This article originally appeared in issue 90 of Linux User & Developer magazine. Subscribe and save more than 30% and receive our exclusive money back guarantee – click here to find out more.

There was a time when UNIX desktops were developed by big corporations. It took not just one but many large firms to come together to build a desktop for the UNIX OS. The end result was CDE (Common Desktop Environment), developed jointly by Sun Microsystems, HP, IBM and UNIX Systems Laboratories. To its credit, it was a popular desktop environment used in almost all UNIX systems, but it was (and still is) not even close to being a decent desktop for most users. CDE was announced in June 1993. Windows was already available and CDE looked quite primitive by comparison. In 1996, the KDE project was started, followed a year later by GNOME, and the world of UNIX desktops changed for ever. It was quite surprising back then because neither KDE nor GNOME was a commercial project. Both started as open source, and both shared the same goal: to make Linux the best desktop operating system. But the philosophy was different. The KDE project wasn’t concerned with open source idealism – which is why GNOME was born, to create a desktop environment with fully GPLed software. This philosophical disagreement led to two completely different and innovative desktop environments for Linux.

Typical GNOME desktop

Typical KDE desktop

For your convenience the article has been broken down into a number of sub-sections which weighs up the various pros and cons for GNOME and KDE in various situation for both users and developers. Turn the page to get started…

Want to make sure your applications play nicely on both KDE and GNOME DE’s? Check out our quick guide here…

This article originally appeared in issue 90 of Linux User & Developer magazine. Subscribe and save more than 30% and receive our exclusive money back guarantee – click here to find out more.

“Mr Zuckerberg has an enviable record. He has done more harm to the human race than anyone else his age,” said Software Freedom Law Center founder Eben Moglen, in his ‘Freedom in the Cloud’ talk in New York, in February. The same talk that prompted four NYU students towards the idea of Diaspora* [the ‘*’ is part of the name] – a distributed social network, to be built using free software and open standards.

“The price – being spied on all the time – is too high,” said Moglen in the lecture. “I’m not suggesting it should be illegal: it should be obsolete. We’re technologists; we should fix it.” This (successful) call to arms came between two changes to Facebook that have people questioning what they have given away in return for a few pennies’ worth of hosting from the social network with half a billion users.

Who owns your data?
Discontent with Facebook emerged strongly when the terms of service were changed in February 2009, to allow Facebook the right to “Do anything they want with your content. Forever”, as The Consumerist’s Chris Walters put it. Facebook’s most recent change – turning all of your interests, favourite films and books, etc, into ‘Like’ pages – makes all of these interests public. Information to be sold.

Openbook uses Facebook’s search service to show public Facebook updates, exposing an alarming amount of personal information left open to the world. It’s plain that many Facebook users have no idea that recent changes to privacy settings have left them open to the whole internet, and probably don’t understand exactly how their information is read in order for them to be advertised to.

But for many the particular problem with Facebook – after all, let us remember that Google fillets Gmail messages for cues for advertisers too – is the way it sucks in all outside data, but lets none out. Just as once Compuserve users couldn’t email AOL customers, so Facebook users cannot use any of FB’s features to reach the outside world.

Nor can you take your contacts (the real value of your network) with you. Facebook won’t allow users to export their data – and have suspended those who tried web scraping to retrieve content, notably Robert Scoble. FriendCSV, which allowed users to export their data to Gmail accounts, seems to have disappeared from Facebook Apps.

Appleseed and OneSocialWeb
This analysis of the privacy problem, and a proposed federated solution, first emerged in 2004, when Appleseed was founded by Michael Chisari. Appleseed has working code, and recently development has resumed. Particularly interesting is an implementation of IM2000, an underused alternative to SMTP. Chisari has also talked open standards to OneSocialWeb.

Lacking the longevity of Appleseed, but slightly ahead of the rest of the pack, OneSocialWeb has the aim of providing a “free, open and decentralised alternative to the social networking silos that are Facebook and Twitter.” Developer Laurent Eschenauer is looking to work with Diaspora co-operatively if the chance arises, and has been on the GNU social mailing list since its inception.

This article originally appeared in issue 90 of Linux User & Developer magazine. Subscribe and save more than 30% and receive our exclusive money back guarantee – click here to find out more.

