which date

For most servers it is:

/bin/date

This is the same for perl and php

Lately, I’ve been noticing stories about how to use Linux you need to know half-a-hundred Linux shell commands and the like. Ah, what century are you from? Today, if you can see a window and handle a mouse, you’re ready to use Linux.

And no, I’m not talking about how we’re all already using Linux in devices like the TiVo or the Droid smartphone and through Linux-powered Web sites like Google. I’m talking about using Linux on the desktop.

There is nothing — I repeat, nothing — that requires any special knowledge to use Linux on the desktop today. If you’ve already mastered Windows XP, you’ll have little more trouble moving to a Linux desktop like Red Hat’s Fedora 12; Novell’s openSUSE 11.2; or Canonical’s Ubuntu 9.10 than you would in switching over to Windows 7.

I’m not saying using Linux isn’t different from running Windows. It is. For example, you’ll need special software like Crossover Linux to run Windows-specific software.

The interfaces also aren’t the same — but then, Windows 7 and Vista’s interfaces aren’t the same as XP’s, and Mac OS X’s Aqua interface doesn’t look anything like the others. Besides, can any other operating system besides Linux let you set up the interface so that it duplicates XP’s look and feel? I think not!

What you don’t need to use desktop Linux is to learn dozens of obscure Linux shell (aka command line) programs to get work done. Neither do you need to know how to edit configuration files by hand to get Linux set up properly.

Sure, it can help to know how to use the Unix/Linux shell. I was writing shell (awk, sed, and grep) scripts to get work done in Unix, and later Linux, before many of you played your first game of solitaire on Windows 1.0. My point is, for ordinary, everyday use, you don’t need to know anymore about those things than you need to know how to edit Windows’ registry to run Windows.

I use desktop Linux every day, and I’m a Linux expert. Do you know how often I turn to a terminal to get to a shell to run commands? Maybe once a month, if that.

Between the two major Linux desktop interfaces, KDE and GNOME, Linux has you covered. For applications, many of the most popular applications, such as Firefox and OpenOffice, run just the same on Linux as they do on Windows. For other end-user programs, Linux programs such as Evolution for e-mail and Pidgin for IM are just as good, if not better, than their Windows equivalents. And again, you don’t need to know anything special to use them.

Installing new software on Linux isn’t any trouble either. Better still, major Linux distributors like Ubuntu are continuing to make installing Linux software easier than ever with programs like Ubuntu Software Center.

Don’t get me wrong: if you’re running a Linux server, you really need to know Linux’s technical guts. But you know what? If you’re running a Windows server, you also need to know Window’s version of the shell, the PowerShell.

No matter what desktop operating system you’re running, if you really want control over exactly what it does, you need to know how to manage its command line tools. But for day-to-day use, Linux’s graphical interfaces makes it just as easy to use as Windows or Mac OS X. Pretending that you need to be some kind of computer wizard to run Linux on the desktop today is just downright silly.

At the OSCON open source convention in Portland last week, Neuros CEO Joe Born explained how Linux-based embedded devices will bring open source to the set-top market and the consumer electronics space. He also demonstrated how to build applications for the Neuros OSD, his company’s programmable DVR product.

We looked at Neuros last year when the company teamed up with TI to build an open source reference platform for TI’s DaVinci DSP hardware. Neuros is building a community around the platform and has sponsored bounties to encourage third-party developers to participate in the process.

In our recent post-mortem of the HTPC concept, we noted that specialized set-top boxes are rapidly ascending to a position of dominance in the living room. HTPC systems can’t compete with media devices that are smaller, less costly, and more energy-efficient. Neuros aims to offer a product and platform that can fill the set-top niche in an open and extensible way.

“The conventional view of convergence has largely been surrounding PCs,” said Born. “But if you take a step back, you see a lot of things that PC isn’t the answer for. Embedded devices still remain cheaper.”

The growing ubiquity of Internet-enabled always-on consumer electronics products will change the way that people use technology, Born contends. He believes this trend will create new opportunities for building an ecosystem of services and applications that are accessible to such devices. He also notes that the high level of fragmentation that currently plagues this market has created a space where open source software can achieve some traction. Software freedom will come to the masses through embedded devices, he says, not desktop computers.

