuration is always overwritten, though.

Installation items:

MariaDB (MySQL drop-in replacement)
PHP 8.2
Nginx
Mautic

Installation script

#!/usr/bin/env bash

set -euo pipefail

trap 'echo "Error on line $LINENO"; exit 1' ERR

# Variables

DB_NAME="mautic"

DB_USER="mauticuser"

DB_PASS="YourSecurePassword" #CHANGE THIS

MAUTIC_DIR="/var/www/mautic"

PHP_VERSION="8.2"

DOMAIN="YourAwesomeDomain.xyz" #CHANGE THIS

SERVER_IP="$(hostname -I | awk '{print $1}')" # auto-detect primary IP

SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"

MAUTIC_ZIP="${SCRIPT_DIR}/mautic.zip"

# Ensure root

if [[ $EUID -ne 0 ]]; then

echo "Run as root (sudo)."

exit 1

install_packages() {

echo "Checking required packages..."

local packages=(

nginx mariadb-server mariadb-client

php${PHP_VERSION} php${PHP_VERSION}-cli php${PHP_VERSION}-mysql

php${PHP_VERSION}-curl php${PHP_VERSION}-xml php${PHP_VERSION}-mbstring

php${PHP_VERSION}-zip php${PHP_VERSION}-intl php${PHP_VERSION}-fpm

unzip curl git composer

)

local to_install=()

for pkg in "${packages[@]}"; do

if ! dpkg -s "$pkg" >/dev/null 2>&1; then

to_install+=("$pkg")

else

echo "✔ $pkg already installed, skipping."

done

if [[ ${#to_install[@]} -gt 0 ]]; then

echo "Installing missing packages: ${to_install[*]}"

apt-get update -y

DEBIAN_FRONTEND=noninteractive apt-get install -y "${to_install[@]}"

else

echo "All required packages already installed."

}

setup_database() {

echo "Configuring database..."

mysql -u root -e "CREATE DATABASE IF NOT EXISTS ${DB_NAME} CHARACTER SET utf8mb4 COLLATE utf8mb4_unicode_ci;"

mysql -u root -e "DROP USER IF EXISTS '${DB_USER}'@'localhost';"

mysql -u root -e "CREATE USER '${DB_USER}'@'localhost' IDENTIFIED BY '${DB_PASS}';"

mysql -u root -e "GRANT ALL PRIVILEGES ON ${DB_NAME}.* TO '${DB_USER}'@'localhost';"

mysql -u root -e "FLUSH PRIVILEGES;"

}

install_mautic() {

echo "Installing Mautic from local zip..."

if [[ ! -f "${MAUTIC_ZIP}" ]]; then

echo "❌ Mautic archive not found at ${MAUTIC_ZIP}"

exit 1

if [[ ! -d "${MAUTIC_DIR}" || -z "$(ls -A "${MAUTIC_DIR}")" ]]; then

mkdir -p "${MAUTIC_DIR}"

unzip -o "${MAUTIC_ZIP}" -d "${MAUTIC_DIR}"

chown -R www-data:www-data "${MAUTIC_DIR}"

chmod -R 755 "${MAUTIC_DIR}"

else

echo "Mautic directory already populated, skipping extraction."

}

configure_nginx() {

echo "Configuring Nginx..."

CONF_FILE="/etc/nginx/sites-available/mautic"

if [[ ! -f "${CONF_FILE}" ]]; then

cat > "${CONF_FILE}" <<EOF

server {

listen 80;

server_name ${DOMAIN} ${SERVER_IP} localhost _;

root ${MAUTIC_DIR};

index index.php index.html;

location / {

try_files \$uri /index.php\$is_args\$args;

}

location ~ \.php\$ {

include snippets/fastcgi-php.conf;

fastcgi_pass unix:/run/php/php${PHP_VERSION}-fpm.sock;

fastcgi_param SCRIPT_FILENAME \$document_root\$fastcgi_script_name;

include fastcgi_params;

}

location ~ /\.ht {

deny all;

}

EOF

ln -sf "${CONF_FILE}" /etc/nginx/sites-enabled/mautic

else

echo "Nginx config already exists, skipping."

# Disable default site if present

if [[ -f /etc/nginx/sites-enabled/default ]]; then

echo "Disabling default Nginx site..."

rm -f /etc/nginx/sites-enabled/default

nginx -t

systemctl reload nginx

}

setup_systemd_unit() {

echo "Configuring systemd unit for Mautic jobs..."