Drupal is a content management system (CMS) and also a content management framework (CMF). Hence it’s both an application that enables non-programmers to build complex websites, and a framework that developers can use to quickly create custom functionality. There is a large and rapidly growing community that has developed around Drupal, which is passionate about improving the software and supporting its usage.

More than two years in the making, and containing 6,600 community-contributed commits, the release of Drupal 7 is imminent. Major changes have been made which impact end users, site builders and developers. This article explores the most notable new features, but first some Drupal basics for the uninitiated…

Drupal basics
There are five layers in Drupal:
1. Data – These are ‘nodes’ (articles, comments, users).
2. Modules – Functional plug-ins to expand your site’s functionality.
3. Blocks & Menus – Provide output for sections of a page, which can include navigation.
4. User Permissions – Role-based, control what users can see and do.
5. Themes – Templates rendering the site in XHTML and CSS from the underlying data.

Central to Drupal is the concept of ‘nodes’, which is a generic term for a piece of content – for example a blog entry, a news article, a recipe. As administrators can create as many content types as are needed, each with its own set of fields, there is complete flexibility in what content can be added to a site. Content can also be tagged and categorised using a flexible taxonomy system.
Modularity is one of the major reasons for Drupal’s success, and there are currently over 6,000 modules that add various features, all available for free. At the framework level, modularity is provided by a hook system which enables modules to define functions to catch events generated from other modules, thus giving Drupal APIs great flexibility.

Now the basics are out of the way, let’s take a look at the new features in Drupal 7…

Drupal 7 Usability
With great power comes great user interface complexity, and Drupal has traditionally been perceived as being confusing and hard to use. To improve Drupal’s user experience, Mark Boulton and Leisa Reichelt were hired to create and test a major redesign for Drupal 7. This was funded by Acquia, the company founded by Drupal creator Dries Buytaert. The aims of the redesign were:

1. Make the most frequent tasks easy, and less frequent tasks achievable.
2. Design for the 80%.
3. Privilege the content creator – people spent most of their time managing content.
4. Make the default settings smart.

The most notable improvement to Drupal’s user interface are overlays. Usability lab tests observed that many users coming from other CMS systems were accustomed to a dedicated back end for administration and content creation, and found it difficult to distinguish the administration layer from the rest of the website.

Drupal 7 uses overlays to combine the best of both worlds, edit where you are, and the UI simplicity of a dedicated back end. The advantage is that you can instantly go into admin mode from anywhere on the site to change settings and then return back to where you were, thus maintaining context (Fig 1). Another major addition to Drupal’s interface are Dashboards, which provide the user with a quick means to access the relevant information and actions for the user’s role (Fig 2).

Forms have been redesigned to use fieldsets and vertical tabs to compress the size of forms on the screen (Fig 3). Contextual links are a handy new feature. A gear icon appears when hovering over blocks; upon clicking, a menu list available options. These menus are available for every block on the page and are programmable (Fig 4).

This article originally appeared in issue 89 of Linux User & Developer magazine. Subscribe and save more than 30% and receive our exclusive money back guarantee – click here to find out more.

Gadgets are awesome. We buy them, use them and program them. But do we build them? Most of us will say no. This is largely because it’s a project too big for an individual to do. It’s highly complex and needs a lot of investment in terms of hardware and software licences. But let’s say a fairy godmother came along (the good one, not the evil one) and made all these problems go away. Would you design your own gadget then? Of course you would. And all you need are some affordable hardware components and this guide to set you off on the right foot…

Designing your own gadget is not very different from building your own computer. The major difference is the form factor – the components are a lot smaller than for a laptop or PC. After all, you need to fit a whole computing system inside a small box, together with other components like the display or wireless radio.

Most modern gadgets (like the Google Nexus One smartphone or the iPad) are made with the following components:

1. SoC: Also known as system-on-(a-)chip, this is the most important part of a modern computing device. The SoC houses the complete system on one silicon chip. Modern SoCs cram the CPU, RAM and GPU onto a single tiny chip.
2. Storage: Storage can be provided by various means. The popular ones are NAND flash, SD cards and HDD (hard disk drive).
3. Display: Typically LCD, this could also be touch screen or another type of display system.

