Interview with Mandriva (ex Mandrake) Linux Founder Gael Duval
Editors Note: As LQ interviewed Gael in 2003, a couple of the intro type questions were skipped. See the other interview for those questions.
LQ) It's been about a year since the company exited 'redressement judiciaire' and things seem to be going quite well now. To what do you credit the turnaround? What lessons have you learned along the way?
GD) We started to clean things out in 2001 because we felt that the situation started to become dangerous. But when you take actions to save money and make more business, it takes time, and actions can only be taken progressively.
For instance, starting from 2001 we continuously reduced expenses and developed new income sources such as Club and Store which are now our main sources of revenues. But we reached the break-even point only in spring 2003, approximately at the same date as the beginning of the chapter-11 like period.
LQ) With the merger of Conectiva and Mandrakesoft now complete and the new company named Mandriva, in what direction will the company be headed? Have you run into any integration problems with the two companies having such a disparate geographic focus? Has the difference in primary language been a barrier?
GD) Of course, several issues have to be addressed such as the language barrier, but this one is not a showstopper since most of us are already used to speaking English in the IT industry!
Regarding the geographic focuses, Conectiva will boost our research & development forces, but on the business side, they will start to concentrate on what they know: selling Linux & Open-Source solutions to the Brazilian government and local corporates. Progressively we will introduce additional Mandriva services such as the Club to former Conectiva users.
LQ) What has the response to the new name been like? What others names were being considered?
GD) Of course, many different names have been considered. Each of them had its qualities and problems. A key issue was the availability of the related trademarks & Internet domain names. Also we wanted a name that would remind both Mandrake & Conectiva, and which would sound simple and efficient. As a result, "Mandriva" was chosen.
There was a controversial discussion about how English speakers would pronounce "Mandriva", but definitely, "Mandriva" has its own sound as well: it's "Man-dree-vah" (or "Man-dree-vuh").
LQ) Are there any plans to make a larger push into the US market, or will the focus stay on Europe and South America? There has been a recent Linux push into the "emerging" markets (China, India, etc.) - is this a direction the company is looking into?
GD) Mandriva traditionally had a large userbase in the US, since we mostly started there with MacMillan/Pearson in 1999!
We currently have a partnership with O'Reilly who is distributing versions of Mandriva Linux bundled with books in many US stores.
We also have a new program for SMBs in the USA.
The US market is definitely a market we will address more and more, although for retail, it's clear that electronic distribution is gaining more and more momentum, with the development of broadband Internet access.
LQ) There have been rumors that some Linux distributors, including Novell, may follow what Red Hat has done and have an Enterprise release and a "Community" release. Is this direction something that has been considered by Mandriva?
GD) No. Mandriva Linux will still be distributed as both a download edition and commercial offers, with full official support for updates (bugfixes, security).
LQ) Conectiva had a lot to do with the porting of apt to RPM-based distros, while Mandrake has historically used urpmi. What tool is planned for future Mandriva releases?
GD) It's likely that we will have a package-management tool which is going to combine the best features of URPMI & the best from Smart, for RPM packages.
LQ) A couple years back, Mandriva (then Mandrakelinux) went from being KDE-centric to being much more neutral between GNOME and KDE. What prompted this change? Looking back, is it a change that you're glad was made?
GD) It was in year 2000 actually :) Yes we're very glad to have made this change because GNOME & KDE are both very popular desktop environments, so it was natural to provide both to our users.
LQ) Any update on the Linux Core Consortium? Have Red Hat or Novell made a final decision on whether to join?
GD) There is nothing new on the LCC front, we are working on it though. I don't think that RH or Novell would consider joining before they see first results of the implementation.
LQ) Limited Edition 2005 is being considered a "transitional version" for the company. How is that transition going? How has customer and community response to the new roadmap been?
GD) The transition is going well, and the Limited Edition 2005 is actually very appreciated by users. And the number of available packages for LE2005 is becoming impressing: more than 10,000! I feel that it's going to be a reference version of Mandriva Linux for a long time actually! (besides the fact that it's the first with the new name...)
LQ) I've seen you say recently that Linux was ready for the business desktop, but not for the home desktop. What's still lacking and when do you think we'll be ready?
GD) OK, I can try to explain my point of view: "is Linux ready for the home desktop?" The question is ambiguous actually!