UNIT_DIR="/etc/systemd/system"

SERVICE_FILE="${UNIT_DIR}/mautic-jobs.service"

TIMER_FILE="${UNIT_DIR}/mautic-jobs.timer"

if [[ ! -f "${SERVICE_FILE}" ]]; then

cat > "${SERVICE_FILE}" <<EOF

[Unit]

Description=Mautic consolidated background jobs

After=network.target mariadb.service php${PHP_VERSION}-fpm.service

[Service]

Type=oneshot

User=www-data

Group=www-data

WorkingDirectory=${MAUTIC_DIR}

ExecStart=/usr/bin/php ${MAUTIC_DIR}/bin/console mautic:segments:update --env=prod

ExecStart=/usr/bin/php ${MAUTIC_DIR}/bin/console mautic:campaigns:update --env=prod

ExecStart=/usr/bin/php ${MAUTIC_DIR}/bin/console mautic:emails:send --env=prod

EOF

if [[ ! -f "${TIMER_FILE}" ]]; then

cat > "${TIMER_FILE}" <<EOF

[Unit]

Description=Run Mautic jobs periodically

[Timer]

OnBootSec=5m

OnUnitActiveSec=15m

Persistent=true

[Install]

WantedBy=timers.target

EOF

systemctl daemon-reexec

systemctl enable --now mautic-jobs.timer

}

ensure_services_running() {

echo "Ensuring Nginx and PHP-FPM are running..."

systemctl enable --now nginx

systemctl enable --now php${PHP_VERSION}-fpm

}

health_check() {

echo "Performing health check..."

STATUS=$(curl -s -o /dev/null -w "%{http_code}" "http://localhost")

if [[ "$STATUS" =~ ^2|3 ]]; then

echo "✅ Health check passed: Mautic web interface responded with HTTP $STATUS"

echo " Access it at http://localhost or http://${SERVER_IP}"

else

echo "❌ Health check failed: HTTP $STATUS returned. Check Nginx/PHP-FPM logs."

exit 1

}

main() {

install_packages

setup_database

install_mautic

configure_nginx

setup_systemd_unit

ensure_services_running

health_check

echo "✅ Mautic 6.0.6 installed with Nginx + MariaDB, idempotent and safe for reruns."

}

main "$@"

2025-02-05T20:48:00.000-08:00

How to connect to Turnkey Linux application hosted in a VirtualBox virtual machine

If the IP address of the VirtualBox virtual machine is something like 10.0.2.15, the virtual machine networking mode is probably set to NAT (network address translation).

That means the host machine is acting as a router for internet traffic to that virtual machine and you won't be able to connect it.

What you need to do is set the virtual machine networking mode to 'bridged', which means the host machine and the virtual machine will be able to see each other on a virtual network connecting them together and to the internet.

The snag here is that the network drivers for VirtualBox aren't installed by default because TurnKey Linux is owned by VMWare and the virtual machines are built for VMWare platforms, so you need to install the VirtualBox guest additions.

Here's how to do that for the Debian Linux distribution:

Update Package Indexes:
sudo apt update

Upgrade Installed Packages:

sudo apt upgrade

Install Required Packages:

sudo apt install build-essential dkms linux-headers-$(uname -r)

Insert Guest Additions CD Image:

In the VirtualBox menu, go to Devices > Insert Guest Additions CD Image.

Mount the CD Image:

sudo mkdir -p /mnt/cdrom

sudo mount /dev/cdrom /mnt/cdrom

Run the Guest Additions Installer:

cd /mnt/cdrom

sudo ./VBoxLinuxAdditions.run

Shutdown the System:

sudo shutdown now

Configure the Virtual Machine to use "Bridged" networking:

The physical network adapter of the computer should be shown as the network adapter name.

Start the virtual machine:

The IP address should now be something similar to 192.168.1.78

Navigate to the new IP address via the browser of your host machine:

Lemon squeezy.

How to get AutoIt ConsoleWrite to output when running as an application .exe

2023-01-19T14:18:00.004-08:00

An AutoIt script the using ConsoleWrite() function will output as expected to the Scite console when run as a script, but if you use the compile tool in the "Extras" folder, the .exe doesn't output information when run from Windows command prompt.

The solution is to:

Execute the "Compile" option from the Tools menu (or CTRL + F7)
Select the "Create CUI instead of GUI EXE" option in the Output arch section.

Using that workflow will compile the script into an .exe that will output to a command line window without having to modify the script to use complicated Windows system calls to get your progress/information/error/whatever messages displayed.

The Tools Menu Option

The Compiler Option

Scylla Table Join

2022-12-01T22:19:00.004-08:00

Scylla Table Join

Scylla can't join tables.

It's a NOSQL database which is designed for key-value lookups in a single table. It's not flexible, but Scylla is very, very fast if you want to find a particular piece of information.

Do you have an order number and you want to find items in that order? Scylla excels at that.

That's different from a relational database which usually has multiple tables

The main difference between NoSQL and relational databases is the way they store and organize data. Relational databases store data in structured tables with rows and columns and use a structured query language (SQL) for managing and accessing the data. In contrast, NoSQL databases are designed to store and manage data in a more flexible and scalable way. NoSQL databases typically use a non-relational data model, such as key-value pairs, document-based, or graph-based data, and often use a more flexible and scalable query language.