BeagleBoard is a low-power, low-cost single-board computer (SBC) that integrates SoC, storage and GPU on one board of size 3” x 3”. It was designed with open source and hobbyists in mind. And there are numerous open source projects around. But before we jump into the software side, let’s dig deeper into the hardware. BeagleBoard-xM is a revision of the original BeagleBoard and has only just been launched (with pre-orders being taken on the website from 7 June 2010). The xM makes use of an OMAP35x SoC which is based around the ARM Cortex A8 CPU.

You might be wondering about the ARM CPU here. It is the same processor that powers 95 per cent of the world’s embedded devices. It is the de facto standard for embedded computing and is being used in some of the world’s most popular gadgets, such as the Google Nexus One, Apple iPad/iPhone and Nokia Symbian smartphones. ARM is also the name of the processor company which is designing it, ARM Holdings plc, but it is not a manufacturer like Intel. ARM licenses its processor designs rather than making them. ARM CPUs can be manufactured by any of ARM’s licensees, such as Texas Instruments and Samsung. Most of the embedded world today runs on ARM (Advanced RISC Machine) architecture.

Android is changing the way that Linux is perceived. It has become the single most widely adopted type of Linux on embedded devices. It is not only popular in the smartphone space but also expanding its coverage to tablets, set‑top Boxes, televisions and appliances. For an Android application developer, this means a broader market to reach out to. We have already covered the introduction to Android development back in issue 83, so this time we go beyond the ‘hello world’ basics and give you the tips and recipes you need to become a better Android developer…

Using an alternative integrated development environment
an important part of any development platform. The IDE lets you make the most of your time by helping you focus on the code and logic rather than doing redundant tasks. Unlike other platforms, the Android SDK does not come with an IDE. Google provides an Eclipse plug-in called ADT (Android Development Tools) for Eclipse. This lets you use Eclipse as an IDE for Android. While ADT is good at basic tasks, it still is very limited in terms of functionality. As you grow as an Android developer, you’ll need more features and power from an Android IDE. So here are some alternative IDEs available for Android that make development both fun and easy:

MOTODEV Studio for Android (Free)
MOTODEV Studio is from phone maker Motorola. It includes the standard Google ADT and builds on top of it. Motorola has added several features that make Android development easier. Important features are…

1. Code snippets: You can use frequently used parts of code as snippets and use them in any number of applications you want. An example of a code snippet would be a code that is needed to initialise a database connection.
2. Database management tools: You can work with SQLite databases using a GUI interface without leaving the IDE.
3. Localisation file editor: This helps you manage strings to create localised applications.
4. Application creation wizards: It is possible to create essential Android classes – such as Content Provider, Service, Activity and Broadcast Receiver – quickly and easily using application creation wizards.
5. Automated SDK download: MOTODEV Studio is capable of executing automatic download, installation and configuration for the latest Android SDK. This is useful for newbies starting out in Android development.

Other features include application signing, Android Market integration and a built-in emulator. You can download MOTODEV Studio for Android from here.

IntelliJ IDEA (Commercial)
IntelliJ IDEA dares to be different in a world dominated by Eclipse-based IDEs. IntelliJ IDEA supports Android using an open source plug-in called idea-android (now part of the commercial distribution starting from version 9). Important features are:
1. Code insight support including code completion and navigation for Android projects.
2. Seamless Android SDK integration.
3. Android project deployment support.
IntelliJ IDEA is a popular Java IDE and has offers a host of excellent features that apply to most Java projects in general, as well as Android ones. IntelliJ IDEA can be downloaded here.

Don’t forget to follow us on Twitter or get your first digital copy of the magazine for iPhone and iPad free – just search for ‘Linux User’ on the app store now!

Spring is a lovely time of year, when the flowers bloom, the birds sing and community Linux projects release the fruit of their winter labours. Specifically, the Fedora and Ubuntu projects come to the end of their six-month cycles in the April/May time frame. This year’s yield is a bumper crop, with Fedora 13 including a number of interesting technologies and improvements and Ubuntu putting on its finest polish for a Long Term Support (LTS) release.

On the surface, one might expect that there would be little to distinguish the projects, since they each comprise the same core components: the Linux kernel, GNU utilities, X.Org, Firefox, GNOME and so on. But the reality is that the raw materials alone do not a distribution make. The focus of Fedora and Ubuntu is drastically different and it shows through in many ways.

