On Fri, Jul 28, 2006 at 03:35:10AM +0300, Shem Multinymous wrote: > On 7/28/06, Vojtech Pavlik <vojtech@xxxxxxx> wrote: > > note: It's absolutely necessary to limit the API to a well usable > > SUBSET of a superset of the features of all drivers/devices, > > even sacrificing obscure features to keep the API sane. One > > example would be the HID Power spec, which simply can't be > > supported to full extent by any sane API. > > Non-standard functions must be handled reasonably within the > framework, otherwise drivers will have to build duplicate interfaces. > > How about > /sys/whatever/battery0/voltage for standard attributes > and > /sys/whatever/battery0/thinkpad/inhibit-charge-minutes > for non-standard ones? I think the thinkpad features actually fall into the 'standard API', if not for any other reason, then because some 80% of kernel developers use ThinkPads. ;) > > + and the kernel can change the polling frequency based on power > > saving state changes > > Likewise for cached attributes (query hardware only if N jiffies > passed since last querry, other return cached value). And that way, > hardware query frequency is never higher than what userspace actually > needs. -- Vojtech Pavlik Director SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