The software stack

There is a lot more to the project than ideology, however. The Neuros software platform is pretty darn cool. It’s built on version 2.6.26 of the Linux kernel and uses VLC for its multimedia capabilities (more details about the VLC port are available at the Neuros open source blog). Neuros has legally licensed the codecs, so it can handle a variety of common proprietary formats. The middleware stack developed by Neuros is distributed under the GPL.

Support for third-party application and service integration is extensive. The platform offers a D-Bus interface for controlling media playback and has a full API for managing scheduled recording. Neuros applications can be developed with Trolltech’s Qt toolkit, which means that there is a clean glide-path for porting software from the desktop. During the presentation, Born showed how to use the Qt graphical interface designer on a laptop running Ubuntu to build a program, which he then cross-compiled and ran directly on the Neuros. The Neuros SDK provides a set of scripts that hide much of the complexity of cross-compilation, so the whole process was impressively seamless.

For developers who prefer a higher-level development path (I personally prefer to avoid C++ whenever possible), the Neuros platform will support scripting in several different ways. The current model of the OSD supports scripting with Lua, and the next-generation model will add Python with full Qt bindings. For an even lighter development solution, the platform provides a framework for writing simple applications with web technologies. This runs on top of an integrated WebKit HTML renderer which is also used for the Neuros browser.
Third-party stacks

It will also be possible to replace the default Neuros software stack with something completely different. Developers can use the Neuros OSD boxes as testbeds for building new platforms. Born hopes that users can eventually choose from a wide range of third-party stacks that offer a multitude of different capabilities.

For instance, it could eventually be possible to port other preintegrated media center solutions such as MythTV or XBMC so that they can run on the device. Nokia’s recently-sponsored Ubuntu ARM port is another potential candidate for clever hackers who want to repurpose the OSD. We have seen a lot of awesome stuff ported to ARM devices by members of the great Internet Tablet community (such as Penguinbait), so it’s likely we will see lots of action around the OSD, too.

There is already a very real community forming around the device and it has some nice features, such as Last.fm integration and the Neuros web browser, that have been implemented by volunteer contributors and Google Summer of Code participants.

I’m enthusiastic about the whole Neuros concept. I own the current model of the OSD and I’m really looking forward to the release of the next-generation product, which supports 720p and is housed in a mini-itx case. Born says that the 2.0 prototype units will be available for LinuxWorld, which is coming up next month.

July 8 (Bloomberg) — Google Inc., owner of the most- visited Internet search engine, is developing a computer operating system based on its Chrome Web browser, taking aim at Microsoft Corp. in its strongest market.

The system will be designed at first for low-cost laptops called netbooks, Google said in a blog post. The company is in talks with partners on the project and computers running the software will be available in the second half of 2010.

The plan escalates Google’s rivalry with Microsoft, which extends to Web search, browsers and business applications such as word processing and spreadsheets. Windows, Microsoft’s flagship product, runs about 90 percent of the world’s personal computers. Google is also trying to spur Web-ad sales after reporting its first sequential revenue drop as a public company.

“There is a possibility that the new OS can break the paradigm Microsoft and Intel created over the past 20 years,” said Yukihiko Shimada, a computer analyst at Mitsubishi UFJ Securities Co. in Tokyo. “There is plenty of business opportunity for Google in this market.”

Google said it’s working with computer makers to introduce a number of netbooks next year, without identifying any of the companies. The Chrome OS will be open-source, meaning the program code will be open to developers, Google said. The software will work on top of the Linux operating system.

Netbook Competition

Frank Shaw, a spokesman for Redmond, Washington-based Microsoft, declined to comment. Windows accounted for 28 percent of the company’s $60.4 billion annual revenue in the 12 months ended June 30, 2008.

Microsoft has stepped up its efforts in the netbook market. It said in May it plans to remove a restriction of running three applications at a time on its forthcoming Windows 7 Starter Edition, which is designed for netbooks. The announcement eliminated one of the most significant differences between the basic edition of the operating system and a pricier one.

Google, based in Mountain View, California, rose $2.61 to $399.24 in Nasdaq Stock Market trading at 9:32 a.m. New York time. Microsoft fell 22 cents $22.31. Before today, Google had risen 29 percent this year and Microsoft had added 16 percent.

Online Strategy

The Chrome OS is consistent with Google’s focus on getting people to use software online, which contrasts with Microsoft’s approach of providing programs on the computer itself. Google started offering business software in 2007, allowing users to access spreadsheets and word-processing documents via the Web, just as anyone might access the search engine or Google News.

“We hear a lot from our users and their message is clear — computers need to get better,” Google said. “The operating systems that browsers run on were designed in an era where there was no Web.”

Google is trying to encourage people to spend more time online to fuel demand for Internet ads, which accounted for more than 90 percent of its 2008 revenue of $21.8 billion. In the first quarter of this year, Google had its first sequential sales drop since it went public in 2004 as the recession prompted companies to curb advertising spending.

The open-source Chrome OS will probably win over companies that don’t want to pay for Windows, said Jim Friedland, an analyst at Cowen & Co. in New York. The system may also help Google sell Web-based applications, such as the Gmail e-mail service, that run on top of it, he said.

“They really haven’t cracked the enterprise yet,” said Friedland, who rates Google’s shares “outperform” and doesn’t own any. “We’ve seen some inroads around the edges.”

‘More Options’

Computer makers such as Acer Inc. and Asustek Computer Inc. already have plans to offer computers running Android, an open- source operating system backed by Google and initially designed for mobile phones. Acer, the world’s second-largest laptop maker, said last month it plans to release a low-cost notebook powered by Android. Asustek Computer has also developed a netbook that runs on Google’s software.

“Having another OS or another interface does create more options, and with the weight of the Google name behind it, does lift its prominence,” said Bryan Ma, a computer analyst at IDC in Singapore.

Google said that while the Chrome OS is separate from Android, the two will overlap in some areas. The Chrome operating system is designed to save users from having to deal with viruses and security updates, Google said.

“Google Chrome OS is being created for people who spend most of their time on the Web, and is being designed to power computers ranging from small netbooks to full-size desktop systems,” Google said. “While there are areas where Google Chrome OS and Android overlap, we believe choice will drive innovation for the benefit of everyone, including Google.”

‘Beneficial to Users’

Tony Chen, chief operating officer of Asustek’s notebook unit said by phone the company will consider “anything that’s beneficial to users.” Fujitsu Ltd. spokeswoman Nozomi Endo said the company will monitor market conditions before deciding whether to introduce products using Google’s operating system.

Faith Brewitt, a Dell Inc. spokeswoman, and Hewlett-Packard Co. spokeswoman Liana Teo didn’t answer calls to their Singapore offices. Spokespeople for Acer, Sony Corp., Samsung Electronics Co., NEC Corp., Panasonic Corp., and Toshiba Corp., declined to comment.

The Chrome OS — which will run on traditional Intel Corp.- based x86 chips along with semiconductors designed by ARM Holdings Plc — will work on lightweight netbooks along with more powerful computers, including desktop PCs, Google said.

Google’s Chrome still faces an uphill battle against Microsoft’s browser. Chrome, which was unveiled last year, had 1.2 percent market share in February, compared with 67 percent for Microsoft‘s Internet Explorer, according to research firm Net Applications, which tracks Web statistics.

In May, Microsoft introduced a search engine called Bing that has enhanced shopping, travel and sorting features. Bing’s market share climbed to more than 10 percent in June, according to Comscore Inc.

Google’s search engine is No. 1 in the U.S., holding more than 60 percent market share. Microsoft is No. 3, according to ComScore.

Find the RHEL 5.2 kernel source

For re-building RHEL 5.2, you’ll want kernel-2.6.18-92.el5.src.rpm. It is on the source CD that comes with your CD set. You will not find it on your standard CDs.

