There has been many instances of website hijacking for many WordPress site owners. The root cause is wrong permissions set on WordPress core files. Here’s a general step by step process to harden your WordPress installation and prevent it from being hijacked by asshole hackers.
To prevent security vulnerabilities and hijack attacks, it’s important to give appropriate permissions to WordPress files. The rule of thumb is no file must be given more permission than it needs or you might end up getting screwed. My WP hardening involves a several-step process (For NGINX):
The permissions we set previously allow some code to be executed in special dirs i.e. uploads, themes, plugins, upgrade, etc. To contain this threat, we can add further control to prevent any code from execution in this dirs. In NGINX’s server block for this virtual host, add:
# Deny PHP execution in wp-content/uploads
location ~* ^/wp-content/uploads/.*\.php$ {
deny all;
}
# Deny PHP execution in wp-content/plugins
# This is typically good practice, but be aware some plugins
# *might* legitimately place PHP files here that need to run.
# Assess carefully.
location ~* ^/wp-content/plugins/.*\.php$ {
deny all;
}
# Deny PHP execution in wp-content/themes
# Similar to plugins, themes could have PHP files needing execution.
# Use with caution and test thoroughly.
location ~* ^/wp-content/themes/.*\.php$ {
deny all;
}
# Deny PHP execution in wp-content/upgrade (temporary files during upgrades)
location ~* ^/wp-content/upgrade/.*\.php$ {
deny all;
}
# Add other writable directories as needed, e.g., wp-content/cache
location ~* ^/wp-content/cache/.*\.php$ {
deny all;
}
In Apache, you can create a .htaccess file under the uploads dir or any other dir you are concerned about with this content:
<Files *.php>
Deny from all
</Files>
<FilesMatch "(?i)\.(php|phtml|php3|php4|php5|php7|phps)$">
Deny from all
</FilesMatch>
Note: At the time of this writing, May 26 2025, Kotlin Multiplatform is only available for macOS. So, we will create the project in IntelliJ on macOS and then we can also build it on the other platforms using Compose Multiplatform plugin. here are the steps that need to be taken to create and build a Kotlin Multiplatform app:
There are two ways to create a KMP project:
Press File->New Project and then choose Kotlin Multiplatform on the left. On the right, choose the platforms you target, e.g. Desktop. iOS, Android, etc. I only choose Desktop because we focus on Desktop apps.
To build the the app on macOS, use the default task defined by the project (desktopApp in the list of Run configurations).
Once you installed a JDK version 21 and above, specify the jdk path in the gradle configuration. You can install various JDK versions using winget in Windows Terminal. Note that Gradle prefers forward slash for paths, rather than Windows’ “"” separator. So, change the path accordingly. For example, in my case it is:
org.gradle.java.home=C:/Program Files/Eclipse Adoptium/jdk-21.0.7.6-hotspot
To build distributed packages like dmg (on mac), msi (Windows) and flatpak (Linux), run the following gradle command:
./gradlew packageDistributionForCurrentOS
# on Windows run: ./gradlew.bat packageDistributionForCurrentOS
RabbitMQ can be used for task queue management in applications that need some time to perform something. With containers, it’s quite easy to start a group of RabbitMQ containers, some as workers, some as producers. Here, I will show how to create a setup with a single worker and a single producer. The worker will wait for messages from the queue and the producer will send messages to the worker through a queue.
Note: I will be using Podman to create and manage containers 0- Create env.ini with the following contents:
[rabbitmq]
RABBITMQ_DEFAULT_USER=myuser
RABBITMQ_DEFAULT_PASS=mypass
RABBITMQ_DEFAULT_VHOST=rabbit1
1- Create Podman network:
podman create network rb
2- Create Worker container:
podman run -d --hostname rabbit1 --name rabbit1 -p 8080:15672 --env-file env.ini --network rb docker.io/library/rabbitmq:3-management
3- Create producer container:
podman run -d --hostname rabbit2 --name rabbit2 --env-file env.ini --network rb docker.io/library/rabbitq:3
4- Enter the worker and install pika:
podman exec -ti rabbit1 bash
apt update && apt install -y python3-pika nano
5- Enter worker code: While in the container, do:
mkdir code
cd code
nano server.py
#!/usr/bin/env python
import pika
import time
credentials = pika.PlainCredentials('myuser', 'mypass')
params = pika.ConnectionParameters('rabbit1', 5672, 'rabbit1', credentials)
connection = pika.BlockingConnection(
params
)
channel = connection.channel()
channel.queue_declare(queue='task_queue', durable=True)
print(' [*] Waiting for messages. To exit press CTRL+C')
def callback(ch, method, properties, body):
print(f" [x] Received {body.decode()}")
time.sleep(body.count(b'.'))