There are several other differences between NoSQL and relational databases. For example, NoSQL databases are often more scalable and can handle larger volumes of data but may not offer the same level of consistency and transactional support as relational databases. Relational databases are typically more established and mature but may not be as well-suited for modern web-scale applications that require high performance and flexibility. Ultimately, the choice of which type of database to use will depend on the specific needs and requirements of the application.

Planetside 2 Stuttering - Steam Client Bootstrapper Service Disk Usage

2018-12-15T12:39:00.006-08:00

I bought a new beast of a video card and when playing Planetside 2 (it's free to play!) it became unplayable at certain points in game, particularly right before engaging enemy due to framerate drop. It seemed like right before I would encounter enemy, the game was loading files and it would overburden the system.

I also happened to have a task manager window open, so when I quit the game in frustration, I saw the Steam Client Bootstrapper Service was transferring 30 to 40 MBs of data almost constantly.

I dug around and found that there seems to be something confusing the Steam background downloader and making it keep accessing the disk and the solution is to clear the Steam cache.

To do this in Windows, press Win-R (or type "run") and run steam://flushconfig

You may need to completely exit Steam first. This will not remove any games or save games, just clean the download cache. You'll need to log into Steam again.

This worked for me and Planetside 2 is running smoothly again.

How to root a Samsung Note 5 SM-N920X and flash a Gear VR compatible ROM

2018-01-21T18:40:00.001-08:00

Attention! The "one click root" applications out there will hijack your phone and load it with adware, malware and worse. KingoRoot, KingRoot, and the rest should be avoided.

I bought a Samsung Note 5 SM-920X Live Demo Unit (it's the display model you see at stores) and found this ROM which works with Gear VR.

Here's my post on turning off "Retail Mode", if you need it.

The installation steps will flash a rooted kernel and the TWRP recovery bootloader.

The ROM is Dr. Ketan's Android 7 which has a lot of useful features.

You'll have lots of space to store VR content on your Note 5.

Enjoy!

https://forum.xda-developers.com/showpost.php?p=74653947&postcount=13622

How to Disable Retail Mode on a Samsung Galaxy Note 5 SM-N920X Live Demo Unit

2018-01-18T19:51:00.002-08:00

Device Limitations

First of all, the Samsung Note 5 SM-N920X doesn't have a cellular modem so it is unable to place telephone calls or send text messages. It's intended as a store demonstration unit and is essentially a tablet with wifi.

Unlock

The most common method to disable "Retail Mode" on a Samsung Note 5 is as follows:

Go to the app drawer and open an app called Samsung Retail.
Tap the triple dots and then tap "Configuration Settings" or "Store Settings" and enter the password 5444 to unlock it.
Then uncheck “Disable the factory reset” and enter the following password M729Q16K8546.
Turn your Samsung Galaxy S5 off and boot into Recovery mode.
Go to Wipe data & factory reset and press OK.

Apparently, Walmart and Sam's Club have a unique password to disable Retail Mode.

The password is "walton" in lowercase.

I've also read "rollback" has been used, in case you need another option.

Best Buy uses "blue" as the password.

How to root T-Mobile Alcatel One Touch Evolve 5020T

2014-11-13T07:05:00.001-08:00

The Alcatel One Touch Evolve 5020T from T-Mobile can be rooted using Cydia Impactor from the guys of iOS jailbreak fame. Version for Mac, Windows, and Linux are available.
You can get Cydia Impactor here.

Backup copy for Windows: Cydia Impactor 0.9.14

The tool uses a one-click method. It could not be easier!

Here's how to root an Alcatel Evolve 5020T:

Download Cydia Impactor to your desktop – Windows | Mac | Linux 32-bit | Linux 64-bit
Extract the package to your desktop.
Run the Cydia Impactor tool.
On your Android device, enable USB Debugging from Settings > Developer Options
Connect the Alcatel Evolve 5020T to a USB port on your computer.
On Windows, you would need to have the USB drivers for your device installed. If not, you can also use the Driver Scan function in Cydia Impactor to automatically construct and install a driver for your device.
After making sure that # drop SuperSU su to /system/xbin/su is selected in the drop down menu, click the 'Start' button.
Wait for the tool to do its thing.
Once done, download and install Superuser from the Google Play Store.
Enjoy Root access on your Android device.

Another helpful utility to install is Titanium Backup which allows you archive all the applications and data on the Alcatel Evolve 5020T. After the backup is complete, you can start to delete the T-mobile bloatware.