To determine whether you have installed the source RPM, look under /usr/src/redhat/SPECS. If there is nothing there, then you probably have not installed the correct RPM yet.
Install the kernel-2.6.18-92.el5.src.rpm

The instructions are as follows:

rpm -i kernel-2.6.18-92.el5.src.rpm
(If you see lots of these warnings:

warning: user brewbuilder does not exist – using root
warning: group brewbuilder does not exist – using root

you can ignore them.)

cd /usr/src/redhat/SOURCES
ls *.patch | wc -w
1809
(this directory contains the base kernel source code and 1809 patch files)

cd /usr/src/redhat/SPECS
ls
kernel-2.6.spec
(this directory contains the kernel-2.6.spec file which specifies the
steps to build the kernel)

Build the kernel

You might have to install unifdef-1.171-5.fc6.ppc.rpm which is needed for the build.

The build instructions are as follows:

cd /usr/src/redhat/SPECS/

vi kernel-2.6.spec

(Search for…)
#
# Polite request for people who spin their own kernel rpms:
# please modify the “buildid” define in a way that identifies
# that the kernel isn’t the stock distribution kernel, for example,
# by setting the define to “.local” or “.bz123456″
#
#% define buildid

(Change the commented out line)
#% define buildid

(to something informative – in this case – the normal 64K kernel)
%define buildid .64K

<save and file>

cd /usr/src/redhat
rpmbuild -ba –target ppc64 SPECS/kernel-2.6.spec

(–target ppc64 is used to indicate that you want to build a ppc64 kernel)

(the rpmbuild process will take a while to complete…)

Note: During the compilation, gpg (GNU Privacy Guard) may want some system activity to generate security related keys, which causes the build process to pause.

* If this happens, you can open up another session and run “dd if=/dev/mapper/VolGroup00-LogVol00 of=/dev/null bs=1M count=500″ or anything that generate system activities. The paused build process should resume after the request from gpg is satisfied.

rpmbuild is the clever program that applies all of the patches to generate the overall kernel source, builds the kernels, modules and packages all the resulting files in an RPM for installation.

The rpmbuild step will result in these final messages:

Wrote: /usr/src/redhat/SRPMS/kernel-2.6.18-92.el5.64K.src.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-2.6.18-92.el5.64K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-devel-2.6.18-92.el5.64K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-headers-2.6.18-92.el5.64K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-debug-2.6.18-92.el5.64K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-debug-debuginfo-2.6.18-92.el5.64K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-debug-devel-2.6.18-92.el5.64K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-kdump-2.6.18-92.el5.64K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-kdump-devel-2.6.18-92.el5.64K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-debuginfo-common-2.6.18-92.el5.64K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-debuginfo-2.6.18-92.el5.64K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-kdump-debuginfo-2.6.18-92.el5.64K.ppc64.rpm

Install the newly built kernel

The instructions are as follows:

cd /usr/src/packages/RPMS/ppc64

rpm -i kernel-2.6.18-92.el5.64K.ppc64.rpm

These steps install all the pieces in the right place. The /etc/yaboot.conf file is updated as follows:

boot=/dev/sda1
init-message=Welcome to Red Hat Enterprise Linux Server!\nHit <TAB> for boot options

partition=2
timeout=80
install=/usr/lib/yaboot/yaboot
delay=5
enablecdboot
enableofboot
enablenetboot
nonvram
fstype=raw
default=2.6.18-92.el5.6

image=/vmlinuz-2.6.18-92.el5.64K
label=2.6.18-92.el5.6
read-only
initrd=/initrd-2.6.18-92.el5.64K.img
root=/dev/VolGroup00/LogVol00
append=”console=hvc0 rhgb quiet”

image=/vmlinuz-2.6.18-92.el5
label=linux
read-only
initrd=/initrd-2.6.18-92.el5.img
root=/dev/VolGroup00/LogVol00
append=”console=hvc0 rhgb quiet”

The variable “default” is set to “2.6.18-92.el5.6″ for the new kernel. We strongly recommend that you keep the boot entry for the RHEL 5.2 kernel around, in case the newly built kernel is having some boot issues.
Confirm the boot

Indeed you must confirm that it boots correctly.

