With the impending demise of Windows XP (even though it has recently been announced that XP will now continue to receive updates until July 2015), the prime time for migrating casual Windows users to Linux is nigh.
However, one crucial aspect remains: driver support.
Some will be swift to point out that in-kernel driver support has come leaps and bounds lately, and most things just work “out of the box.” Unfortunately, that is not sufficient in the Real World.
Unless you stick to fairly standard peripherals and ones from major manufacturers, there are still instances when you don’t have kernel support for your device and you have to go fish for one – and the experience is hit and miss. If you have hardware from a major manufacturer, you may strike lucky and be able to get a Linux driver for it.
Then, if that fails, you need to find a Windows driver for it. Here’s the catch:
- ndiswrapper was specifically made to cater to wireless cards, but is used for just about any driver including printers, mice, etc, but without any guarantee that it’s a suitable hack
- ndiswrapper does not support anything other than Windows XP drivers. That’s what it was designed to handle, and it rejects drivers for later versions of Windows.
What options do we now have for the migrators (and those of us who help such Linux immigrants) to allow them to continue to have their choice of hardware, major or minor manufacturer, despite some lack of native driver availability?
From a Freedom standpoint, we obviously need to push manufacturers to publish their driver sources, failing which the second best is at least to publish the design specs.
But from a user standpoint, and a computer admin standpoint, and a migrator standpoint, this does not work in the Real World.
“Use hardware that respects your freedoms” urges Stallman – but only two consumer devices are widely known to do that. Other than that, the FSF publishes some “complete” systems that also meet FSF criteria – they are bare motherboards.
In my view, this goes very much against a consumer user’s freedom of choice currently. User-base and market share drives manufacturing companies, and if we have no weight to pull in numbers of computer consumers, the Freedom rhetoric is cute at best.
The ndiswrapper project still hasn’t seen fit to extend their scope, and nobody has yet picked up the mantle on that front either – we’ve all been content with ndiswrapper until now, for decent reason, but the writing is on the wall and we must act now. Windows XP’s lifetime extension will buy us some time to get a new system out and working, and iron out the kinks before we get a new opportunity to woo Windows users over – but who knows how long third party manufacturers will keep XP on life support? Once software and hardware houses stop putting out Windows XP drivers (when they don’t provide open source or native ones), we’ll be left scrambling again, and it’ll be hard to argue a case.
That or we can cave in to the idea that Linux will never be for the casual, non-technical user, and the idea of “Free Software” is just an academic dream, impractical if not impossible to fulfil in the desktop market.