cvs commit: src/sys/i386/conf DEFAULTS GENERIC

[Scott Long]

Maxime Henrion wrote:
> Scott Long wrote:
> 
>>David O'Brien wrote:
>>
>>>On Thu, Oct 27, 2005 at 05:34:35PM +0000, John Baldwin wrote:
>>>
>>>
>>>>jhb         2005-10-27 17:34:35 UTC
>>>>
>>>>FreeBSD src repository
>>>>
>>>>Modified files:
>>>>  sys/i386/conf        GENERIC 
>>>>Added files:
>>>>  sys/i386/conf        DEFAULTS 
>>>>Log:
>>>>Create a default kernel config for i386 and move 'device isa' and
>>>>'device npx' (both of which aren't really optional right now) and
>>>>'device io' and 'device mem' (to preserve POLA for 4.x users upgrading
>>>>to 6.0) from GENERIC into DEFAULTS.
>>>
>>>
>>>I may be missing something.  I don't quite follow the benefit of the new
>>>'DEFAULTS' file.
>>
>>It's been 2+ years since the io and mem devices were made optional, and 
>>the mailing lists are still filled with people who don't understand why
>>X doesn't work after they remove them from their kernel config.  We 
>>expect there to be a large migration of people from 4.x to 6.0 who
>>never tracked the change in 5.x, or who want to bring their 4.x kernel
>>config files over with as few surprises as possible, so this will make
>>their lives easier.
>>
>>
>>>I'm also curious why we don't explicitly 'include'
>>>DEFAULTS in GENERIC vs. the new automagic include feature.
>>
>>Because if it was specifically mentioned in the GENERIC config file then
>>it would be deleted by people who don't understand what it does or why
>>it's important, and it would be missed by people writing config files
>>from scratch or migrating from previous versions of FreeBSD.
> 
> 
> While I'm all for making FreeBSD less prone to errors like the one
> discussed here, I feel that having: a DEFAULTS file, a good comment
> explaining what purpose it serves in it, an explicit include DEFAULTS in
> GENERIC and a big scary comment next to it inclde explaining why one
> should not remove it ought to be sufficient.  I believe this approach
> would give enough seatbelts to our users while being more pleasing
> technically speaking, since we wouldn't have an automagic include
> feature in config(8).  (It should be noted that there were no comment
> next to the mem and io devices which can probably at least partly
> explain why it has been such a recurent problem).
> 
> Just my 2 cents...  Thanks to you and other re@ members for all the
> great work accomplished with 6.0-RELEASE.
> 
> Cheers,
> Maxime

Big scary messages have a tendency to be ignored just as well as little
scary messages or no messages at well.  See also: npx.

Probably the biggest omission here was a good explaination to the
mailing lists and a section for the handbook and manual pages.  If
anyone would like to help with this, I'd gladly appreciate it.

Scott