You can verify the base page size by the following command:

getconf PAGESIZE
65536

Page size of 65536 bytes clearly indicates that we are running on 64KB pages.
Build a kernel with the 4KB support

After you have completed the above steps successfully, we have a good base for building other kernels. Let’s build a 4KB kernel.

Let’s follow the instructions below.

cd /usr/src/redhat/SPECS
vi kernel-2.6.spec

Change %define buildid .64K
To     %define buildid .4K
<save and file>

cd /usr/src/redhat/SOURCES

vi kernel-2.6.18-ppc64.config

Change CONFIG_PPC_64K_PAGES=y
To     CONFIG_PPC_64K_PAGES=n
<save and file>

Do the same to kernel-2.6.18-ppc64-debug.config and kernel-2.6.18-ppc64-kdump.config

cd /usr/src/redhat
rpmbuild -ba –target ppc64 –without kabichk SPECS/kernel-ppc64.spec

“–without kabichk” is used to disable ABI compatibility check. Without specifying it, the kernel cannot be built. And we assume that ABI compatibility is not crucial for what you are doing.

This should succeed. We will see the following at the end of the build.

Wrote: /usr/src/redhat/SRPMS/kernel-2.6.18-92.el5.4K.src.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-2.6.18-92.el5.4K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-devel-2.6.18-92.el5.4K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-headers-2.6.18-92.el5.4K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-debug-2.6.18-92.el5.4K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-debug-debuginfo-2.6.18-92.el5.4K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-debug-devel-2.6.18-92.el5.4K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-kdump-2.6.18-92.el5.4K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-kdump-devel-2.6.18-92.el5.4K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-debuginfo-common-2.6.18-92.el5.4K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-debuginfo-2.6.18-92.el5.4K.ppc64.rpm
Wrote: /usr/src/redhat/RPMS/ppc64/kernel-kdump-debuginfo-2.6.18-92.el5.4K.ppc64.rpm

Install the 4KB kernel

rpm -e kernel-2.6.18-92.el5.64K
(The 64K kernel might have conflicts with the 4K kernel.)

cd /usr/src/redhat/RPMS/ppc64
rpm -i ppc64/kernel-2.6.18-92.el5.4K.ppc64.rpm

The /etc/yaboot.conf will be modified accordingly.

boot=/dev/sda1
init-message=Welcome to Red Hat Enterprise Linux Server!\nHit <TAB> for boot options

partition=2
timeout=80
install=/usr/lib/yaboot/yaboot
delay=5
enablecdboot
enableofboot
enablenetboot
nonvram
fstype=raw
default=2.6.18-92.el5.4

image=/vmlinuz-2.6.18-92.el5.4K
label=2.6.18-92.el5.4
read-only
initrd=/initrd-2.6.18-92.el5.4K.img
root=/dev/VolGroup00/LogVol00
append=”console=hvc0 rhgb quiet”

image=/vmlinuz-2.6.18-92.el5
label=linux
read-only
initrd=/initrd-2.6.18-92.el5.img
root=/dev/VolGroup00/LogVol00
append=”console=hvc0 rhgb quiet”

And reboot.

You can verify the base page size by the following command:

getconf PAGESIZE
4096

The page size clearly indicates that we are on 4KB pages.
Apply a kernel patch

Let’s follow the instructions below.

cd /usr/src/redhat/SPECS
vi kernel-2.6.spec

Modify buildid as specified above.

Search for Patch99990 which is almost at the end of all patches.

Assume that the patch right before Patch99990 is Patch12345.

Add “Patch12346: abc.patch” where abc.patch is your patch.

Search for %patch12345 -p1.

Add %patch12346 -p1.

<save and file>

Add abc.patch to /usr/src/redhat/SOURCES.

Run rpmbuild as specified above.

Using the re-built RPM on other RHEL 5.2 systems

It is indeed desirable to install the built RPM on other RHEL 5.2 systems, but you need to be very careful.

Note that RHEL 5.2 systems are different in many ways, e.g., the setup of the volume manager.

Once the RPM is installed, follow the instructions below.

cd /boot

mkinitrd initrd-2.6.18-92.el5.64K.img 2.6.18-92.el5.64K
(Generate a fresh initrd image for booting purposes.)

