Phoronix: Canonical: Ubuntu To Soon Ship On 5% Of PCs
Chris Kenyon, the VP of sales and business development for Canonical, just spoke this afternoon at the Ubuntu 12.10 Developer Summit about what Canonical does with OEMs and ODMs. He also tossed out some rather interesting numbers about the adoption of Ubuntu Linux...
The biggest problem for Ubuntu and other Linux distros shipping to the masses is that, in a lot of cases, customers are installing Windows after they get the unit. And not just dual booting; they wipe Linux and never come back as soon as they figure out that it won't run whatever exe trash they download from the internet. And then they immediately stick in a Windows CD (often pirated) and don't look back.
This can actually be a good (illegal) business proposition for individuals looking to buy hardware: instead of the hassle of trying to return an unopened Windows license that comes OEM with your hardware, just buy a computer with a free Linux distro on it like Ubuntu, pirate Windows, pirate all the apps you want, and you get an OS that does what you want it to (in the simple sense of "runs exactly what I want and what I'm used to", not in the sense of running what might actually be better) for free. No paying for Windows and no suffering through learning Linux (ohnoes!).
I think it's a terrible and extremely hypocritical idea to do it, but still, it's very common. And it hurts desktop Linux each time it happens.
IMHO we should start doing something like this with desktop Linux :-
1. When an OEM is building a Linux distro, they should pre-install various things that will make the user's experience much easier. For example, they can pre-install Wine and wire it up so that you can literally download an exe and click it with Firefox and it'll run. Or they can pre-install a licensed copy of Crossover Linux (purchased at a volume discount from Codeweavers) to do the same. They can also ship the Fluendo codec pack and Adobe Flash, so that people will have out of the box Flash video, decoding of all common proprietary media, and Windows file compatibility.
2. Strike deals with game developers to provide very low-cost games as a bonus/add-on when buying a computer, but only give them the Linux version. So they have an incentive to keep the Linux: you want your game, you run Linux. The OEM should eat this cost out of profit so that they will increase the value of their unit, even though the cost will come out of profit (hopefully it'll be a very low cost if good deals for $5 - $10 per sale can be worked out).
3. Build up the support infrastructure. Give end-users a number they can call and get very good, robust support for doing things they want to do, even if their particular use case isn't exactly anticipated ahead of time. Hire support technicians who are innovative, think on their feet, and can help out with remote desktop at a very low cost to the customer. Let's not only match the Windows tech support infrastructure; we have to be better than theirs.
4. Make sure that shipping distros provide the latest stable updates to end-user-visible applications! This is critically important. Firefox, LibreOffice, and anything "of that sort" (where stable updates are released independently of the distro and are generally well-maintained and QAed upstream) should be automatically imported into the stable distro post-production, following a brief in-house testing period. There's really no reason not to. Same goes for proprietary stuff too, unfortunately; people want their Flash and Chrome and Fluendo codec and Skype updates as well.
Gotta reach out there and grab people and tell them "hey, DID YOU REALIZE that the Linux distro that comes on your PC is so much better than Windows, and all you have to do is give it a little try?" And don't give them much chance to say no -- incentivize keeping Linux. Obviously we can't enforce vendor lock-in quite like Apple or Microsoft can, but we can certainly hold a carrot on a stick... they're free to not bite it, but they'll want to.
1. When an OEM is building a Linux distro, they should pre-install various things that will make the user's experience much easier. For example, they can pre-install Wine and wire it up so that you can literally download an exe and click it with Firefox and it'll run. Or they can pre-install a licensed copy of Crossover Linux (purchased at a volume discount from Codeweavers) to do the same.
Almost all people want what everyone else has. They're not going to want something different until it's the cool thing to do, or until everyone else is doing it. Everyone wants to stay on Windows because that's where everyone's at.
The only people that are going to make the switch are those who are technically inclined enough to see the benefits of Linux and want something better / different.
But Canonical needs to do something with this God-awful color scheme.
the problem is that for a desktop user linux hardly has any compelling reason over windows. there is no selling point like iLife. you get less functionality with more bugs. that's the cold hard truth. apple had a similar problem in the 90s when there was a severe lack of apps.
This isn't the 90's though. I'd love to see some numbers that show percent of time a user spends in their web browser vs. all other apps combined. I'm betting the browser trumps like 10:1
In which case, the Windows browsers have better graphics acceleration, more mainstream browser choices, more available plugins, and better sandboxing/security. So, yeah. Go Linux.
Also, that small fraction of non-Web apps still freaking matter. People run all kinds of little (or big) apps that are not Linux-friendly. Even if it's just "non-essential" items like games. Your computer experience is only as good as the worst part, so even if Linux is a fantastic Web OS, when it can't run a particular desired app the sum of the experience is going to be "it can't run what I want."
Put in a crass metaphor: Your ratio of non-sex to sex every day is quite likely higher than 10:1 but that doesn't mean that you're going to opt into a sex-free life to save a few bucks on the costs of dating.