Gaël Duval's own Mandriva Linux related FAQ

About Me

Backgrounder: I'm 32 and counting, European, living in Normandy, France. Most of my time is devoted to Mandriva Linux as initial founder. I've spent the latest seven years to put much energy in Mandriva Linux to make it known as widely as possible, and a real alternative to proprietary operating-systems.

I recently switch from the Communication department to a newly-created departement which is going to focus on Community actions and connexions.

Never been to space yet. Will try to improve my chinese connexions to become a Taïkonaut.

* Of course, some of you will notice that this own-interview is a wink to the recent Mark Shuttleworth FAQ *

FAQs: Why and Whither for Mandriva Linux?

I want to talk a bit about Mandriva Linux, why it started, what's its purpose. I also want to describe how Mandriva can be a real Free Software project and can support Free Software while being a business, thus paying many people to work on it.

Of course this piece of writing is purely from my own perspective, and his written in MyOwnEnglish(Copylefted) language. That's why it's published on my personal website, and I apologize to readers who are going to find typos and grammar and as a result become upset.

*Why did I start Mandrake^H^H^H^H^H^H^H^HMandriva?*

First of all, I think that Free Software is the future of computing. If you look at competition in the software area, design costs and product price factors make Free Software is natural challenger to propriary software. So for me it was more natural to launch a Free Software project instead of starting a proprietary software business.

Back in 1998, Linux wasn't easy to use at all unless you had significant UNIX background. Installing a new hardware meant you had to recompile the Linux kernel with correct option. Using a cd-rom or a floppy drives meant you had to type, for instance, "mount -t iso9660 /dev/hdc /mnt/cdrom" on the command line. Which is not really an obvious task.

The desktop environment at this time, was also more of a "decoration environment". It was more a way to justify the use of X-Window, than anything else :)

So I started Mandrake Linux, alone against everybody ;) At home. Later, when Mandrakesoft was created, it's been clear that the focus of the company would be to deliver easy-to-use Linux products, while supporting Free Software. And it didn't change.

On the other hand, I must confess that Mandriva Linux's initial design was to focus on the desktop only. I thought that it was better to offer a "specialized operating system" - on the desktop - by removing most server features and offering only one desktop option. I think that this was a real mistake to think like that.

Why the hell should we limit operating system features or specialize it? In my opiniont, this is a "cosmetic thought" for minds who tend to like "purity of things" over "boiling messes". And it's also a reminiscence of the proprietary software industry's way of thinking which tends to impose unique standards.

But Free Software is about diversity, and can also fits all needs. FOSS OSes have no limits in functionality, and I'm convinced that this aspect is a real force over traditional/proprietary OSes.

As a result I'm extremely happy to see that Mandriva Linux as turned to be capable of being used as a fast and efficient desktop OS or as a solid server OS, or even both at the same time. Users have different needs, users have to choose, not the OS maker.

*Will Mandriva ever demand licence fees or royalties?*

Absolutely not. We do Free Software. All the code created by Mandriva is published under the GPL. We don't do proprietary software.

But people will tell me: "you put proprietary software in your commercial packs". That's true.

For instance we included Netscape for a long time until a Free Software alternative - Mozilla - was good enough as a web browser to throw it away. And many people want their XYZ proprietary drivers or Z plugin, because there is not yet a good alternative in Free Software.

We put a few proprietary components in commercial packs, but not in the download tree which consists of more than 12,000 packages.

As a result we support Free Software while providing people a choice.

*How can Mandriva Linux be sustainable?*

Mandriva Linux employs more than 130 people to work full time. Half of them are Linux engineers who can devote 100% of their time to Free Software. I think - or at least I hope so - that they really feel better working on Free Software at Mandriva than working for any other grey and boring proprietary software company.

It means that we have to get income to pay all these people! How could it happen with a Free Software product? It's not easy, but it's not *so* difficult.

First of all, we have developed many services which provide a real added-value to our users. Typically this is the Mandriva Linux User Club, but there is also Mandriva Online, and others to come. Furthermore, we sell packs: physical packs, and now virtual packs for those who have fast Internet connexions. We don't sell the software inside, we send a service: we provide all the software, ready to install, plus support, plus access to online services, plus printed documentation. While people can download the software for free, many of them like to buy Mandriva packs because they bring them real benefits.

