OSNews

The problem is the fault of Linux failing to provide a stable driver API/ABI - if Linux provided said stable interface then the responsibility could be then placed back on the shoulders of hardware vendors to maintain their drivers. The reality is that every distribution of Linux isn't compatible with the next, you can't simply just grab a driver compiled against one kernel version of a particular distribution and then have it work without any problems on another - that is the fault of the OS vendor. Compare that to Windows which has a stable API and ABI thus barring any stupid decisions made by the programmers at said third party hardware vendor then things should work without too many problems.



You don't have to be technically inclined to realise "you get what you paid for" - heck, my sister bought an el-cheapo NZ$600 Acer laptop and she knew what she was getting but as she said to me it was a throw away laptop and that should be move onto something better later on. Lets not assume that every person out there is clueless because a fair chunk of people, even if it is by gut instinct alone, that you pay for what you get and that the extra money isn't necessarily for the brand but also the fact that these companies support you for the long haul.



Which mean nothing given that OEM will have newer drivers from their suppliers than what is actually available on the internet. Synaptic for example is already shipping Windows 8 compliant drivers through Lenovo (IIRC 16.2.x) which aren't available on the Synaptic website itself. Regarding ACPI, again, that requires a power management driver which requires the OEM to provide an updated driver - again, if it is anything like the Lenovo situation then the driver will be release either on the day or a week or two after Windows 8 has actually hit the market. Windows 8 has been RTM'ed but it is not general availability, it isn't shipping and very few hardware vendors have Windows 8 test drivers available - again, come back once Windows 8 has actually shipped to the general public then draw your own conclusions base on reality and not someones half baked review based on September 5 drivers that were tested only on Windows 7.



Then you should wait till the driver has been released - goodness gracious me, what is so f--king hard about looking up the hardware vendors website and checking whether they have a Windows 8 compatible driver first before deciding to install? install first ask questions later?

Edited 2012-09-27 08:47 UTC

Actually, as much as the concept of a stable driver ABI is lovable in theory, Windows does not really have it anymore. As an example, Microsoft switched GPU drivers API ABI from XP to Vista, added new features that required new drivers to be supported in Win7, and now they are breaking it once more just for the fun of it with Windows 8. Goes to show that in the realm of hardware support, manufacturer cooperation is much more important than technical merits.

It is true that Microsoft do not usually break their driver APIs during the lifecycle of a given release, though, unlike some unscrupulous Linux distros which ship incompatible kernel and Xorg releases as harmless security updates.


I know a lot of people who understand that a cheap laptop will have crappy hardware, that fail early, overheats, and that kind of crap. Fewer people, however, seem to avoid blaming "Windows" or "PCs" for awful OEM installs.


Are you trying to sell me that RTM copies of Windows do not include at least basic support for one of the world's most widespread touchpad brands ?


Basic ACPI operation (such as sleep and resume) shouldn't require a driver and per-manufacturer support. It has been designed precisely to standardize power management and avoid that kind of mess. Some device drivers should explicitly support ACPI and run routines on sleep and resume in order to avoid problems on resume (like, say, assuming that a network card is still connected to a network), but failure to do so should not cause a full kernel crash.


I guess we will indeed find out in a month, but if drivers are still not ready at this point I would tend to predict a Vista-like scenario with very few mature drivers at release time myself.


Hardware vendors will claim very early to have a Windows 8 compatible driver. What they will not necessarily announce is when said driver will actually be in a stable and usable state.

Edited 2012-09-27 17:30 UTC

Though expensive, high-sticker-price hardware doesn't really assure future support. It being rarer, more niche, can in fact decrease the chances...

Around here I have two webcams from the end of last century. One is as-classic-as-it-gets Logitech Quickam Express, very inexpensive even at the time ( http://en.wikipedia.org/wiki/QuickCam#Cameras and mine is even a black-logo OEM package, was notably less expensive) ...and with quite horrible quality, fairly crappy overall. The other is a Philips Vesta Pro/Scan, with decent quality, exceptional low-light performance (which gave it a prominent place in amateur astrophotography http://homepage.ntlworld.com/molyned/web-cameras.htm http://www.pmdo.com/wintro.htm http://keithwiley.com/vesta675mod/vesta675mod2.shtml http://thierrylambert.free.fr/materiel/vestapro/vestaprocielprofond... http://www.astrosurf.com/astrobond/ebmate.htm ), and quite pricey in its day, I believe (it found its way to me as a surplus hw) ...remaining so for a while, I think, with how it was sought-after for astro.

Now, I'll let you guess which one of those webcams is supported by all popular Windows versions in a totally plug'n'play fashion (as in, you connect it and everything works, the driver is included with the OS - despite the webcam predating USB video class standard), and which is supported only up to XP... (and requiring 3rd party driver there)

Similar with one "pro" audio card that also found its way to me, E-mu APS, essentially abandoned a long time ago. Except, it's related to SB Live! (I believe APS was the first card using the EMU10K1 chip - it sure has visibly "old style" packaging there), so it's supported by a sort of community project targeting this family: http://kxproject.lugosoft.com/


If you want to be safest from driver woes, get what's popular and still suitable to your needs.


Oh I don't know if this one would be so bad - from what I see, most laptops are used with a mouse attached, anyway.