The origin and ultimate disposition of the distributions gives a clue straight away as to the intended audience. Consider, Ubuntu is based on Debian testing, pulling packages from Debian after they’ve already received some work and been put through at least a minimal amount of testing and quality assurance on the Debian side. Basically, the Ubuntu Project spends time polishing and improving bits already worked on by Debian. That’s not to say that Ubuntu doesn’t do development or innovate above and beyond what Debian does. But the entire concept of Ubuntu was that Debian provided a solid base to provide a user-friendly Linux distribution, but Debian doesn’t itself provide a user-friendly distro.

In contrast, Fedora is the precursor to Red Hat Enterprise Linux. Whereas Ubuntu is a final product, Fedora itself is a staging ground for software that may ultimately wind up in Red Hat’s commercial offerings. Whereas Canonical pursues OEM deals with Ubuntu LTS releases, Fedora has an approximate 13-month shelf life, after which the project consigns the release to the end-of-life bin. In short, both projects produce usable desktop distributions, but for fairly different audiences.

As the version number implies, Fedora 13 is the 13th release coming from the Fedora Project. Lucid will be the 12th release for Ubuntu. The Fedora Project precedes Ubuntu by nearly a year, with its first release in November of 2003 and Ubuntu’s first in October of 2004.

The respective origin of each of the distros is also instructive. Fedora was offered by Red Hat as a substitute for Red Hat Linux after the company stopped its long-standing practice of releasing Red Hat Linux publicly, when it began a new product line of enterprise-focused releases. It was several releases before Fedora became a true community distribution in its own right, allowing significant community contribution and guidance.

Ubuntu, on the other hand, was designed to offer the same release to all-comers. Whether the distro was obtained from OEMs as a pre-load or downloaded by a hobbyist, Ubuntu should be the same distro. And unlike Fedora, Canonical started out with a development process that embraced the community from day one. At this point, both projects are legitimately created by co-operation between community contributors outside the companies and those employed by Red Hat and Canonical, respectively, to guide development. Let’s now take a look at how they stack up…

This article originally appeared in issue 86 of Linux User & Developer.
Linux User & Developer, one of the nation’s favourite Linux and Open Source publications, is now part of the award winning Imagine Publishing family. Readers can subscribe and save more than 30% and receive our money back guarantee – click here to find out more.

There are many alternatives to VMware’s expensive and proprietary software. Join us as we investigate four of the most prolific tools kickstarting the revolution in open source virtualisation…

Let’s start with a brief overview: Virtualisation leverages computing hardware by allowing you to run multiple computer workloads on a single system. Computer workloads take various forms: from shared instances of the same operating system to fully virtualised computers running as if they were standalone physical systems. Using virtualised infrastructure saves money by decreasing the amount of rack space used by physical hardware,
reducing power consumption and heat generated, and decreasing your
overall hardware budget.

The most popular software for desktop and server virtualisation is VMware. Its array of solutions is the industry standard. However, it’s expensive – so much so that it’s prohibitive for some small and medium-sized businesses to adopt. In a reaction to commercial virtualisation’s high entry costs, several companies (including VMware) and independent development communities have created free, open source software that compares favourably to VMware’s offerings. We shall examine four such products: KVM, QEMU, VirtualBox and Xen.

VMware might surface as the world’s best third-party-supported software of its kind, but is it significantly better than the free alternatives? Our exploration will give you the opportunity to find out for yourself.

These solutions, like VMware, are all of the full virtualisation type. Full virtualisation means that each virtual machine (VM) receives its own complement of virtual hardware, usually consisting of a virtual hard disk, a virtual NIC, a virtual CD/DVD drive, a virtual CPU and a virtual display adaptor. Optionally, a VM may also have virtual sound devices, floppy disk drives, multiple CPUs, multiple NICs, multiple hard disks and serial ports (including USB). In a fully virtualised VM, you install an operating system to the VM’s hard disks just as you would to a physical machine’s disks. VMs can boot from their virtual CD/DVD drives or from ISO files.

Alternatively, you’ll hear this type of virtualisation referred to as host/guest virtualisation, which describes the relationship between the physical machine (host) that provides the disk space, network connectivity, processing power and other peripherals and the VMs that reside on it as ‘guests’.

