Free born to run software for linux

Configuration HOWTO would be the main documentation for configuring most common hardware and services.

This document is one of the most important for LDP, because by configuring hardware and software you can get your own Linux box. This HOWTO was born in the Golden Age of developers, mainly for the command line. As it became too big and old for the current distros, I rewrote it more simple as I could.

Fundamentally, to configure the system, Linux users have to write some configuration files. To do it easyer, today programs and wizards are avaliable to manage them. This programs may be quite different for the various distributions.

In this HOWTO, I will speak about Mandrake-Linux and about Red Hat.

Bio::Tools::Run::TribeMCL is a method for clustering proteins into related groups, which are termed protein families.
SYNOPSIS
use Bio::Tools::Run::TribeMCL;
use Bio::SearchIO;
# 3 methods to input the blast results
# straight forward raw blast output (NCBI or WU-BLAST)
my @params = (inputtype=>blastfile);
# OR
# markov program format
# protein_id1 protein_id2 evalue_magnitude evalue_factor
# for example:
# proteins ENSP00000257547 and ENSP00000261659
# with a blast score evalue of 1e-50
# and proteins O42187 and ENSP00000257547
# with a blast score evalue of 1e-119
# entry would be
my @array = [[qw(ENSP00000257547 ENSP00000261659 1 50)],
[qw(O42187 ENSP00000257547 1 119)]];
my @params = (pairs=>@array,I=>2.0);
# OR
# pass in a searchio object
# slowest of the 3 methods as it does more rigourous parsing
# than required for us here
my $sio = Bio::SearchIO->new(-format=>blast,
-file=>blast.out);
my @params=(inputtype=>searchio,I=>2.0);
# you can specify the path to the executable manually in the following way
my @params=(inputtype=>blastfile,I=>2.0,
mcl=>/home/shawn/software/mcl-02-150/src/shmcl/mcl,
matrix=>/home/shawn/software/mcl-02-150/src/contrib/tribe/tribe-matrix);
my $fact = Bio::Tools::Run::TribeMCL->new(@params);
# OR
$fact->matrix_executable(/home/shawn/software/mcl-02-150/src/contrib/tribe/tribe-matrix);
$fact->mcl_executable(/home/shawn/software/mcl-02-150/src/shmcl/mcl);
# to run
my $fact = Bio::Tools::Run::TribeMCL->new(@params);
# Run the program
# returns an array reference to clusters where members are the ids
# for example :2 clusters with 3 members per cluster:
# $fam = [ [mem1 mem2 mem3],[mem1 mem2 mem3]]
# pass in either the blastfile path/searchio obj/the array ref to scores
my $fam = $fact->run($sio);
# print out your clusters
for (my $i = 0; $i
print "Cluster $i t ".scalar(@{$fam->[$i]})." membersn";
foreach my $member (@{$fam->[$i]}){
print "t$membern";
}
}
This clustering is achieved by analysing similarity patterns between proteins in a given dataset, and using these patterns to assign proteins into related groups. In many cases, proteins in the same protein family will have similar functional properties.
Enhancements:
- Perl

Hotplough is a Linux hotplug subsystem support tool. It is started whenever the hotplug event occurs - e.g. when you insert a USB device, ifconfig up your network interface, plug in the PCMCIA card, etc. The task of hotplug support tool is to load appropriate modules into kernel and do whatever else the administrator of the system wants it to perform.

Once I upgraded my kernel to 2.6 branch I found out that the murasaki hotplug support system isnt working anymore. Checking out the murasaki homepage at that time was of little help. The usual upgrade story.

So I tried the hotplug support script, mentioned somewhere in kernel documentation or the post-halloween document... Well, it worked. But, even though I dont reboot daily, I thought I cant afford the luxury of wasting 10-15 minutes on the mere bootup. Besides, hotplug events may happen to occur not only when I reboot and to put such a load on the machine when I, say, synchronize my handheld doesnt sound appealing either.