Handily, Cydia Impactor also includes an ADB (Android Debug) shell that you can use to list all the packages on the phone. You can also backup the packages as well using the ADB shell:

Here's how to use the shell:

Backup apks

You should back up any apks that you plan to remove. That way, you can easily re-install them if you change your mind. You have two main options:

Nandroid backup from the recovery console, or
Manual backup with

$ adb pull /system/app/ ~/Desktop/app/

How to find applications

To find what apps are installed on the device, you will need to run the following, either in adb shell, or the device's term:

$ ls /system/app

$ ls /data/app

You can also list installed packages using the package manager, either in Terminal Emulator or via adb shell:

$ pm list packages

To see a list of installed packages along with their associated files:

$ pm list packages -f

At a lower level, you can pull the packages.xml and the packages.list files using adb and examine them directly to determine what is installed and what the related .apk file is called and where it is stored. packages.xml will show you the package name and the .apk file installation path and name. In packages.list you can look up the package name and locate the data path. Both of these files will tell you a great deal more, these are the things that pertain to the present article.

$ adb pull /data/system/packages.xml
$ adb pull /data/system/packages.list

Generally, any apps you install from the market or otherwise manually install will be in /data/app, while any apps that come pre-installed will be in /system/app.

How to remove applications

Remember to watch the case of all files, directories, apk names, and package names!

Case matters. Example.apk and example.apk are two different things.

To remove applications you can use aps on the market Titanium Backup, ROM Toolbox Pro, ES File Explorer or do it through ADB First of all you restart adbd as root

$ adb root

ADB

/system apps	/data apps
`$ adb remount $ adb shell rm -f /system/app/.apk`	`$ adb uninstall`

Term (on device)

/system apps	/data apps
`$ su $ mount -o rw,remount /system $ rm -f /system/app/.apk`	`$ su $ mount -o rw,remount /data $ rm -f /data/app/.apk $ pm uninstall`

How to reinstall applications

Change your mind? No problem! You have two options, depending on how you backed up your files:

Restore your Nandroid backup from the recovery console, or
Manually restore your app with

$ adb install ~/Desktop/app/.apk

What are Python decorators?

2014-11-09T14:00:00.002-08:00

First we need to get the concept of higher-order programming. This is when we treat functions as just more things to work with; we can write functions which take other functions as arguments, return new functions, etc.

So, for example, suppose we have this simple function:

def add(x, y):
    return x + y

But let's say there's something in our codebase that often requires adding 2 -- and not any other number -- to something. So we could do something like:

def add2(x):
    return add(x, 2)

This is a technique that's called currying -- we're making a new function which has fewer arguments, because the arguments that have disappeared now have fixed values on every call (in the case of add2, one of the arguments ofadd is now fixed as 2, instead of potentially varying on every call).

We could even write a function which generates specific "adder" functions:

def make_adder(number):
    def _adder(x):
        return add(x, number)
    return _adder

This works because the value of number gets "closed over" (the usual term is that this relies on a closure) -- Python remembers that value, and so whenever we call the function returned by make_adder we'll get the value of number from that specific time we called make_adder. So, for example, we could do:

add2 = make_adder(2)
add3 = make_adder(3)

and so on.

Now, decorators. Decorators are just another example of higher-order programming. A decorator is a function which takes a function as an argument (and may take other arguments, too), and returns a new function.

The simplest case for decorators is when you have an existing function, and you want to make a new function that does some extra stuff before or after. Like, say, if you need to debug by printing out the arguments a function was called with, you could write this:

def argument_printer(func):
    def _wrapper(*args, **kwargs):
        print "Called with positional arguments: %s" % list(args)
        print "And with keyword arguments: %s" % kwargs
        return func(*args, **kwargs)
    return _wrapper

Now, suppose we want to always know what arguments our original add function gets called with. We could do:

add = argument_printer(add)

Or, using the decorator syntax:

@argument_printer
def add(x, y):
    return x + y

And now any call to add will print out the arguments it was called with, before actually calling add and returning the correct result.

And... that's it. That's a decorator. But that's just the tip of the iceberg as far as what they can do. Decorators can do pre- and/or post-processing, as in this example. Or they can implement more complex logic, or choose what to do based on what arguments the function gets called with, or... well, all sorts of things, really.

Bluetooth Mouse Works or Sleeps Intermittently

2014-05-01T12:33:00.000-07:00

If you find that your Bluetooth mouse stops working every couple of minutes, it could be that the Bluetooth device manager in your system is configured to save power by turning off the radio.

This seems to be the default behavior for Windows and I've seen many posts about it happening with Linux installations as well so I hope this post helps all the people who are frustrated and probably angry with this annoyance of their mouse falling asleep every couple of minutes. This problem with my Bluetooth mouse always plagues me when I get a new laptop until I get frustrated enough to fix the issue. This post is for me as well as you.

