Linux and its closing window of opportunity with OEMs

I am planning on changing the world with this article. I can’t do it on my own: I need your help.

Well, I must admit that changing the whole world might be a little ambitious. For now, I will settle for the “computing world”.

Right now, the following factors are true:

  • Linux has a very viable desktop and office suite—for free. OpenOffice being bloated is basically not an issue anymore, since even a basic computer today will run OpenOffice completely fine. Thanks to Ubuntu, end users can now use Linux and not notice the difference.
  • Again thanks to Ubuntu, Linux is amazingly simple to configure.
  • Computers are getting cheaper and cheaper. I am using a $950 laptop, and it’s an amazing machine which will probably last many years.
  • Linux’s hardware support is impressive.
  • Vista is being released. All the anti-piracy procedures will annoy users immensely. Plus, Vista is a new system: a big break from XP.

Despite what people say, Linux does not have a significant slice just yet. By “significant slice”, here I mean 20% to 30%. We are nowhere near it, in fact.

Linux has, right now, a huge window of opportunity to actually grab a huge slice of the desktop market. The time is ripe. The opportunity is right there.

And yet, there is one missing piece that is stopping it. In fact, “stopping” here is an understatement. This missing piece could well kill this fantastic opportunity.

The missing piece has a name: OEM. Or, lack thereof.

There is no easy way to go to a major brand (see: Lenovo, HP, Toshiba) and buy a Linux laptop with Ubuntu preinstalled on it.

Why not?

I can see three main reasons:

  • Lack of demand. Yes, not enough people ask for Linux. This is sadly true. Linux users are still a minority.
  • Microsoft lobbying. There is no conspiracy theory here: it’s something everybody knows. Executives of big companies are being pressured (see blackmailed) so that they will never allow people to buy a Linux laptop.
  • Support. Laptop manufacturers have dealt with Windows since day 0. Linux is scary—and I am talking about money, here: big vendors are scared by how much supporting a brand new system will cost.

Lack of demand will not rise until they start offering Linux laptops. This is a catch 22 well known to the computing industry. Microsoft lobbying will never stop. In fact, I suspect it has grown stronger and stronger. Support is an issue that manufacturers will only face if they are forced to.

Linux needs a range of laptops that will never, ever give even a single problem.

Some can argue that Linux can “make it” without the manufacturers’ help.

This is simply not true. Go out there, read the mailing lists, read the bug reports being submitted for sound, suspend support, video cards. You will find countless hackers struggling and trying to guess which bit should be raised at which point. You will see users endlessly applying those patches, and reporting back to the drivers’ developers (who are real saints, if you ask me).

This is not a good way to go about it. It might have worked so far, but it is rapidly becoming unsustainable.

Why? Because the amount of available hardware and devices is growing—and the available documentation is diminishing. Because there is a limited number of people who are skilled (and patient) enough to hack a driver so that it corrects the weirdness of the chipset XYZ that only shows up when the BIOS ABC is being used with parameter QWE on. If you go out there, you will see that it’s a constant battle. It feels like Linux is defending itself, but the enemy is getting bigger and bigger.

Also, Laptops tend to become better supported once they age a little—six months to a year. In that time, hackers have time to fix drivers, support more devices, and so on. This is a problem still: people want to buy new, fast laptops rather than last year’s models. Plus, this year’s version of a laptop might use slightly different chipsets—starting a driver-fixing war again.

So, what does an OEM need to do in order to create a “Linux range”?

  • Make sure that free versions of drivers covering every single piece of hardware used by the machine is fully supported by Linux. By “fully supported”, I mean no weirdness, no strange behaviours, no quirkiness. If there are problems, the OEM manufacturer needs to either solve them (most likely by working with the chip’s maker), or not use the defective piece of hardware.
  • Make sure a popular distribution works. For example, they could make it “Certified to work for Ubuntu”, and then list “Dapper Drake, Edgy”.
  • Make an installation disk that install a system that always, always, always “just works”. Again, this is relatively easy with Ubuntu.
  • Offer Linux support to people with problems. This could mean outsourcing some of it, for example. Or training people within the company. They need to keep in mind that here “support” doesn’t mean “OpenOffice won’t change font”. This kind of support is not offered to Windows users either.
  • Have a big range of laptops, from small portables (12" screen laptops, at about $950) up to higher end machines (17" monitors).

When this happens, the world will change. Why? Because OEMs will be directly involved in the tuning up of the hardware they provide. They will either fix existing drivers, or buy components that work 100% in Linux. No matter which way they go, Linux will benefit immensely: component makers that don’t support Linux will lose money.

Microsoft knows for sure that if one of the largest OEMs does exactly what I described above, others will follow. It will become an unstoppable trend, which will cost Microsoft market share—the same market share that brings value to their stock and sales dollars into their bank accounts. This is why Microsoft has worked so hard, politically, to make absolute sure that OEMs supporting Linux properly just didn’t happen. Sadly, they managed.

How can we make this happen?

  • Show that there is demand. This could be done in several ways. One is for example by calling laptop makers, and asking which Linux laptops they sell. If they don’t, then it’s important to ask the person to speak to a teamleader and have the complaint logged. Another way, is to create a new project with a full range of latest, bleeding edge laptops which work in Linux out of the box. Since this is aimed at desktop users, “certifying” these laptops for Ubuntu would probably work.
  • Lobby OEMs at higher levels. Executives need to understand that by not providing Linux laptops, they are actually losing money. This is tricky—it’s hard to get to talk to people “at the top”. And, normally Microsoft has the potential to buy expensive “presents” for these executives... But, the Linux community is large, and we all know the principle of “six degrees of separation”...
  • Send a clear message to OEMs: they can outsource support to start with. Canonical could provide that support, for example.

Not acting right now could cost Linux dearly. Very dearly. There is a window of opportunity open right now. It will eventually shut. Microsoft might be able to keep their stronghold on OEMs forever. Trusted computing is their next card to play—Linux will be even harder to use on a lot of hardware. And things will only get worse.

What are your comments about this? What do you think I got wrong? How do you think we could achieve what I got right?

I am here, listening to your comments.


Verbatim copying and distribution of this entire article are permitted worldwide, without royalty, in any medium, provided this notice is preserved.