These unfortunate events took place, doubtlessly in order that the next KIN might be born. And yes, I decided to rewrite the whole thing. No wonder that when redone in c, things work tons faster.

Thus came the hotplough. If you run the tool with no parameters (kernel will never do so) with root privileges and /proc filesystem mounted it will install itself as a hotplug handler in the /proc/sys/kernel/hotplug. This should be done somewhere in the very beginning of your initscripts.

Speaking of which. You should receive a word of warning here. So, let me explain why you dont want to use hotplough. Most likely youre using one of the countless linux distributions that do everything for you and in return expect you to maintain compatibility with them. Since I do not use any distribution (real men compile from sources, dont they?) I didnt have to and didnt want to make hotplough compatible with any existing implementation, even though it may work in similar manner.

And for the brave souls who keep reading on, another warning: the current version lacks features. In particular theres no support of PCI hotplug which I am not using and dont expect to start using in the nearest future. It also lacks PCMCIA hotplug support which is likely to come as soon as I will install linux on my notebook which is not used now. If you feel like implementing missing features yourself, your patches are welcome.

Now that the weak of will are scared off, the only thing left to explain is where to store your hotplug agents (i.e. executables to be launched at particular hotplug events). Unless you specify otherwise, the base directory for agents is ${prefix}/etc/hotplough. It is expected to have subdirectories for different device classes, such as bluetooth, net, tty, usb and the most special (or better the least special) of them - generic.

At this time if you want to know more details on how to lay out the aforementioned directory or what d-bus signals hotplough sends you will have to either look up the code or ask a specific question.

Bio::Tools::Run::PiseWorkflow is a class to create a Pise workflow using Pise application objects as methods. A workflow is defined by a set of methods which all instanciate the class PiseApplication.

SYNOPSIS

# First, create a Bio::Tools::Run::AnalysisFactory::Pise object:
my $factory = new Bio::Tools::Run::AnalysisFactory::Pise();
# Then create the application objects (Pise::Run::Tools::PiseApplication):
my $clustalw = $factory->program(clustalw);
$clustalw->infile($my_alignment_file);
my $protpars = $factory->program(protpars);

# You can specify different servers for different applications :
my $protdist = $factory->program(protpars
-remote => http://kun.homelinux.com/cgi-bin/Pise/5.a//protpars.pl,
-email => your_email);

# Create a new workflow object :
my $workflow = Bio::Tools::Run::PiseWorkflow->new();

# Define the workflows methods using the application objects:
# the application method $protpars will receive the output of
# type readseq_ok_alig from the application method $clustalw.
$workflow->addpipe(-method => $clustalw,
-tomethod => $protpars,
-pipetype => readseq_ok_alig);

# The application method $clustalw will be piped to a second
# application method ($protdist) using the output of type readseq_ok_alig.
$workflow->addpipe(-method => $clustalw,
-tomethod => $protdist,
-pipetype => readseq_ok_alig);

# The application method $protpars will be piped to the application
# method $consense using the output of type phylip_tree.
my $consense = $factory->program(consense);
$workflow->addpipe(-method => $protpars,
-tomethod => $consense,
-pipetype => phylip_tree);

# Run the workflow.
$workflow->run();

rinse software was born out of a frustration with the currently available solutions for installing minimal copies of RPM-based distributions.
The need to perform simple chroot() installations of other distributions is not common, but the ability to do such a thing is very useful when it comes to testing new tools, and working towards the creation of new Xen guests.
The following distributions are supported:
- Centos 4 & 5
- Fedora Core 4, 5, 6 & 7
I plan to support OpenSUSE in the near future.
Enhancements:
- This is primarily a bugfix release, which restores the ability to install new distributions of Centos 4 & 5.

