mouse@Holo.Rodents.Montreal.QC.CA (der Mouse) writes:
>>>> So wouldn't it be usefull to reserve some numbers for attapi, or
>>>> will it be never included to the official tree some day?

>>> The atapi source are unoffical changes so we can't really care about
>>> the device number for every unoffical driver.

>> What a load of nonsense.  You either have to start a registry for
>> device majors for third party drivers or implement the tools that
>> make it possible to add a device driver and assign it a major number
>> completely automagically

>Why do we/they "have to" do anything of the sort?

I guess I was unclear.  That was meant as an answer to the question
how one could possibly know about what majors are in use.

>If you hack on
>conf.c, you've _always_ been on your own when it comes to avoiding
>conflicts with future releases - or conflicts with what the guy across
>the hall is hacking in, for that matter.

I've never had to "hack on conf.c".  What I had to do was explain to
customers how to "hack on conf.c" to get our drivers installed.  And
frankly, none of the customers that payed for our products had the
expertise or interest to do so.

As I mentioned before, Syvish systems have used a machine generated
cdev and bdev switch for well over ten years.  It's simply not
necessary to hack conf.c when adding a device driver.

>I agree it would be a good thing (not "necessary", as you seem to
>think) to reserve a range of major device space as "never will be used
>by a released system", for private or otherwise unofficial drivers.

That, could be a first step in the right direction.  It's also easy to
implement.  It would be a good idea to set aside a range of majors for
port-specific drivers.  Of course, the best solution would be to
generate the necessary configuration files from an abstract
description of the driver's needs and capabilities.

-- 
Christoph Badura	bad@flatlin.ka.sub.org

You don't need to quote my .signature.  Everyone has seen it by now.
Besides, it doesn't add anything to the current thread.