On Windows:

Open Device Manager by clicking the Start button
Click Control Panel
Click System and Security
And then, under System, click Device Manager
Expand the Bluetooth list item
Right click the Bluetooth Adapter device and click Properties
Click the Power Management tab
Uncheck the 'Allow the computer to turn off this device to save power' option
Click OK

Now enjoy your constantly connected Bluetooth mouse!

How to move the command line cursor to the beginning of the line in C#

2014-04-16T16:18:00.004-07:00

I was working on a little C# application to process online documents and didn't want to clutter the window with a running list of items processed. I just wanted a single line to update with the item being worked on.

Using Console.WriteLine outputs the data and starts new on the next line. Nope.
Using Console.Write outputs the data in one big long stream. Nope.

I needed a way to return the cursor to the beginning of the current line.

Using Console.Write('\x000D') does that.

Here's why:

Windows uses two character to create a new line when displaying information in the shell window. It sort of works like a typewriter (or teletype for that matter) by moving the cursor to the beginning of the line and then advancing the cursor one row on the screen.

Those two ASCII characters are '\r' and '\n', being 'return' and 'new line' respectively.

Here are the ASCII, decimal and hexadecimal values for those characters:

ASCII Decimal Hex

\r 13 0D
\n 10 0A

Including the '\x' in the Console.Write parameter tells the application to that the data is encoded in hexadecimal. You could use Console.Write(char(13)) to type cast the ASCII code if you wanted to.

The Console.Write('\x000D') example above includes two extra zeros to account for Unicode's two-byte character set... just FYI.

A possible use for this could be to create progress bar/completion percentage in your application.

How to compute an MD5 and SHA1 cryptographic hash values for a file in Windows

2014-03-01T15:48:00.001-08:00

Microsoft has a handy command line tool to compute the MD5 or SHA1 value for file and makes it available for free.

So if you need a quick way to check your torrent hashes, this is probably the tool you want. You could throw together a quick GUI using AutoIt Script if you want to be fancy.

The name of the utility tool is File Checksum Integrity Verifier (FCIV).

To compute the MD5 and the SHA-1 hash values for a file at the same time (how handy!), type the following command at a command line:
FCIV -md5 -sha1 path\filename.ext

For example, to compute the MD5 and SHA-1 hash values for the Shdocvw.dll file in your %Systemroot%\System32 folder, type the following command:
FCIV -md5 -sha1 c:\windows\system32\shdocvw.dll

I can think of examples of where having a command line hash generator - in both MD5 and SHA1 - would be very handy.

How To Choose A Good Password

2013-08-07T00:14:00.002-07:00

Contemporary password design confirms long simple passwords are more secure than complicated short passwords.

Long story short, the explanation is here: http://xkcd.com/936/

The upshot of this is that you have the chance to make your unique poem.

You could start with a haiku:

ExplosionBoomBoom
WelcomeToTheShowWelcome
PleaseComeInSideBoom

I modified 'Welcome Ladies and Gentleman" by Saliva (I'm on vacation and feel like ROCKING) for that example. But you grasp the concept, right?

You could idealize an idea and then write a sentence about it:

"Monster Movie Season Is So Fun!"

That is a very secure password.

Per the rules of my organization, I have to change my password every 10 weeks. The choice of formats I have are either one that's less that sixteen characters with so many capitals and non-alphanumeric symbols as to make the password unrememberable and gymnastic to type or a "passphrase" that is much more manageable, like a sentence.

Would you mind being reminded to change your password every so often if you could skip the 'more capitals and punctuation' message? I don't now that I know I can use a lyric from my favorite song of the moment.

How to import Windows Performance Monitor Data Collector Sets

2013-08-05T17:03:00.001-07:00

It's possible to export Windows Performance Monitor Data Collector Sets from perfmon.exe and then reopen the template on a different machine.

For a couple data collector sets on a couple different machines, this is not much of a problem. But if you have many performance data collector templates or many machines or both, you need an automated way to import the data collector templates.

An easier way is to use logman.exe to import the performance monitor data collector set.

Here's the magic:

logman import -name "MonitorName" -xml "MonitorTemplate.xml"

Create a batch file or Powershell script with that command for each data collector set and you can deploy multiple performance monitors to multiple machines quickly.

2012-10-10T22:38:00.002-07:00

How to open an Xbox 360

This is great guide on how to open an Xbox 360 because the instructions are simple and the photos are descriptive themselves.

I was able to open my Xbox 360 with a screwdriver from a kit I got a dollar store and a torx 10 driver.

I wish I had a torx 9 driver so I could take the heatsinks off and apply some Arctic Silver 5 to them, but so it goes.

Here's the "Open your Xbox 360" guide.

