The Trouble With Mandriva
Feb 9, 2012 5:00 AM PT
Now that Linux distributions like Ubuntu and Mint are enjoying such widespread attention and success, it's increasingly difficult to remember that not all distros are sharing in the same good fortune.
Take Mandriva, for example. This venerable distro dates back to 1998, when it was born as Mandrake Linux, but its last few years have been tempestuous. Bankruptcy fears have long loomed on the horizon, and today things aren't looking any better, thanks to shareholder infighting and a variety of other problems.
What's a distro to do? According to the blogger masses over at Slashdot, looking beyond just the OS itself could help.
'Any Room for a Straight-Up Distro?'
One of Mandriva's main problems is that "it never grew into more than just an OS vendor," wrote jfruh on Slashdot recently.
"The big players in the commercial Linux space -- Red Hat, SuSE, Canonical -- all built Linux into their larger computing visions," jfruh added. "Is there any room in the marketplace for just a straight-up Linux distro anymore?"
Nearly 300 comments later, the answer still wasn't clear. Linux Girl took to the blogosphere's rowdy Punchy Penguin Saloon to learn more.
'It Didn't Keep Up'
"Mandriva's problem isn't that it only sold Linux, it's that it didn't keep up as a viable distro and instead fell behind and became yet another Red Hat clone," consultant and Slashdot blogger Gerhard Mack told Linux Girl.
"It's not a shock that there is no money in being yet another knock-off product," Mack added.
Alternatively, "I think the Slashdot question confuses distros and vendors," offered Chris Travers, a Slashdot blogger who works on the LedgerSMB project.
Free as in 'Speech' and 'Beer'
"There is room for stand-alone Linux distros (Debian, for example, or Slackware), but there is no room for a vendor that does nothing but sells the OS and related services," Travers explained.
"The fact is that Linux is free both in the 'speech' and 'beer' senses, so if you try to make it un-free in one way or another (by bundling proprietary apps, and/or by charging money for it), you are competing against folks with larger, more active communities and a better price point than you can offer," he added.
Another problem arises when it comes to support contracts, Travers suggested.
'It Hurts Your Bottom Line'
"Why build your own distro for your support contracts when you could tie into a free one that you don't have to maintain yourself?" he pointed out. "It's overhead, and it hurts your bottom line."
For this reason, then, "I think that Linux vendors really have to start with a computing vision, an idea of how customers can better their operations using free and open source software, and then build around that, whether it is a custom Fedora spin, a fork of Debian, or whatever," Travers concluded.
"Otherwise, the vendor is not really adding value sufficient to cover their expenses," he added.
'A Millstone Around the Neck'
Indeed, "because FLOSS is available for $0, anyone charging money for it is going to be squeezed for cash," opined blogger Robert Pogson. "It is almost necessary to bundle FLOSS with other goods and services, especially for GNU/Linux, a commodity available from multiple sources."
Very few of the large, noncommercial distros are failing, Pogson pointed out. Rather, it's the commercial ones like Mandriva that are "dependent on sources of funds that can dry up," he added.
"From its early days as Mandrake, the 'insider club' that paid to use Mandrake for 'extras' made the organization dependent on constant growth and the generosity of users," he explained. "That is hard to sustain, and the cost of advertising and payroll are a millstone around the neck of such a distro."
'Maybe Hard Times Are Needed'
Roberto Lim, a lawyer and blogger on Mobile Raptor, saw fragmentation at the root of Mandriva's problems.
"If there was one default Linux desktop OS and one default interface, and power users could 'upgrade' the operating system with add-ons and replace the UI with whatever they liked, I think Linux would have a much better chance at success," Lim told Linux Girl.
"I will be sad to see Mandriva go," he added. "I used Mandrake for some time back in 2003 or 2004. But maybe hard times for Linux are needed: From a purge, you will get one dominant Linux desktop operating system."
Otherwise, "we might as well stop calling it Linux, and just refer to each as Debian, Fedora, Ubuntu, Mint and the like," Lim suggested. "In the end, they have little in common to the typical consumer, and referring to them under one banner probably makes things more confusing."
'Be Ready to Trade Money for Market Share'
Indeed, "is it even rational to expect vendors to survive by offering the same product as everyone else, when the moment you come up with something better anyone else can just copy all the good bits?" mused Barbara Hudson, a blogger on Slashdot who goes by "Tom" on the site.
"If there's one industry that badly needs consolidation, it's the various linux distros," she added.
"It's a cruel world out there," Hudson explained. "The days of hoping to support your distro by selling CDs and DVDs were killed by fast Internet. The paid-support model is a bust for most distros -- you need people pounding on doors, and previous success stories, to have any hope of getting business and government contracts.
"OEM deals are drying up as manufacturers either 'roll their own' or go with Android," she continued. "Bundling paid me-too services such as 'the cloud' doesn't cut it when there are so many platform-agnostic solutions from specialists. Bundled Music and eBook stores or hardware? Be ready to trade money for market share."
'You Ain't Seen Nothin Yet'
The real issue, however, is whether the concept of the distro is changing, Hudson suggested.
"Why not just use a web interface to pick and choose what you want, both in terms of packages and customizations, or mix-and-match from a selection that other users have created?" she explained. "You could even say, 'I want the same one my friend Jane is running,' have it bundled up, and download it."
In fact, the beginnings of such capabilities are already visible in "online build services that take care of issues such as resolving dependencies for different platforms," she pointed out. "Releasing your own distro has never been easier, and if you believe choice is good, it's just going to continue to get better.
"The only downside? If you think that the current situation of having to choose from a thousand distros is hard, trust me, you ain't seen nothin' yet!" Hudson concluded.
'The FOSS Model Simply Doesn't Work'
Last but not least, Slashdot blogger hairyfeet pointed to money as the root of the problem.
"The FOSS model simply doesn't work outside the enterprise and embedded markets," hairyfeet told Linux Girl.
"It cost over $300 million to create Windows 7, something like $200 million to create OSX lion," he pointed out. "Now where are you gonna find a company willing to invest THAT kind of money in Linux?"
'They Would Never Break Even'
The answer is, "you won't, simply because they would never break even, much less have a ROI," hairyfeet explained.
Red Hat makes its money via support contracts, "which home users simply won't buy," hairyfeet noted. Yet "it's the consumers that buy, it's the consumers that will replace a perfectly working device for the latest 'shiny,' it's the consumers driving the market.
"So how can a Linux vendor possibly invest the tens of millions --I would argue hundreds of millions, because you'd need to fork the whole thing away from the itch-scratching developers -- if they can't sell what they've invested in?" hairyfeet concluded.