Let’s get started with the Kernel-based Virtual Machine (KVM) over the page…

This article originally appeared in issue 84 of Linux User & Developer magazine.
Linux User & Developer, one of the nation’s favourite Linux and Open Source publications, is now part of the award winning Imagine Publishing family. Readers can subscribe and save more than 30% and receive our exclusive money back guarantee – click here to find out more.

Linux has come a long way from being just a geek’s operating system. But there is no doubt that Linux is still the best operating system for geeks and developers. The effect of Linux as a development environment has not just been limited to Linux, but has spread on all systems including embedded devices. Projects like Firefox, FileZilla, Qt and SuperTux were originally created on Linux and then made their way to different platforms.

It’s certainly not just another operating system, but in fact a complete stack
of technologies that virtually any developer will enjoy. It doesn’t matter
if he or she is coming from Windows, Mac or any other OS – there is something for everyone. This feature looks behind the scenes and finds out what really makes Linux tick as a development platform…

Application development libraries
Application development libraries are the heart and soul of any program. Even for something as simple as a ‘Hello World’ program written in C, you’ll need the stdio (standard input/output) library, because the latter contains a function to display text onto a standard I/O device. So you see, the life of a developer would be nothing without such libraries.

Linux comes with a really large collection of development libraries for a wide variety of tasks. Most of the libraries available on Linux are often cross-platform. Essentially, this means that when you develop an application for Linux, you can easily port it onto other platforms such as Mac OS X and Windows.
Let’s look at the some of the popular libraries available on the Linux platform…

GUI toolkits
GUI toolkits help developers write GUI (graphical user interface) applications

Qt toolkit
Qt is a cross-platform application and UI framework. Qt is probably the best C++ toolkit currently available. And to prove its stability and functionality, there is KDE (K Desktop Environment) which uses Qt from top to bottom. Until recently Qt had a dual-licensing model which made its use in commercial applications very costly. With the release of new LGPL licence, Qt is friendly to both open source and commercial application developers.

Qt is very easy to use and it comes with its own set of development tools, including a multiplatform make tool (qmake), a GUI designer and a translation tool. Qt is not only supported on all the major operating systems (Windows, Mac, Solaris and so on), but also on embedded platforms like Maemo, Symbian and Windows Mobile. The best thing about Qt is that when you are porting your Qt application to other platforms, even on an embedded platform it requires very minimal code changes. It supports all the popular compilers, including GCC, Intel C++ Compiler and Sun Studio.

Qt is not just a collection of UI libraries: it also has support for non-UI-specific things like XML parsing, threading and networking, plus the wrappers around popular open source libraries such as CLucene and WebKit. This means that you will rarely need to use any native libraries when using Qt. Qt applications are very easy to translate. The Qt translation system consists of Linguist, which is an easy-to-use application for translation of your application user interface into many different languages.

Qt is available with almost all Linux distributions and can be installed using from your distribution’s package manager or from the Qt website. Lot of applications use Qt, including KDE, Opera, VLC, VirtualBox and Google Earth.

Every year was going to be ‘the year of Linux on the desktop’, until many began to wonder when the FOSS (free and open source software) breakthrough would come. As we watched for signs of hope in netbook sales, Drupal sites and partly-FOSS Android phones, a revolution was taking place all around, in the physical world but not entirely away from the internet.

Enter, the Arduino: a low-cost, open source, tiny hardware board for connecting the real world to your computer, and/or to the whole internet. What can be done with it? Everything. The limit is the imagination, and as you’ll see from a few of the example creations we review here, imaginative use is the norm.

Open hardware
Just as GNU/Linux software spread because everyone owned it, and could improve it, giving people the confidence to get involved, so with open hardware. The reference designs for Arduino are distributed under a Creative Commons licence (the software is GPL/LGPL, naturally), and the Italian company behind the boards, Smart Projects, is keen to welcome new contributors and variants. Boards are manufactured in different formats, sold all over the world, and the web is full of different designs if you want to build one yourself, whatever your skill level.
Numbers of boards in use are estimated in the hundreds of thousands, though like Linux distros, the freely copiable nature makes an accurate count difficult. What’s not difficult to see is the genuinely open nature of the online communities, and hackspace meetings, around Arduino projects. This has led to a cornucopia of wonderful projects from diverse artistic and creative people – but first, some history…

Those with long memories of geekery and the free software movement will know of the Tech Model Railroad Club (TMRC) – a group of MIT students who got together and played with trains. Some liked the modelling aspect, but the legendary Signals and Power Subcommittee implemented, in the 1950s and 1960s, a semi-automatic control system of great digital cunning, before acquiring a PDP-11 in 1970. TMRC members embodied early hacker culture, giving it much of its current vocabulary and terms of reference, and many became leading lights in DEC and other early computer companies, but this hacker culture fitted the American stereotype of the nerd, the socially awkward genius who could never get the girl (TMRC was, inevitably, an all-male club).