Selenium selectWindow method is broken

2012-03-06T13:27:00.000-08:00

The Selenium method 'selectWindow' has a longstanding bug in that Selenium cannot see new windows created by hyperlinks with a target specified. The Selenium method selectWindow will only select the newest window if the 'target' field is null or blank. In this regard, selectWindow functions like selectPopUp. Here is more detail:

Selenium selectWindow failing A ton of research and testing has gone into my conclusion that the method selectWindow simply does not work for anything but an empty string or NULL. The investigation started with Selenium IDE and continued in SeleniumRC with java in Eclipse. As of this writing the versions used were Selenium IDE 1.0.2 Selenium Remote Control 1.0.1 Firefox 3.0.11 IE 6 Use Case: 1. Navigate to the site’s main page 2. Click on a link that spawns a new window/tab 3. Select the new window 4. Validate stuff in the new window with asserts and verifies 5. Close the new window 6. Return to the main window 7. Rinse and repeat with other browsers and platforms ( IE6, IE7, FF2, FF3, WinXP, Mac, Vista, …) 1st attempt Firefox + Selenium IDE record/replay All steps recorded as expected except for the close method at the end open / assertElementPresent link=Texas Accessibility Policy click link=Texas Accessibility Policy selectWindow name=AccessibilityPolicy waitForTitle AccessibilityPolicy verifyText //div[@id='fullPage']/h1 Electronic And Information Resources Accessibility Policy This has never replayed correctly. Each time we get to the selectWindow step it fails and says “[error] Window does not exist. If this looks like a Selenium bug, make sure to read http:// selenium-core.openqa.org/reference.html#openWindow for potential workarounds.” The suggested URL was wrong so I had to search for their “workarounds”. In short, none of the workarounds work, which violates the basic premise of the term. They should be called “time wasters” instead because I have about 60 hours into this problem so far with no working tests in either IDE or RC that the QA group could use. What I have tried and observed so far: The first case that we tried was complex, in that the new window was a completely different domain as the original window. Main URL https://www.texascollegesavings.com/ New Window URL http://www.window.state.tx.us/accessibility.html This led me to believe that we were violating the JavaScript Same Origin Policy (http:// www.devarticles.com/c/a/JavaScript/JavaScript-Security/). In order to eliminate the same origin policy as a problem I changed to a different site Main URL New URL Since neither the domain or the port changed for the new window, this should not violate the security policy. Next step was to create a script that would conform to the use case. In the Selenium IDE recording produced: open waitForTitle verifyTitle clickAndWait selectWindow waitForTitle selectWindow waitForTitle At first I was ecstatic that it recorded very much the way I thought it should look. However on replay I got the exact same problem. On the step selectWindow | name=NEW | it returned more than 1 error. Output: • [info] Executing: |open | /HomeSite/HomePage/HomePage.asp | | • [info] Executing: |waitForTitle | OPnet, The OFI Corporate Intranet | | [info] Executing: |verifyTitle | OPnet, The OFI Corporate Intranet | | • • [info] Executing: |clickAndWait | //table[@id='TVAds_Table']/tbody/tr/td[2]/a/img | | http://opnet/HomeSite/HomePage/HomePage.asp http://opnet/departments/hr_contacts/learningdev/hrportal/Index.html /HomeSite/HomePage/HomePage.asp OPnet, The OFI Corporate Intranet OPnet, The OFI Corporate Intranet //table[@id='TVAds_Table']/tbody/tr/td[2]/a/img name=NEW E-HR Home Page name=null OPnet, The OFI Corporate Intranet • [error] Timed out after 30000ms [info] Executing: |selectWindow | name=NEW | | • [error] Window does not exist. If this looks like a Selenium bug, make sure to read http:// selenium-core.openqa.org/reference.html#openWindow for potential workarounds. • So the clickAndWait did not work since the original page was not ever re-loading, so turn that back to a click. But the last error is confounding because Selenium recorder captured the “name” of the new window so it must be able to “see” the window, only that’s not true on playback. Further research on the forums and documentation shows I should try using the waitForPopUp method waitForPopUp ( windowID,timeout ) Apparently name and windowID are synonymous, so I used the value of NEW which is the value of the target attribute on our main page that creates the new window. Again, it did not work • [info] Executing: |waitForPopUp | NEW | 2000 | • [error] Timed out after 2000ms • [info] Executing: |selectWindow | name=NEW | | • [error] Window does not exist. If this looks like a Selenium bug, make sure to read http://selenium- core.openqa.org/reference.html#openWindow for potential workarounds. This is when I bailed out of the IDE and started working with Java in Eclipse with Selenium Remote control. This way I could use some other methods like getAllWindowIds() getAllWindowNames() getAllWindowTitles() And I could also get some more logging options from the RC server. This led to some interesting discoveries. If the browser creates a new window using the method Selenium cannot see the new window with the name “NEW”. I came to that conclusion because all three getAll… methods return only the main window, not the child. windowIds String[1] (id=51) [0] "undefined" (id=55) windowNames String[1] (id=53) [0] "selenium_main_app_window" (id=56) windowTitles String[1] (id=54) [0] "OPnet, The OFI Corporate Intranet" (id=57) No wonder the selectWindow | NEW statement does not work, because the collection of available windows does not have that window in it. I can see the window open and investigate it’s properties with Firebug or at the Javascript console, but Selenium cannot. This means that an extremely common method of opening new windows in web programming is unusable and not testable using Selenium. Another interesting discrepancy is that the windowIds array shows the window id to be “undefined” which does not match the documentation for Selenium. Out of curiosity I modified the steps to perform the open window myself instead of emulating the click behavior on the object in the page. The steps: - capture the HREF attribute of the object and store it in a variable - use the openWindow method from Selenium to create a new window with an ID that I control and the href from the page The code: String eHRhref = selenium.getAttribute("//table[@id='TVAds_Table']/tbody/tr/td[2]/ a@href"); selenium.openWindow(eHRhref, "eHR"); Now when I run Selenium CAN see both the main window and the child window windowIds String[2] (id=47) [0] "undefined" (id=828) [1] "undefined" (id=829) windowNames String[2] (id=52) [0] "selenium_main_app_window" (id=54) [1] "eHR" (id=824) windowTitles String[2] (id=53) [0] "OPnet, The OFI Corporate Intranet" (id=823) [1] "E-HR Home Page" (id=826) However the window IDs still come up as “undefined” which does not match they say the ID should be. Continuing on to the selectWindow step, based on the Selenium documentation I should be able to use the name of “eHR” to select the new window, a name I made up. The code: selenium.selectWindow("eHR"); The output/error: com.thoughtworks.selenium.SeleniumException: ERROR: Window does not exist. If this looks like a Selenium bug, make sure to read http://selenium-core.openqa.org/ reference.html#openWindow for potential workarounds. at com.thoughtworks.selenium.HttpCommandProcessor.throwAssertionFailureExceptionOrError( HttpCommandProcessor.java:97) at com.thoughtworks.selenium.HttpCommandProcessor.doCommand(HttpCommandProcessor.ja va:91) at com.thoughtworks.selenium.DefaultSelenium.selectWindow(DefaultSelenium.java:343) at OpnetNewWindow.testOpnetNewWindow(OpnetNewWindow.java:60)… Window does not exist?! I made the window, I named the window, I asked Selenium if it can see the window and sure enough it can, now it says it cannot. This is what is called a defect. The thing just does not work as designed. I then played some more with the selectWindow method and found that only and empty string “” or the value of NULL would run without error.