Literally, the question itself means: "is the Linux system (technically) capable of handling home applications and work?". Obviously, the answer is yes! Linux is a very advanced OS, it's easy to install, it supports most of existing hardwares, it provides very neat desktop environments etc. Everybody can easily switch from Windows to Linux, it's not like switching from Windows to a mainframe system for instance, or to DOS...
Now, I also have my own interpretation of the question "is Linux ready for the home desktop?" ... For me it means: "Can Windows users massively adopt Linux instead of Windows (and stop using Windows)?". Obviously, the answer is no!
People who claim that Linux is 100% ready to replace Windows just don't know what they are talking about, or are just liars :) I've been converting people to Linux for years, and even now, I wouldn't recommend it to everybody for a simple reason: not enough common commercial applications or FOSS equivalents are available yet.
Let's take a simple example: do you think a graphics designer could use Linux exclusively for common design work? I don't think so. Take the example of most popular icons in Linux, most of them have been designed under... Adobe Illustrator! It's possible to find similar examples in several areas, and there is also the games issue.
As a result, I think that Linux is ready to replace Windows if you have a "basic" desktop use of it, for instance for Office/Internet and Multimedia.
Anyway, I think that we are getting closer and closer to the "ignition point". It's certainly only a matter of a few software publishers who will start to release their applications to Linux, and for instance the recent release of Acrobat Reader 7.0 by Adobe is an excellent sign of hope.
The best we can do to make this adoption go faster is to grow the Linux userbase so it becomes a market that traditional software can't ignore anymore.
LQ) What is your take on the fact that Munich selected a "non-commercial" distro (Debian) for their high profile Linux rollout?
GD) It's certainly an excellent idea that they selected Linux! Using Linux X, Y or Z is certainly not very important though. I just wish them they found motivated volunteers in the Debian community to help them :)
LQ) Thanks again for taking the time to do this interview. Any final comments?
GD) Thanks to you!
If you have any follow up questions, please post them here and I will forward them. I'd like to thank Gael for taking time out of his busy schedule to chat with LQ (again).
Good job, very interesting!
Follow up question.
When is Mandriva going to get serious about the Artistic appeal of the Distribution. Sure, Mandriva is easy enough for a small child to use, but that doesn't mean it has to look like a childs desktop. Scary looking animals with stars in their eyes, washed out blue backgrounds; come on, when is Mandriva going to get some of those Connectiva graphic artist to give the distro a fresh modern look. Sure beauty is in the eye of the beholder, but I've yet to hear anyone describe the LE as remotely artistically appealling. When is Mandriva going to Follow Redhat's lead and design it's own set of Icons, a professional looking set of desktop backgrounds, and a new window decoration to match both. How about Mandriva just adopt some of the Modern looking Open source icon sets Like Minimum, or H30 and build around those. I know that artist Like Sascha Hohne (http://www.rad-e8.com) and Milosz Wlazlo (http://www.miloszwl.com/index.html) would be more than happy to give the Mandriva desktop the type of attention it needs to be taken seriously for both its value and artistic appeal. It's time to put a new look on the desktop along with the new name.
you do realize that you can do all of this yourself. You dont need to sit in a forum and wait to get invited to do somthing like creating new eye-candy regardless of distro or variant of unix.
If it involves waiting for someone else to fix your problem, even if I do agree that mandrake makes for one fugly visual effect, you have nothing to complain about. Don't like it? Change it.
Now this is what I love about forums. Now I know you have no clue who I am, but let me say that I've ported or created over 5 different Icon sets for Linux alone in the past year, that can readily be found on Kde-look.org, so while your comments are welcomed, they have zero merit. I've been a Mandrake user since version 7 and the one thing every Mandrake/Mandriva version has in common is a subpar look to the desktop. Mandriva/Mandrake has been the most popular Home based Linux desktop for the past 2 years and while the distro has improved by leaps and bounds, the look has slowly deteriorated to what is currently called 2005 LE. The Lilo Boot screens, kdm login background, and Desktop Backgrounds are absoultly horrible.
I spend more time changing my desktop than I spend using it.
In fact, before I make any modifications to a fresh install, I absoultely have to rid my PC of all the default Mandrake graphics, Icons, and window decorations. I used Redhat for a brief period when I couldn't get Mandrake to work with my new nforce Nvidia motherboard and while I prefer Mandrake over Redhat, I was absolutely blown away by how professional the desktop looked out of the box. With every Mandrake release, I'm absolutely blown away with how little time they spend on making the desktop look remotely professional. I didn't even feel the need to change anything other than the windec in Redhat, because it looked like a modern OS. I'm not a big fan of Blue curve, but at least it's got polished look to it. I have a hard time recommending the download version of Mandrake to friends, becuase when they do a little research to determine what Madriva is, they always come back and snicker about how it looks. There is a reason Apple poured millions of research dollars into all that OS X eyecandy. Beautiful astetics improve a users perception of the product. LE 2005 may work well, but it's got to be the most astetically disapointing release Mandrake/ Mandriva has ever released into the wild.
