Finally, after testing, using and experimenting with dozens of GNU/Linux distributions during the last 4 years, I’ve found my ultimate distro of choice. It’s ArchLinux. I’ve never been so happy to be a GNU/Linux user as today.

ArchLinux, is exactly what I was looking for. I’m strong believer in the concept of rolling updates because I think it’s the best way for GNU/Linux to take on in the Desktop usage.

Out of many GNU/Linux Distro I’ve tried, only two caught my interest. ArchLinux followed by PCLinuxOS. The only difference between them is that PCLinuxOS is more easy to install while ArchLinux is more easy to maintain. As a result ArchLinux has more recent packages than PCLinuxOS.

Another great thing about ArchLinux is the quality of its documentation. It’s excellent, everything is detailed and simple to follow.

And I’ve never seen any full-fledged GNU/Linux distro running so fast as ArchLinux. That’s because it’s not bloated with useless apps, the user chooses what to install and what not, and because its packages are optimized for i686.

ArchLinux Community plays an important role in providing help and maintaining AUR packages. I can confidently say that ArchLinux Community is a remarkable and vibrant one.

So far, I’ve had only one issue with it. My DVD recorder couldn’t burn above 2x which was due SATA Kernel module. It’s needless to say that not only I was able to fix the issue in a few moments but the way ArchLinux reconfigured the Kernel was lightning fast.

All I can say is Kudos to ArchLinux members and to ArchLinux Community. Really, a Big Thank You to everyone.

Now, I can focus on doing my work instead of fighting with my OS.

throughout my experience with Debian Sid (nearly a year now), at least I had three show-stoppers. One was a serious package management conflict that required manual hacks to solve. The second one was a complex Xorg breakage that required 2 hours of work to fix. and the last one was yesterday when I was unable to bootup the new installed kernel 2.6.26-1-686

Here is nice screenshot for those who have never seen Linux kernel crash. It’s rare to happen. In 4 years, I’ve only seen it once and it was due to hardware failure.

I’m not going to be defensive, but the Kernel crash you are going to see has nothing to do with the quality of the Kernel itself but it shows what can happen when distro maintainers start to mess with things 🙂

It seems that I wasn’t the only one complaining about Ubuntu’s specific bugs. But the road to Debian Lenny is not that rosy too.

Although Debian Lenny get fairly new and stable packages it suffers from incoherence in the way packages are pushed into Debian repositories. I’ll explain :

It seems Debian Lenny Packages are released under the policy “if it’s fairly stable then push into the repositories.” In theory, this sounds good but in practice it’s not. Let’s say that Gnome 2.20 is released today, and after 2 months from today it will become approved to show up in Debian Lenny repository. However, instead of pushing all Gnome 2.20 packages and the ones that depend on it in one shot, these packages get released gradually which will make your system having a mixture of the old Gnome packages (2.18) and the new ones (2.20) thus causing some of your installed softwares not to work or crash because some packages (the ones they depend on) are not yet released.

I have taken some time to try to figure out what to do in this case. Well, I think I have found a way :

It’s creating my own local Debian Repository. The concept is to move to Debian Sid (experimental = the most recently introduced packages with some risks involved) instead of Debian Lenny (Testing = fairly new, but may suffer from incoherence).

Then create a snapshot of the entire Debian Sid Repository. This will allow me to install any package later on without having to resolve any dependency issues. Then use it for 3 months or so, then re-snapshot it again and start using the latest and greatest things in the Free andOpen Source software world.

This is my plan for now, although I’m still evaluating other alternatives Like Sidux and Mint Linux. Probably CentOS and Fedora Core are the next interesting thing to use.

Sharing The Power

November 28, 2007

As human being, it seams it’s our nature to disobey, to break free from the crowd, to want be free and to gain control over things.

To control, you need to communicate; and communication is all about passing and getting information.

While information is power, Its power comes from the way it can be presented or manipulated. For instance, you can manipulate information you gained to reach your goal (your interest), or counter it with another information (to protect your interest).

