On Mon, 26 Sep 2011 17:54:47 +0200, Uwe Kleine-König wrote: > Hello Guenter, > > > renaming the file makes it technically impossible to review the changes. > oh, I forgot to add -M to format-patch. I can resend when we agree on > renaming or not. But note that even then there are many hunks as I did > s/mc13783/mc13xxx/ in the driver. Please don't do that, it's pointless. Naming drivers and functions after the first supported device is a very common practice. > > Please refrain from it, and use the mechanism used for other hwmon > > drivers - add support for the new chip to the old driver, and reference > > the supported chips in Kconfig and the driver documentation. > Ah, I missed to rename Documentation/hwmon/mc13783-adc. Will update that > on v2. > > > Besides, the new name implies that all future MC13XXX chips would be > > supported by this driver, which seems to be a bit difficult to predict. > hmm, but today mc13xxx is the best name we have. And I prefer a driver > named mc13xxx over mc13783 because when looking for a driver for the > mc13892 and seeing only mc13783-adc I'd assume it doesn't support > mc13892. OTOH if there is a mc13xxx-adc driver I'd look in more detail That's simply a wrong assumption from you. > if my chip is supported. With name mc13xxx-adc you'd have to check the details just the same, as you have no guarantee that the xxx's match the chip you have. > So if you ask me, use the best name available today and think about the > best naming again when the next chip comes up that fits the naming > scheme but not the driver. I don't think there's any name to find. The driver has a name today, stick to it, and properly document which devices it supports. Thanks, -- Jean Delvare _______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