The hacker and free software movement has always suffered from an image problem – such that as female participation in professional IT has declined from 50% to 20% in the last 50 years, some free software projects have a derisory 1% female participation. It’s pathetic, boys, really – but there are signs of hope.

There’s more on page two…

The following links lead to downloads for the ‘Developing for Andoid’ tutorial that followed Adrian’s feature.

Icon Files – Click here to download the icon images
TextEdit – Click here to download the textEdit.java

With the right attitude, and some tips and tricks from industry experts, you can soon turn your homegrown Android app into a nice little earner. Adrian Bridgwater explains how…

Advisor:
Adrian Bridgwater
Adrian has been a technology journalist and IT communications consultant for nearly two decades. His journalistic creed is to bring forward-thinking, impartial, technology editorial to a professional and hobbyist software audience around the world.

There’s a fantastic book by former Intel CEO Andy Grove called Only The Paranoid Survive, which is themed around the turning points in the technology roadmap that led Intel to switch its business model and become the microprocessor giant that we know so well today. These ‘strategic inflexion points’, as Grove called them, were the points at which the paranoid business brain would realise the need for change and the opportunity to make money.

With the world of open source in mind, there comes a time in every developer’s life when however fulfilled they are as a result of working on their homegrown hobbyist projects, the allure of cold hard cash just can not be ignored any longer. Almost every open source project has an arm connected to a potential commercial channel where money can be made; and hey, even Sun Microsystems eventually had to “hand the stewardship” (as they carefully put it) of the Java language over to Oracle.

Commercial reality-check
So with these commercial realities under our belt, let us turn our focus to the Android mobile operating system, which of course runs on the Linux kernel. Initially developed by Android Inc and later bought by Google, Android now sits under the watchful stewardship of the Open Handset Alliance. Since the wraps came off the initial Android distribution back in late 2007, the operating system now sits with most of its code under the free and open source Apache License.

But enough of the basics, as most of you will know this preamble. When it comes to Android, the question remains: just when should you start to be paranoid and how and when can you expect to start making money from it? You’ve built your Android-based application and now you want to know how you should set about turning it into a revenue stream, right?
One of the problems is that developers face inherent platform fragmentation issues when launching new mobile content and services to the wider market. Companies such as Mobile Distillery produce technology that claims to reduce porting cycles by up to 80% as well as achieving cost savings by giving developers the chance to bring their mobile applications faster to market across more mobile devices.

Not so strange BREW
Having already simplified the porting process for mobile apps to native BREW (Binary Runtime Environment for Wireless), Windows Mobile and BlackBerry, Mobile Distillery says that it recognised the rise of Android and launched its Celsius Android solution to optimise and port applications to Android phones.
“We recognised that Android is becoming a platform of choice for mobile application developers and end users and wanted to lend our expertise to Android applications and so developed Celsius Android,” explains Vincent Berge, the co-founder and CEO of Mobile Distillery.
“Using Celsius, all aspects of mobile application development are optimised and, most importantly, porting cycles are faster. This in turn reduces the time to market of new mobile applications on any Android phone. The launch of Android offers many possibilities for the mobile market and with Celsius we want to help overcome any fragmentation issues so that the developer can be more powerful across all platforms and all phones,” adds Berge.
But of course optimisation and porting process improvements are only just part of the total equation. This is a complex cutting-edge segment of the application development market, after all; with so many impacting factors to juggle, surely a healthy level of paranoia and a wide-angle view are the most prudent tactics to adopt? Android proponents argue that the comparatively new application rankings capabilities are bringing the required commercial touch to the party.

Advisor: Swayam Prakasha
Swayam has been working in information technology for several years, concentrating on areas such as operating systems, networking, network security, electronic commerce, internet services, LDAP and web servers. Swayam has authored a number of articles for trade publications, and he presents his own papers at industry conferences.