Re: Zealots Unite?
So, either explain yourself more fully in your posts or lay off posting comments such as this.
guppetto is right. Mandriva's default look is ugly. They should change it. A proffessional Linux distro should look proffessional. We can change our wallpaper/icons etc. but only they can change the default look. Since only club members allowed to write to their official forum, this is the right place to express his opinion about the look.
Follow up question:
The official www.mandrivalinux.com website is a mess. Any plans to change it? At least the structure.
No central place for the blogs. Mandriva emplyoees don't write to forums etc. are there any plans to act more open?
I've done and posted a Portuguese translation of Duval's interview on my website: http://www.linuxdailylog.com/2005/05...ndador-da.html .
Please feel free to make suggestions or corrections that you want.
I think guppetto is right in all points. I am also a Mandrake user since version 8.0. Duval and his coworkers have done a wonderful job, as I have witnessed the improvement of Mandrake Linux from release to new release. But NOT all is PINK. esthetics is not a strong point in Mandriva. I also spend a lot of time changing the defaults of the desktop, such as themes, colors, icons, etc. because they were probably chosen by consensus (its a guess), and so it looks rather ugly. When I see the MacOS X desktop, I am amazed at the beauty. Its artistic.
Being said that, Duval said something that deserves further analysis. He believes that Linux is ready for the busines desktop bot not ready for the home desktop. Well, I use it for my desktop at home and do 99% of everything I want (What is that <1% left?: for ex., I have to use DVDShrink in Win98 to backup my DVDs; I have used ldvd, but it is not yet as good as DVDShrink). That is not really a big handicap. But I believe that a Linux busines desktop will have more limitations, because the best professional software (commercial) are not yet available for Linux. Yes. That is the principal reason for saying Linux may not be 100% ready. But if you can do what you want with the available applications, then it is 100% ready for your desktop. Ahhhh, what a small detail, yet so big that makes people hesitate in switching from Windows MS to Linux. More explanations like this should be discussed. Of course, the other point is, how much of your time you are ready to invest to learn Linux. After 4.5 years or so, I am still learning how to use Linux. My sporadic and slow learning is moved only when I have to find a solution to a problem. And so, far, I dont say it has been easy. Nope. But I really decided not to give up (many times I was about to!). In part is my pride to learn something new that is pushing me. Hopefully, Linux is evolving into an even easier to use OS and newcomers will not face difficult situations. We"ll see.
Re: Interview with Mandriva (ex Mandrake) Linux Founder Gael Duval
Specifically the "it's easy to install" part. This part is of course dependant on distribution, and even in easiest of distros, isn't neccisarily true.
Installing a system, isn't mearly getting the console up and running, it's getting the system from no OS.. to a working OS that handles all of the hardware. And can infact include all the way to the point when it has all of the software up and running that the user needs/wants.
This proccess can be a complete PAIN in the arse. Especially compared to other OSes such as Windows, and even OSX. And the problem lies in hardware support. Not so much in lack of support, as it is, in ease of support. Sure I got my wireless running properly on my laptop, but it took me days to get going, and I had to scour the web and modify scripts that are in a language I don't know, to achieve this. This is not "easy".
And I have yet to get the video properly configured on it.. ATI.. I do not own a subscribition to the club, so didn't get the included ATI driver.. but that should be a moot point, I should be able to install a video driver without scouring the web for what so far seems to be non existant documentation.
To me, Linux's biggest problem when it comes to desktop support, is that of Hardware Drivers. So far as I can tell, Linux lacks a well developed standardized API for supporting many pieces of hardware. What needs to be done, is the guys who know what they are talking about(unlike me) need to sit down, at a forum, and start working out a set of standards for driver development and support under Linux, that will make it possible for a company like NVidia or ATI, to with some ease write a piece of software that can be downloaded by the user, run, and then will install its own self, and supply any configuration needed, via GUI!
