Artem Kachitchkine wrote:
> On 03/02/09 03:29, Cecilia Hu wrote:
>> I am sponsoring this case for Tim Chen.  It is to provide a new
>> wireless(IEEE802.11b) driver, atu(7D), for Atmel AT76C50x USB card.
>
> I'd be the last to complain about naming, so let me start by saying: 
> +1 on the case. But a note for the future: we've tried to name USB 
> drivers with names starting with "usb" or at least "u", though I guess 
> that line has faded since rum(7D) and zyd(7D).
>
> -Artem

IMO, naming conventions for drivers is a doomed effort.   You wind up 
with either long names to avoid collisions in the namespace, or names 
that don't make sense.

The only thing I'll say further about it is, I would really like if 
*device drivers* (I don't care about misc/ modules) could please avoid 
using *numbers* in their names.  Names like e1000g, audiovia823x, and 
audio810 are a PITA.

I'll even go one further than that -- I'd like to suggest that 
henceforth, no driver should be allowed to integrate which uses trailing 
digits in its name (audio810 being an example).  It breaks all kinds of 
parsing and forces folks to come up with special hacks to print device 
information.  (Instead of just <driver><instance>, you have use a 
delimiting colon.)   Note that this rule already exists for NIC devices 
--  I just would like to see it expanded to all other devices.

    -- Garrett

Reply via email to