print(" [x] Done")
ch.basic_ack(delivery_tag=method.delivery_tag)
channel.basic_qos(prefetch_count=1)
channel.basic_consume(queue='task_queue', on_message_callback=callback)
channel.start_consuming()
# Run the code
python3 server.py
6- Go inside producer container:
podman exec -ti rabbit2 bash
apt update && apt install -y python3-pika nano
mkdir code
cd code
nano client.py
#!/usr/bin/env python
import pika
import sys
credentials = pika.PlainCredentials('myuser', 'mypass')
params = pika.ConnectionParameters('rabbit1', 5672, 'rabbit1', credentials)
connection = pika.BlockingConnection(
params
)
channel = connection.channel()
channel.queue_declare(queue='task_queue', durable=True)
# Create a list of files to simulate list of operations
files = [
'file1.mp3', 'file2.mp3', 'file3.mp3'
]
i = 0
while True:
message = files[i%len(files)] # iterate over the list of files to simulate operations
i = i+1
channel.basic_publish(
exchange='',
routing_key='task_queue',
body=message,
properties=pika.BasicProperties(
delivery_mode=pika.DeliveryMode.Transient
))
print(f" [x] Sent {message}")
connection.close()
# Run the producer
python3 client.py
If you did all the above steps, then the management panel is available at the following address:
http://localhost:8080
Once you opened the page, you will be asked for username and password. Enter myuser and mypass, the same values entered in env.ini file, respectively for this. After being authenticated, you will be able to see Overview, Connections, Channels, Exchanges, Queues and Streams and Admin sections, each of which contains valuable information.
Sometimes you need to create partial slides for product carousels, in such a way that only 1 product and half of the next product is visible in the viewport. This is appropriate especially in mobile view. It also encourages the user to slide the carousel to see the next item in the list.
There are probably some plugins that allows creating partial slides. You might also want to use some custom CSS. However, in recent versions of Elementor Pro you can do this in the Loop Carousel using its settings. Here are the steps:
bcdedit /enum firmware
Select EFI firmware identifier e.g. {abcdxxxx}
Delete the entry from store:
bcdedit /delete {abcdxxxx}
Delete files for the EFI partition of the deleted entry:
🟢5-1:
diskpart
🟢5-2:
list disk
🟢5-3:
select disk [1]
Replace [1] with the correct number from “list disk”
🟢5-4:
list vol
🟢5-5:
select Volume [0]
Replace [0] with number for the volume that is 100 MB (System Partition)
🟢5-6:
assign letter z
z can be any letter
🟢5-7:
Exit diskpart
🟢5-8:
Go to z
z:
🟢5-9:
dir
🟢5-10:
Go to EFI directory
cd EFI
🟢5-11:
Find the directory for the deleted Linux distro (e.g. Ubuntu)
🟢5-12:
Remove the directory:
rmdir /s Ubuntu
(Replace Ubuntu with actual distro name)
]]>In order to set a web proxy on MacOS’ terminal, we will use networksetup utility:
networksetup -listallnetworkservices
It will show a list like this:
An asterisk (*) denotes that a network service is disabled.
Thunderbolt Bridge
Wi-Fi
Freedom VPN
V2BOX
v4freedom
FoXray
From the output, we know that Wi-Fi is what we’re looking for.
networksetup -setwebproxy Wi-Fi 127.0.0.1 19999
networksetup -setsecurewebproxy Wi-Fi 127.0.0.1 19999
That’s it. You should be able to use the proxy now.
networksetup -setwebproxystate Wi-Fi Off
networksetup -setsecurewebproxystate Wi-Fi Off
Using On in the above commands will turn on the proxies.
TODO
]]>Here are the steps that I took in trying to root a Nokia 3.2 device. However, because of the problems that process created, I had to reverse the process and thankfully get to a working state.
Today I had to install driver for HP LaserJet 1200 on a client’s computer and I had a hard time making the printer work. The reason: The latest Universal PCL 6 printer driver offered on the HP site turned out not to be compatible with Windows 10, through there nothing written about it on the downloads page. I kept seeing an error message on the printed test pages.
Error was something like: Pcl XL error Unsupported Protocol
It struck my mind that the printer, being an old machine, might need an older printer driver like PCL 5, and a search on Google showed that PCL 6 (a.k.a PCL XL) is very different from PCL 5. So, I looked for LaserJet 1200 printer PCL 5. There are not many options on the Internet. Finally, I found the driver’s files in a zip on Driver Pack’s website. You don’t need the Driver Pack app to install the driver.
I assume you downloaded and unzipped the driver for HP LaserJet 1200 PCL 5 on your system. To install the printer correctly, go to printers and hit Add Printer. Then select the model from the list and then find the button Have Disk instead of selecting Windows drivers. Then, give it the path where you unzipped the driver. It will install the driver and might even launch an installer mid-way. Nevertheless, you should have no problem after the driver gets installed. Hit Print Test Page on the final page and get your first (hopefully) successful printout.
In case your printer does not print multiple copies of a page despite being told so in the print page, do the following:
Caching in web applications is important because there are many requests sent to the server all the time and without good caching mechanisms, the performance of the site and users’ browsing experience will not be optimal.
W3 Total Cache is a plugin that allows you to configure several types of caching mechanisms for your site. Once installed, it will show you a wizard guiding you through the configuration steps. The wizard consists of several parts the follow next.
This speeds up Time to First Byte. There are several options e.g. Disk,Disk: Enhanced, Redis, etc. The recommended one is Disk: Enhanced as in my case, it shows the least amount of time in ms (22.49).
Database cache allows caching queries. The difference between the default database performance and cached ones is not significant, so we should leave it as None.
This cache allows reusing objects cached by WordPress during page renders. I choose Memcached here because it takes only 1.77 ms
This is an important setting that allows setting cache headers for assets, saving browsers from downloading assets like CSS and JS files on each page render. So, you should set it to Enabled.
Check Enable WebP Converter to convert image files to lightweight web-friendly format WebP, thus improve performance.
Check Lazy Load Images option to reduce page loads by deferring image loads until they are needed.
]]>