Featured Post

Using USB CDMA modem to setup a home/private WiFi network

The idea here is to give a ‘how-to’ on setting up a a home WiFi network using mobile, CDMA based USB modem, hoping that bandwidth that you obtain from this service is satisfactory :-) Hardware Requirements: Linux PC/Notebook/Netbook Wifi Access point connected to ethernet port of the Linux box CDMA-1x USB Modem (Device used: Reliance M880) Software/OS Requirements: Linux Operating System with “usbserial” and “cdc_acm” device driver modules “pppd”, “wvdial”, “iptables”, and “dnsmasq” applications and their dependencies installed on the System Setup: The steps followed in setting up this CDMA based connectivity can be split into 3 steps, as follows. Configuring kernel modules for auto-loading at bootup Configuring “wvdial” application for dialup / ppp connection Initializing / starting the CDMA modem connection Setting up “iptables” for NAT forwarding Setting up “dnsmasq” for resolving DNS and as DHCP server Configuring kernel modules for auto-loading at bootup Open

Debian Squeeze and Mac OSX Lion - Dual boot

With Mac OSX Snow Leopard, things were much simpler to install and run a Linux operating system just by using rEFIt boot manager utility, on my Mac Mini(4,1 Mid-2010 model). With the introduction of Lion, Apple has introduced a hidden, OS recovery partition. In total, there will be 3 partitions for Mac OSX alone, for Lion. They are ESP (EFI System Partition), Mac OSX installation partition and, the OS recovery partition. Also, as per the GUID Partition Table(GPT) which Mac uses, there can be maximum of 4 primary partitions, with no scope for Extended partitions like in typical MSDOS Partition Table setup. It is also recommended to follow GPT when using Mac. With all these constraints, I'm actually left with one partition where I can install and run GNU/Linux (in this case, Debian Squeeze). I didn't try resizing an existing Mac OSX installation partition and creating an Ext4 partition for Linux. Rather, I choose to go for fresh installation. You can try with the resizing metho

Making persistant liveUSB from liveCD image

“Universal Serial Bus” is today one of the most common interface available in the computer to communicate with it in various aspects. Having a mobile storage device that seamlessly (well, at least in possible situations) boots on a standard i386 based hardware and allows one to perform his/her work will definitely make the life a lot comfortable. LiveUSB distro is meant to serve this purpose. Creating a LiveUSB distro using Ubuntu LiveCD iso on a typical Desktop PC / Laptop is explained with the following steps. These steps are meant to be for users already comfortable using command line utilities. For rest of the folks, there are readily available tools. Some of them are, UNetbootin and Liveusb . Pre-requisites: 1. Install ‘gparted’ , ‘syslinux’ , ‘mtools’ and ‘squashfs-tools’ using Synaptic package manager or apt-get using a terminal. 2. Download the latest Ubuntu LiveCD distribution for Desktops, from . 3. You need to have superuser access

Creating Linux filesystem images

I’m sharing the following script which creates a filesystem image of desired size (in kBs) for use in building an installUSB/LiveUSB Pen-drive image or for similar uses. I have written this in my effort to create a installUSB image for Debian-Lenny. Hope it is of some help to somebody:-). You may copy the italicized/quoted text below into a new text file and save it as a shell script (.sh extension). Don’t forget to set its executable bit;-) and also, please don’t forget to leave your comments/feedbacks. Note: It currently creates images for fat16, fat32, ext2 and ext3 and you should run it as root/superuser (sudo). You can extend this script to generate fs images for other filesystem types as well. Usage: script-name fs-type output-file required-size-kB Ex.,; ./ ext3 myimage.img 600000 This will create a filesytem image called myimage.img of type ext3 and of size ~600MB. #!/bin/bash if [ $UID -ne 0 ]; then echo “You need superuser/root privileges to

How to extract an initrd image

Well, I had this often repeated problem of searching for how to extract an initrd image only to find that it is basically as cpio+gzip archive but, over a period of time forgetting it so that I come back searching for this again:-) Finally, I made up my mind to put in what all it takes to extract an initrd image in Linux and re-packing it after customization (not covered here, as customization depends on the need), here. Unpacking the initrd image: Create an empty directory /tmp/initrd (this is an example; you are free to use any other directory name) and cd into it, using the following command (Replace /tmp/initrd with your directory name, if using a different directory name) mkdir -p /tmp/initrd; cd /tmp/initrd; Extract the initrd image using the following command gunzip absolute-path-to-initrd-image | cpio -i —make-directories Now, the contents of the initrd image will be there in /tmp/initrd (or whatever directory you have used). Go ahead and do the necess

Using git over proxy

I was trying to clone Android repository at office when this problem of “Connection refused” started coming when running the ‘repo’ tool. On subsequent googling, I got this really useful link (thanks to Emil Sit) which explained how git can be used over http proxy for those git servers which don’t allow http method as an alternate/bypass to git protocol (As it looks is the case with Android git repository). Ok, in short, here is what I have done; Typed in the below lines (quoted) into a shell script called gitproxy and put it in $(HOME)/bin directory. Of course, its executable bits has to be set with chmod a+x $(HOME)/bin/gitproxy. My $(HOME)/bin directory is already part of PATH so that I can access my custom scripts easily. #!/bin/sh # Use socat to proxy git through an HTTP CONNECT firewall. # Useful if you are trying to clone git:// from inside a company. # Requires that the proxy allows CONNECT to port 9418. # # Save this file as gitproxy somewhere in your path # (e.

Custom d-i banner for your Debian blend

The GUI version of the Debian Installer (d-i for short) comes with a pinkish banner with Offical debian logo and GNU/Linux text embossed. If you are going to use live-installer tool to create a Debian Blend you may as well like to have your own d-i banner for your blend. Here are the quick steps: Open GIMP image editor tool and create a new image of size 800x75 pixels. 800 corresponds to vga=788, which is the default command line option for vga, passed in boot menu. Create the banner that suites your tastes and save the image as banner.png Select menu->image->Flatten Image to flatten the image (i.e., remove Alpha) and save it again. Now, your banner is ready to use. You can use a custom binary-hook script in live-helper such as the following, to get your banner replacing the original d-i banner. The below hook is derived from live-helper configuration for Webconverger . #!/bin/sh # This is an binary hook for live-helper(7) to overwrite the banner # in the graph

Creating GRUB/Syslinux splash images

GRUB splash image creation: Open GIMP application. Create an image of size 640x480 pixels. You may open an existing image and resize it too. Reduce the number of colors to 14 by selecting menu item “Image->Mode->Indexed”. Don’t use any color dithering. Save the resulting image as splash.xpm and close GIMP. Compress the image file by running; gzip splash.xpm Copy the resulting splash.xpm.gz file to /boot/grub. Run update-grub to upgrade the grub configuration to include this splash image. Syslinux splash image creation: Open GIMP application. Create an image of size 640x480 pixels. You may open an existing image and resize it too. Reduce the number of colors to 14 by selecting menu item “Image->Mode->Indexed”. Don’t use any color dithering. Save the resulting image as splash.ppm and close GIMP. Install the package syslinux-common by running; apt-get install syslinux-common as sudo/root.  Convert the saved image to syslinux format by running; ppmtolss16 ‘#fff