And then reboot.
In summary

This quick report takes you through the practical steps of rebuilding a RHEL 5.2 kernel for Power systems. This exercise is intended simply as a way to see how the RHEL 5.2 kernel is built, installed, and booted to.

Because of the rise of small ‘netbook’ laptops like the Asus Eee, which don’t come with any kind of CD/DVD drive, it’s quite handy to have good, clear instructions on installing Linux from a USB stick. Even if you do have an optical drive, why bother burning a CD every time? It’s so wasteful.

So I worked out a really easy way to transfer the contents of ther Ubuntu LiveCD to my USB stick and set it bootable, and I thought I’d document the process here in case it can help anyone else.

This method also works with Edubuntu, Kubuntu and Xubuntu, as well as most other Ubuntu-based distros and even some other distributions too (basically, as long as the CD uses isolinux as the bootloader, which 99% of them do). It won’t hurt to give it a try, and I’m happy to help anyone out who wants to give it a go.

These instructions assume that you’re using a Windows machine to prepare the USB flash drive, as there are plenty of guides out there on how to do this from within Linux itself. Also, the latest version of Ubuntu comes with a USB stick creation tool right on the LiveCD, so if you have another machine with a CD drive you can boot into that and give it a go.

-

Preparing the files and USB flash drive

* First up, we need to format the USB stick in Windows – go into My Computer, right-click on the USB drive, and then click Format. When prompted, select FAT32 as the file system. Don’t use NTFS – it won’t work!
* Download and install 7-zip. It’s free and extremely useful, for this and anything else you might need to do with compressed files. It’s also open source, just like Linux, so that has to be a good thing. Get the latest stable version to be on the safe side – not the beta.
* Download Syslinux, which is the bootloader we will be using to make the stick bootable. Just get the latest version in zip format, and extract it to the Desktop – right-click, select 7-zip, then extract to syslinux-7.1.1 (or whatever) and then rename that folder to “syslinux”.
* Open a command prompt (on the Start Menu under Accessories, or run “cmd”) and cd to the syslinux folder:
* cd Desktop/syslinux/win32
* Run this command to install the syslinux bootlader to the thumbdrive:
* syslinux -ma f: (where f: is the letter of your USB thumbstick)
* Now we need to extract the Ubuntu ISO – Right click on the ISO file, and select 7-zip/extract to ubuntu-8.04…

The preparation is done! You should now have syslinux installed on your USB thumb drive, and a folder called Ubuntu-8.04-somethingorother on your Windows desktop. Now it’s time ot create our bootable Ubuntu LiveUSB Stick!

-

Putting it all together

* Drag all of the files you just extracted from the ISO into the empty USB drive.
* Move everything from the isolinux folder into the root of the drive. So, if your USB thumbdrive’s letter is F, as in our earlier example, move all the files from F:\isolinux\ into F:\
* Rename the isolinux.cfg file to syslinux.cfg
* Eject the drive, insert into the computer you wish to install Ubuntu onto, and boot! Shazam!

-

You may need to configure your BIOS to boot from USB, or at least go in there and change the boot order to try external drives first. It’s quite a simple process, but varies depending on your motherboard manufacturer, and so I can’t really go into it here.

Most modern computers will let you press F12 during the very first boot screen to select a temporary boot device – you should only need this once so that could be sufficient.

-

[Update]

This has now been tested with Zenwalk and gOS (see my gOS Gadgets introduction here) and works flawlessly. However, for an even easier method I can highly recommend UNetbootin, which fully automates the process under Linux or Windows. If you are using the method above and running into problems, why not give it a try!

Also, for some reason this doesn’t seem to work with Ubuntu Server. I am looking into it!

The main feature of gOS 3 Gadgets is its ability to instantly launch Google Gadgets for Linux on start-up, allowing users access to more than 100,000 iGoogle and Google Gadgets applications. These applications, though graphically rich, are small enough to be added to the computer in seconds over an Internet broadband connection. The new operation system will also be loaded with WINE 1.0, Lightweight X Desktop Environment (LXDE), and other Google software for Linux to improve the user experience.