A log file grows without limits unless some action is taken. Growing log files pose many problems since larger files are very difficult to manipulate and file systems can run out of space. Thus there is a need for a solution to prevent the log files from growing beyond a certain size…

Most Linux systems come with a tool called ‘logrotate’ and it is very handy in many situations. It will help the easy administration of systems that usually generate a large number of log files. Using this tool, we can rotate the log files, compress the log files, remove the log files, mail the log files etc. Rotating a log file means taking a copy to a back-up file and creating a new log file. We can then remove the back-up logs when they are out of date. We do not have to have any special privileges to run this logrotate. It is usually installed to run nightly as a cron job. For each service we install, we need to have a configuration file which tells logrotate how to rotate the log files for that specific program. In simple words, logrotate involves regular (typically nightly or weekly) moving of an existing log file to some other file name and then starting afresh with an empty log file. A part of a typical logrotate configuration file is given below.

# see “man logrotate” for details
# rotate log files weekly
weekly

# keep 4 weeks worth of backlogs
rotate 4

# create new (empty) log files after rotating old ones
create

# uncomment this if you want your log files compressed
#compress

The beauty of the logrotate tool lies in the fact that it allows us to handle each log file daily, weekly, monthly or whenever it grows too large. With logrotate, one can keep logs longer with less disk space.
The practice is to run logrotate as a daily cron job. Thus it is not needed to modify a log file several times in any day. If we have a program that keeps a log file and if that log files grows with each passing day, then logrotate can help up in cleaning it up for you. When called, logrotate reads the following two files:
a. The logrotate configuration file /etc/logrotate.conf.
b. The files in the logrotate configuration directory – /etc/logrotate.d. Note that most of the services (Apache web server, MySQL etc) installed on your system create a configuration file in /etc/logrotate.d.

The logrotate can be executed with many command line arguments. If no command line arguments are given, then logrotate will print version and copyright information along with a short usage summary. If it encounters any errors while rotating logs, it will exit with a non-zero status. Logrotate comes with many command line options and some of these are pretty useful. The following are some of the popular used options with logrotate:

-v – Turn the verbose mode on.
-d – Debug mode is enabled.
-f – Forcibly rotate the log files, even if it is not necessary.
– usage – This option prints a short message, giving information on its usage.

Configuration file format
The format of the log rotate configuration file is defined by the manpage for
logrotate. In a nutshell, the format is
as follows:
<global configuration options>
<definition for logfile 1>
<definition for logfile 2>
<definition for logfile 3>

As mentioned earlier, logrotate comes with a default configuration files and it is ‘logrorate.conf’ located at /etc. It is a normal practice to place the service-specific configuration files in the /etc/logrotate.d directory.

Now let’s see how we can use logrotate for rotating the log files…

1. Running Python scripts
On most of the UNIX systems, you can run Python scripts from the command line like so:

$ python mypyprog.py

2. Running Python programs from Python interpreter
The Python interactive interpreter makes it easy to try your first steps in programming and using all Python commands. You just issue each command at the command prompt, one by one, and the answer is immediate.
Python interpreter can be started by issuing the command:

$ python
kunal@ubuntu:~$ python
Python 2.6.2 (release26-maint, Apr 19 2009, 01:56:41)
[GCC 4.3.3] on linux2
Type “help”, “copyright”, “credits” or “license” for more information.
>>> <type commands here>

In this article, all the code starting at the  >>> symbol is meant to be given at the
Python prompt. It is also important to remember that Python takes tabs very seriously – so if you are receiving any error that mentions tabs, correct the tab spacing.

3. Dynamic typing
In Java, C++, and other statically typed languages, you must specify the data type of the function return value and each function argument. On the other hand, Python is a dynamically typed language. In Python you never have to explicitly specify the data type of anything. Based on what value you assign, Python will keep track of the data type internally.

4. Python statements
Python uses carriage returns to separate statements, and a colon and indentation to separate code blocks. Most of the compiled programming languages, such as C and C++, use semicolons to separate statements and curly brackets to separate code blocks.

5.  == and = operators
Python uses ‘==’ for comparison and ‘=’ for assignment. Python does not support inline assignment, so there’s no chance of accidentally assigning the value when you actually want to compare it.