How to install Puppy Linux from DOS

2012-02-08T16:59:00.000-08:00

After finally getting the PCMCIA CD-ROM to work, I was able to copy the contents of the Puppy Linux Live CD to the notebook.

The next problem to overcome was how to get Puppy Linux up and running from DOS.

Here's how to do it:

1. Download GRUB.EXE
2. Copy GRUB.EXE to c:\boot\grub
3. Create a menu.lst file in c:\boot\grub

If you're using a Puppy Linux 3.00 or later like I am, this for a bare bones get-up-and-running menu.lst worked for me.

title Puppy Linux 4.00
rootnoverify (hd0,5)
kernel /vmlinuz pmedia=idehd 
initrd /initrd.gz

GRUB.EXE requires absolute paths, so if you copied the Puppy Linux to a sub-directory on your hard drive, say to puppy400, your menu.lst would look like this:

title Puppy Linux 4.00
rootnoverify (hd0,5)
kernel /puppy400/vmlinuz pmedia=idehd psubdir=puppy400 
initrd /puppy400/initrd.gz

And just to reiterate: GRUB.EXE and menu.lst have to be installed in \boot\grub

I hope this helps.

How to check if Linux is 32 bit or 64 bit

2012-02-08T16:56:00.000-08:00

How can I determine what the bit architecture of my Linux distribution is? It would be a command or an environment setting, right? And the answer is:

>$ uname -m

x86_64



Stop Windows From Stealing Focus
2012-01-26T09:44:00.000-08:00
Here's how to stop Windows from stealing focus with 
   The easy method: Use the Tweak UI tool
Download Tweak UI from Microsoft. Scroll down the page and look for it on the right-hand side. Be sure to download the program to a location where you can easily find it, like your Desktop.
Note: Tweak UI is freely available software provided by Microsoft. Tweak UI is used to make a number of changes to your system that would otherwise require edits to the registry.
Double-click the downloaded file to install it. Follow the instructions provided during installation.
Navigate to Start, then All Programs (or Programs), then to the Powertoys for Windows XP folder.
Click on the Tweak UI shortcut.
In the Tweak UI program, click the + next to General on the left.
Click on Focus.
On the right, check the Prevent applications from stealing focus checkbox and then click the OK button.
From this point forward, programs should no longer steal the focus from the window that you're currently working in.
The Advanced Method: Use RegEdit.exe
Click on Start and then Run.
In the text box in the Run window, type regedit and click OK. This will open the Registry Editor program.
Locate the HKEY_CURRENT_USER folder under My Computer and click on the (+) sign next the folder name to expand the folder.
Continue to expand folders until you reach theHKEY_CURRENT_USER\Control Panel registry key.
Select the Desktop key under Control Panel.
On the right-hand side of the screen, locate and double-click on the ForegroundLockTimeoutDWORD.
In the Edit DWORD Value window that appears, set the Value data: field to 30d40.
Note: Make sure the Base option is set to Hexadecimal when entering the DWORD value.
Click OK and then close Registry Editor.
Reboot your PC for the changes you made to take effect.
From this point forward, programs should no longer steal the focus from the window that you're currently working in.


Uploading test scripts to HP Quality Center from Excel
2011-02-09T13:40:00.000-08:00
I got stuck at Step 5 of uploading test scripts from Excel to HP Quality Center using the HP Excel HP Add-in. Step 5 is the part about selecting a map name.

I would enter a map name and click 'Next' and the process would not progress.

The problem is that the Excel Add-in needs to have adminstrator privilege to create the map in QC.

The solution is to run Excel as administrator, open the spreadsheet with the tests, and then start the HP QC Excel Add-in to upload the test scripts / defects / requirements.

Answer at HP.


Change Windows XP Key
2010-11-08T11:11:00.000-08:00
Here is what you need: Windows Product Key Update Tool
There are a bunch of scripts on the net that supposedly remove the non-working Windows key from your Windows installation but many of them do not work. This tool from Microsoft does what you need it to do with a minimum of fuss.

If you're like me and have a bunch of keys from second-hand machines and can't remember which goes with which, this tool is handy for changing the key that was allowed for installation of Windows but won't validate with Windows Update or the Genuine Windows service. 





How to fix eDonkey LOWID problem in Shareaza
2010-02-24T14:09:00.000-08:00


Having a LOWID with eDonkey is a bad situation because when your client gets to the front of the download queue, it won't send the request for the file. Below is how to fix your Shareaza connection.



Your client will just still there waiting for the server to tell it that it may download the file and with the Shareaza implementation (a bug, maybe?), that will never happen.



So here is the solution so you can get a HighID:



1) Click "connection" in the Shareaza settings and enter port number below 65535. Do NOT select "random."
2) Open/forward the port in your firewalls and/or routers.        Port forwarding tutorials for almost all routers are at portforward.com.
3) Connect Shareaza to eD2k and check that your Shareaza port is  really        open on http://www.canyouseeme.org.        If it isn't, go back to step 1.
4) Low ID with open port? Disconnect from eD2k, hit F9 to open  "Discovery        Services," right-click and make sure only "Show server met's"        is checked, then delete all Server met's.
5) Hit F8 to open the Host Cache, select "eDonkey Cache," select        all eDonkey hosts, and delete 'em all.
6) Quit Shareaza, restart, and connect to eD2k


PCMCIA / CardBus drivers for Windows NT, XP, 98, DOS
2010-01-25T11:38:00.000-08:00
Here are some 16 and 32 bit PCMCIA / CardBus adapter drivers for Windows NT, XP, 98, and Dos.


Getting a video to work on a Sony NWZ-E343 / E344 / E345
2009-11-06T04:01:00.001-08:00
I tried a bunch of different methods to convert DivX videos to WMV so they would be watchable on my Sony NWZ E344 -- copying the files directly to the device, syncing via Window Media Player 11, using FormatFactory, but nothing worked.





I finally downloaded and installed the Sony Content Transfer application and it worked.




It's drag-and-drop and super easy to use. 




I should have started there.


How to upgrade Open Office on Ubuntu 9.04 Jaunty Jackalope
2009-05-28T16:48:00.000-07:00
How to upgrade from your earlier Open Office to the new version of Open Office 3.1 in Ubuntu 9.04 Jaunty Jackalope .  1. Add the repositories to "Software Sources"
  deb http://ppa.launchpad.net/openoffice-pkgs/ppa/ubuntu jaunty main
  deb-src http://ppa.launchpad.net/openoffice-pkgs/ppa/ubuntu jaunty main
 2. Copy this pulic key and save it to your desktop under the name "key"
-----BEGIN PGP PUBLIC KEY BLOCK-----
Version: SKS 1.0.10
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=rx3y
-----END PGP PUBLIC KEY BLOCK-----
 
3. Import the key using this command 
  sudo apt-key add /home/YourUsername/Desktop/key
If need be, change   /home/YourUsername/Desktop to location where you saved the key file.
 4.  Now update and upgrade
  sudo apt-get update && sudo apt-get dist-upgrade