While WINE has been known to allows users to use thousands of Windows applications on Linux platform, LXDE is a renewed effort to develop more lightweight desktop applications for Linux environment. By supporting LXDE, gOS 3 Gadgets, apart from desktops, would also make a good choice for ultra small mobile laptops, which are generally suffer from having slow hard drives and processors.

In addition to Google Gadgets for Linux, gOS 3 Gadgets can also run other Google applications more well-known in Windows platform including Google Desktop, Google Picasa, Google Earth and Google Maps. In the new gOS 3 Gadgets, other Google’s web-based applications such as Documents, Calendar, and Mail launch have a closer appearance and functionality to desktop applications than other platforms.

Version 2.6.26 was released on Sunday, July 13. What’s most interesting is the inclusion of a kernel debugger, Linux contributor Jonathan Corbet writes. Torvalds declined to include a debugger for years but finally caved to the wishes of developers, Corbet maintains.

Additional KVM support is the other key thing about 2.6.26. The update offers ports of KVM to the IA64, X390 mainframe and PPC architectures and basic paravirtualization support. KVM, an open source virtualization hypervisor, was introduced in the 2.6.20 kernel.

The update also offers support for PCI Express Active State Power Management, early support for the forthcoming 802.11s wireless mesh standard, enhanced webcam support provided by UVC drivers and a built-in memory tester.

The kernel also provides features for developers including support for read-only bind mounts, Lx86 Page Attribute Table and BDI statistics. The kernel includes a lot more features and enhancements that are thoroughly listed and explained here on the kernel newbies site.

Every year or so, there’s a new Linux distro darling. Last year’s was Ubuntu, and I’ve lost track of what this year’s favorite penguin might be. But a lot of newcomers to Linux seem to get the impression that a new distro is better than an older package because there’ll be more attention paid to new technologies and easier user interfaces. In fact, pretty much the opposite is true.

For example, I’m a Fedora moonie. I don’t hate any of the others, but Fedora got me way back in Version 3, and now that we’re at 9 it’s just getting better. I use only Linux for certain tasks, and Fedora does them all with aplomb, with almost no maintenance on my part–no reason for me to change to some new distro. A more mature distro means the project team has gotten around not just to incorporating new technologies, but also to paying attention to details. Fedora 9’s new NetworkManager, for instance, really makes Fedora on a laptop easier to manage.

Another old distro that many new users simply ignore because it’s got too much of a geek stigma is Debian. This is one of the original distros, first pubbed back in the early ’90s, with a long history of development. And in fact, many of the new distro darling packages are based on Debian, including Ubuntu. Debian’s maturity, however, has several advantages:

A much smoother installation. Debian has long history of hardware support, including drivers for mainstream manufacturers like ATI, Broadcom and Nvidia.

The largest library of compatible software repositories. We’re talking over 18,000, which is bigger than anything else, even my Fedora.

Officially supported software. So not just 18,000 packages, but just about that many that are officially supported by the Debian team–that means compatibility and performance testing as well as maintenance updates.

Slick details. I’m most impressed with Debian’s installer, which lets you not only find compatible software, but automatically lets users decide how much risk they’re willing to take by choosing the minimum maturity level of installed software, ranging from ‘experimental,’ to ‘unstable,’ ‘in-testing,’ or finally ’stable.’ Cool thing is, most Debian users will tell you that ‘unstable’ is stable enough and choosing this option gives you access to almost the entire library of 18,000 software packages.

I’m still not a big believer in Linux as a mainstream desktop player–a desktop for everyone. But it works great for many, especially IT folks, programmers and serious power users. But if you’re looking to give the penguin a try, don’t immediately go for the latest distro to hit the streets. Check out some of the old timers, too. You might be pleasantly surprised.

Best Buy Ubuntu 

Synopsis

You’re right in the middle of an important procedure when your computer freezes and crashes, erasing your data and costing you hours of extra work. For the thousandth time, you wish you had an easy-to-use alternative to your current operating system. Look no further than Ubuntu Linux, a community-developed, Linux-based operating system designed to give new life to your old PC or Mac.