On the other hand, we do some business with large corporates and governement organizations, which is now a big part of our revenues. We don't sell them Free Software, we sell them our expertize in Free Software, we sell them full solutions, plus support and so on. Additionally, OEM sales with partners such as HP, Dell and others is also a growing source of revenue.

Last but not least, a significant part of Mandriva Linux users understand that supporting Mandriva Linux by regularly purchasing our products is an excellent way to see the product improve at each release. This is not the traditionnal way of thinking, but it's a way of thinking.

As a result, we have built an economical ecosystem which provides income to Mandriva to pay people to write, integrate and release Free Software.

*What about binary compatibility between distributions?*

A lot has been said about the fact that Mandriva Linux is not binary-compatible with Debian, Ubuntu, Red Hat and others. Most of the time, this manifests itself as "I can't install Ubuntu package on Mandriva", sometimes it's more "Why does Mandriva use GCC 4 when Debian is using GCC 3.3?". Or "Why is the kernel and glibc on Mandriva Linux different from that in Debian Sarge?".

First of all I think that "binary compatibility" should really be the ultimate goal for all Linux distributions. While source compatibility is possible between most of Linux distributions and tends to reduce the risk of "UNIX balkanisation syndrom", we need to respect the user's perspective and the ISV perspective.

Users would really prefer a world where they can install a software package designed for X, or Y Linux distribution on their Z distribution. And ISV are not comfortable to have 5 to 10 Linux distributions to support: they prefer to support only one - but which one? Or they'd prefer being able to support all distribtions at the same time.

Worse: speaking of source-code compatibility instead of binary compatibility is not a realistic way of thinking. Free Software source code relies on libraries: for instance Konqueror and Firefox use each one more than 40 different librairies to run. And these librairies are constantly evolving, including their features and their API. Consequently, it's often not easy at all to simply recompile an application to another Linux distribution ("configure, make, make install...") when it was built on the top of different, and sometimes non-common, libs versions.

As a result, I think that the only way to go for Linux distributions is to standardize around common versions of software and libraries to ensure 1) compatiblity between Linux distributions at source level 2) therefore the a possible binary compatibility which is the ultimate goal to reach.

There is a well-knonw organization which focuses on that, supported by all the major Linux distributions and major hardware vendors. This is the Linux Standard Base (LSB).

In order to summarize my point of vue on this topic: compatibility at source level is a necessary step, but not a sufficient step, to reach binary compatibility (at a given time). And only a common organization such as the LSB can improve compatibility at the source level, by defining a common set of libs and a common system structure (file hierarchy).

*But, I heard that Mandriva Linux is LESS compatible than other similar projects?*

If you've heard this specific allegation, it's absolutely not true.

Mandriva Linux, like other Linux major distributions, tends to rely on the same patches and same recent libs at a given time, and adopt the same system structure or similar, as other major Linux vendors. As a result, you will likely find excellent source compatibility and even good binary compatibility between Red Hat/Fedora and Mandriva for instance. Of course, this is only yet "accidental compatibility", but when it happens to be between LSB-certified distributions.

*What about the GCC 4.0 transition? Why did you adopt GCC 4.0?*

We always try to include the latest stable development tools, libraries, and applications. GCC 4.0 was released as part of Mandriva Linux 2006.

*Why is the default desktop in Mandriva Linux blue?*

Because blue is the color of her eyes as well as the color of oceans and sky.

Blue is often the color of children's rooms because it cools them. Blue certainly has a connexion with human minds.

*Will blue always be the default desktop colour?*

While brown is currently a fashionable color on the desktop, blue is for sure part of Mandriva's color scheme since early days. Anyway, some of you have noticed that the Discovery version of Mandriva Linux provides an island picture as default background.

Last but not least, we provide a feature which lets users actually *change* the color of their desktop!

*What's next for Mandriva?*

The future of Mandriva is exciting because we never had as many opportunities as we currently have!

The company is developing its business with big industry partners, extending its presence in the corporate world. And this gives us means to improve the product, and extend its functionalities.

On the other hand, we definately want to extend our involvement in the Linux community, and this is now my new mission at Mandriva.

Gaël, October 19th 2005 .

Back to my homepage

Disclaimer: Everything on these pages is only personal and cannot engage MandrakeSoft/Mandriva's responsability in any way.