PyMusicD is yet another jukebox daemon written in python. It was born out of my frustration with the current mp3 jukeboxes, mostly the lack of features that I want. Ive also been playing with python quite a lot lately, and this is my first attempt at a large project in python.
The player is being written to satisfy my own requirements for a car mp3 player, although theres no reason you couldnt use this for a home mp3 player, or work mp3 player.
It assumes that your network users are (mostly) benign, and doesnt take a lot of precautions against malicious users who want to screw with your mp3s. Take appropriate measures to only allow trusted users to play with it, at least until I secure it a bit better.
Its primary purpose is to run in the background, and play mp3s whenever they are in the playlist.
Usage
Unpack this to its own directory. Edit the pymusicd.conf file to your liking. Run python PyMusicD, then run python pymusic to play around with it. The client operates very similarly to most shells, and includes tab completion for commands (but not arguments... yet.)
If you wish to install this to the system, you may do so by running python setup.py install. This will install the server binary in the default place for your platform (according to distutils.) You may then edit the config file (/etc/pymusicd.conf by default) and run the server by running PyMusicD. The cli client will be named pymusic.
This software is currently in development. It is not ready for end users yet. You should have familiarity with scripting in python before you set about to use this software. That being said, if you find bugs that arent listed in TODO, and/or have problems with the software that arent listed in TODO, email me (zwhite@darkstar.frop.org) and Ill see what I can do to help you. Code patches/suggestions are always welcome.
This is known to work on my Slackware 9.0 machine using the Python 2.2.2 package that came with it. I have also tested it on my Mac OSX 10.2.8 machine with Python 2.2. I have no reason to believe that it wont work on any machine with a Python 2.x interpreter and
either mpg123 or mpg321. Your mileage may vary. If it doesnt work on your platform, please let me know. My goal is to have this work on any platform that python runs on.
Enhancements:
- Fixed a bug when adding an entire directory with %2A instead of *
- Added a debug command to get server state, currently only playstatus is reported, will add more as needed.
- Changed the way the stop command works to eliminate a bug.
- Check to make sure a file exists before we add it to the playlist.
- We no longer start playing music as soon as the playlist has entries.
- Instead, we wait for a play command to be issued.
- Added config file support. Defaults to /etc/pymusicd.conf or ./pymusicd.conf
- Added a setup.py and setup.cfg file. Now users can install PyMusicD using the standard "python setup.py install" method that other scripts and modules use.

SlackPen is a Live CD based on Slackware Linux. Our current goal is to offer everything necessary to perform a complete security audit of your network, in a low overhead environment.
The end goal of SlackPen is to provide an easy installer for our other project SlackWall, a Slackware based firewall distribution. It seemed like a good idea to make the installer useful after the install, SlackPen was born.
Unlike some other live cds with common goals, SlackPen is not intended to be user-friendly (though some of the tools are). We assume that the users of SlackPen have some experience with linux, and know their way around.
Check out the Screenshots section to see how SlackPen looks.
SlackPen distribution was built using Slackware and the linux-live scripts written by Tomas Matejicek.
Main features:
- Slackware Linux based live cd
- Lightweight using Fluxbox, and many console based apps for common tasks such as Instant Messaging.
- Runs nicely on new and old hardware.
- Tools you need. SlackPen comes packed with a large collection of security/auditing tools no whitehat should be without.
- SlackWall installer. (Coming soon!)
Enhancements:
Lots of new tools
- New custom 2.6.13 Kernel
- hydra 5.1 upgraded to hydra 5.2
- ClamAV 0.87 upgraded to ClamAV 0.88
- Added Tor for anonymous web browsing
- Added the ToolBox in fluxbox to launch cli apps.
- Added Ethereal 0.10.8
- Added Nessus 2.2.6
- Added wmnet and wmBinClock
- Added Host-ap 0.4.7 (SlackPen is pre-configured for use with prism2 based cards)
- Added Ndiswrapper 1.7
- Added ipkungfu 0.6.0
- Added chkrootkit 0.46a and rkhunter 1.2.7
- Added BashBurn
- Added Amap 5.2
- Added Etherwake 1.06
- There are a few others Im forgetting at the moment, still need to compile an apps/tools list. The new nmap rolled out while building the iso, so that will come next release (which might be 1.0)
- Other notable changes are a new isolinux bootsplash, 1024x768x256 vga boot.