The most effective way of manipulating information is to use it to stimulate fear. Starting from the fear of losing power, up till the fear of losing life.

So, how to keep that power to yourself? well, hide it. Keep it secret. Prevent others from getting it or knowing about it.

But the real question is do we accomplish anything valuable by keeping that power to ourselves alone? Well, if you’re self-centric and think you’re smart you will accomplish many things by being so; however, the real accomplishment is not getting value from others but giving value to others.

I think the whole idea about Free(dom) Software is about sharing that power instead of keeping it in the hands of privileged. Thanks to Free Software, information is equally distributed, and everyone get the same opportunities, the same benefits, the same advantages and more importantly is that everyone one can participate and make it more valuable.

Well I’ve come to the conclusion that I need to move to Debian Testing codenamed Lenny. it’s the next coming release of Debian currently being under development; and once it’s marked as stable it will become the default version like the actual one, Debian Etch (Debian 4.0).

Why am I moving away from Ubuntu ?

Back in 2005 (my first days with GNU/Linux), I have set a goal to move to Debian when I become ready. First, I tried many distributions but I settled with Ubuntu because it was newly introduced to the market with the incentive of offering the same software quality for both enterprise and community edition while focusing on making GNU/Linux as easy and user friend as possible for newcomers. I was attracted to that principle and to the fact that it was based on Debian which is well known by its stability and its popularity among system administrators.

Since Ubuntu is based on Debian Experimental (Sid), it offered the latest and greatest versions of softwares. Also, it’s means you are on the bleeding edge every six months which is wonderful. However, this bleeding edge has a hidden but justified cost. It’s some bugs that while they can push you to learn more about GNU/Linux trying to figure out what causing them, they may also be the source of serious headaches. Personally, most of my headaches come from FireFox performance, X11 Drivers’ stability, and Linux Kernel issues with some DSL Modems and most of these problems are Ubuntu specific.

Right now I no longer has the capacity to haunt these issues, I want to be using something that’s new and fairly tested. I think, here where comes the advantage of Debian Testing over Debian Experimental or Ubuntu. You sacrifice some of the coolness of using the latest softwares in order to be using fairly new but more tested softwares.

I want a platform that is always evolving, stable and reliable and where I can experiment other things on.

Microsoft latest deals (enveloped as “collaboration agreements”) with some commercial GNU/Linux distributions pushed me to wonder what is the real deal/catch for Microsoft ?

Well, it’s not an easy thing to guess but from closely following the news of such deals I come to these completely speculative ideas. [Update: now they became true and valid.]

Microsoft pays more money than they get from these commercial GNU/Linux vendors in order to

1) Allow these commercial GNU/Linux vendors to buy open source startups. (i.e Xandros acquired scalix) [Update: July 1st, 2008, Xandros acquired Linspire too]

2) Then Microsoft opens their proprietary protocols exclusively to these vendors and thus given them advantage over non-commercial GNU/Linux Distributions. (i.e Microsoft offers specifications and documentations to Xandros about MS Exchange protocols)

3) By doing this, Microsoft preserve its dominance since these commercial GNU/Linux distributions would work efficiently with Microsoft proprietary protocols. This will be beneficial to Microsoft in two ways: 1) Slow down full platform migrations. 2) This gives more time to Microsoft to formulate more effective strategies to deal with this open source trend/threat. [Update: Feb. 21, 2008, Microsoft’s Strategic Changes in Technology and Business Practices to Expand Interoperability announcement ]

4) These deals “may” work in Microsoft advantage in case of later legal actions probably against competing free (Debian, Ubuntu) and/or commercial (Red Hat) GNU/Linux distributions.

However, I don’t perceive any imminent threat from Microsoft as it still trying to figure out what are the driving motives for individuals & SMBs to consider GNU/Linux.

Bravo Ubuntu Team !!

April 22, 2007

Congratulations to all Ubuntu Team and GNU/Linux Community.

You have done a great work in Ubuntu 7.04 Feisty Fawn.