Subject: Re: vmstat, iostat etc no longer work?
To: None <perry@piermont.com>
From: Chris G Demetriou <Chris_G_Demetriou@ux2.sp.cs.cmu.edu>
List: current-users
Date: 11/14/1996 16:31:08
> BTW, I find the notion of options that are not optional disturbing.

Same here.


> If an option is not optional on a given port, it should be on by
> default in that port and the user's config files shouldn't have
> to have the thing at all. "Option" should mean "optional".

I find this one (perhaps the only) great reason to use the
'std.${MACHINE}' config files.  Some things (probably
MACHINE_NONCONTIG) you can put into headers, but some, e.g. the exec
module selection you can't, and those options which are 'mandatory'
shouldn't have to be listed in every kernel config file.  Not only
does that make it harder to understand the kernel config files, it
makes seamless changes harder to accomplish (e.g. on the alpha, the
EXEC_* changes didn't cause any problems for anybody since they were
handled in the std.alpha file).

However, some port maintainers seem ... opposed to the idea of
std.${MACHINE} files (at least for their ports), though i don't know
why.


cgd