None of this is neccisary if Linux is to remain the sole domain of experienced developers, who are the kind of people who don't need to pay any money to figure out what they need. But if you want a community of home users, who are willing to pay some money for the software to be easy enough that even they can install a graphics driver, then a standardised driver architecture like i described is needed.
Cause frankly, if I am ATI, NVidia, Creative Labs, HP, Cannon, etc.. etc.. if I'm a big company, do I want to say my product supports a system where I have to either take the stance, "we don't provide drivers for this OS, you must use third party drivers developed by people who don't have the true specs, and therefore can not provide you support" or "We have gone through hours and hours of tedium to try to write a driver for this OS, and we have, but you have to figure out how to get it working on your system, because we can't afford a highly trained tech to deal with every different system there is out there, and write scripts and figure out how on earth to make the driver work for your system. blah blah etc..."
Companies just are not willing to pay for customer service for this, because drivers are so non standardised under Linux.. it's cludge.
It can be fixed, I know the Linux scene can work out standards to make it feasable for hardware manufacturers to write a piece of software to supprot the hardware, that my mom can install.
If my mum needed to update the video driver on her windows box, I would be able to easily talk her thought it, "yea click that link, save to your desktop, double click that icon, keep hitting next, hit finish" done!
But my lord, if I tried that on Linux, "yea click that link, save to your desktop, ookay now read through the documentation, mmhmm do you have xfree86 or X.Org?, which Kernal are you running? Ookay open a console, and su then your password, now ls to your desktop, ookay sh some-really-annoyingly-long-driver-name-12-5-2005-3-45-6789, now we need to edit a conf file so open up your XF86Conf or -4 file by running.. etc..." And then, it prolly wouldn't work right the first time, cause SOMEONE would have skrewed something up somewhere, or some hardware or software would be the wrong stuff.
All I'm saying is, to get the home user scene, installing hardware needs to be easy as windows if not easier. And to achieve that, to truely achieve that, making a driver, and driver installer needs to be something that a company can do with some ease, and without bending over backwards to support. It's doable, it may mean making som signifigant changes to X, and other pieces of software, but its completly doable. The word "recompile" should just not exist when it comes to a home user doing ANYTHING. And until Linux can do this, it's always going to sit behind Windows and OSX in atleast one major respect, Gaming, cause until GOOD driver support exists for video cards, companies just aren't going to want to write software for Linux, too many driver problems.
I've said too much, I sooo shut up now...
I would like to thank the people at formerly Mandrake and now Mandriva, I have been using Mandrake since version 9.0 and enjoyed its ease of installation especially as I still dual-boot with Windows 98 SE.
There is one gripe however
When I install Mandrake ( haven't yet tried Mandriva ) I am left without any web-support for Java virtual machine, flash and some other web apps needing plugins. Is this going to be addressed in future versions. It would be nice to just install Mandrake / Mandriva and go onto the net without first the hassle of installing plugins and praying they work.
Admittedly I've always used the download versions of Mandrake but with the plugins being free anyway is there a plan to have already incorporated the plugins into the distro so that Firefox and Mozilla etc just work "out of the box".
The main features of the power pack are the NVidia and ATI graphics drivers, and Flash, and Java. The graphics driver is a gimmie when it comes to blackmail.. I mean really, you HAVE to have a working graphics driver to use a computer.. thats essential stuff.. you just can't get away without it. And Flash and Java are essential pieces of the web today, and again, hard to use the web without it. Kinda like giving you a car, but without critical compents of the engine, ones that are difficult enough to get and install into the engine, that they figure you are willing to pay the low fee to get them to install it for you.
I'm still let down by Linux unless you can't tell. The OS scene just hasn't made it what it could be yet, and I dunno anymore if business is helping it, or hendering it. But it seems like progress has been slow, for the amount of funding going into it. I'm left with the impression most of the time is spent trying to get new hardware working, and helping people get the software running right. Things that could be fixed with what my previous post mentioned.
(me = shutup)
Aren't you confusing the definitions of free?
These drivers/ add-ins etc. may be a free download, but they are not open source. So if they are included in the "download" version, the download version is no longer open source. That's against the GPL.
Also, a follow up question: what is the outlook for preinstalled Mandriva systems? It seems they had some HP notebooks with Madriva dual booting offered through the Mandriva website, but the store says this is being discontinued.
Was this a test run? Could it be offered again? Any chance it will be tried outside of France? Because here is the UK there is a grand total of one distributor of pre-installed Linux (Digital Networks UK), and they only do desktops.
|All times are GMT -5. The time now is 10:38 PM.|