Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> I think the idea of $random_component_vendor maintaining an up-to-date Linux driver themselves may be unrealistic

I agree. I wonder what the Windows model is, and is there a way to emulate that in Linux?



The windows model is that you get a binary-only driver that works with the current (certified for that component, etc) version of windows. You maybe get one or two updates over the next half a year. Then nothing.

If there is a bug, exploit, whatever, good luck. If there is a newer version of the component with a minor difference (e.g. different range of serial numbers, different device id, etc) usually the newer driver won't work, so no bugfixes for you. However, you can keep on upgrading your windows version, at least within the same Service Pack and major version, sometimes longer, because the age-old driver will continue to work.

The way to emulate it would be "have a binary compatibility layer for kernel modules". However, this gets you old, broken, weird and incompatible drivers that just happen to run with the new version if you are lucky. Also, because every vendor will reinvent the wheel while doing the minimum amount possible, drivers will be less capable, come with weird special tools, crapware and a generally bad compatibility to devices from other vendors.

Believe me, the grass isn't greener on the other side, just different.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: