On Tue, Jul 27, 2010 at 7:51 PM, Michal Nazarewicz <mina86@xxxxxxx> wrote: >> --- On Tue, 7/27/10, Michal Nazarewicz <mina86@xxxxxxx> wrote: >>> Here are the changes to the composite.c: >>> * https://patchwork.kernel.org/patch/113584/ > > David Brownell <david-b@xxxxxxxxxxx> writes: >> NAK on at least this proposed change, since >> non-managed serial numbers are deeply broken. > >>>> + str = iSerialNumber ?: "0123456789AB"; > > My thinking was that it's better to provide a fake serial then no serial > at all. From what I've been informed Windows in some cases do not like > mass storage without serial number (ignoring other LUNs). > > Later in the code, you'll notice that fake serial number is provided > only when gadget driver explicitly state that it needs a serial number > (so gadgets won't be affected unless they opt for this behaviour) and > that composite will issue a warning about missing module parameter. > > I thought it'll be a good compromise. Still not convinced? http://blogs.msdn.com/b/oldnewthing/archive/2004/11/10/255047.aspx Having unique serial number is the best. Having no serial number at least works in most cases. Having non-unique serial number is the worst. Regards, Xiaofan -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
