Subject: Re: CVS commit: src/sys/kern
To: Jim Wise <jwise@draga.com>
From: Andrew Brown <atatat@atatdot.net>
List: tech-kern
Date: 04/09/2004 11:25:35
On Fri, Apr 09, 2004 at 11:09:11AM -0400, Jim Wise wrote:
>
>On Fri, 9 Apr 2004, Aymeric Vincent wrote:
>
>>
>>Bill Squier wrote:
>>
>>> Knowing briefly what the sysctl knob will change right there in the
>>> output of
>>> sysctl -a seems like a win to me-- and a much bigger one that knowing
>>> that my PIII supports the ABCDEFG-HOOKED-ON-PHONICS-WORKED-FOR-ME flag.
>>
>>This feature does not need to be in the kernel: it can be in sysctl(8).
>>I guess the semantics for a given sysctl variable never or rarely
>>changes, so that if a description exists, it will likely be correct.
>
>This sort of misses the point -- with the new sysctl glue, we support
>locally added and even third-party sysctl's. Even better these can be
>self-documenting.
>
>sysctl.8 is never going to know about these sysctl's. The sysctl
>command is not going to know about them. But the kernel does, and can
>easily return this documentation.
>
>So we're left with a choice -- we can have custom and third party
>sysctl's undocumented, or we can have them return a kernel string. And
>once we've made this choice, are you _really_ arguing that third party
>sysctl's should be documented by the kernel, but `native' ones should
>not?
>
>To save _five kilobytes_?!
let's be fair. when i said five, i was talking about the descriptions
i had done, not all of them (since they're not all done). that five
might also be four or it might be six.
and they're not all done. it might be as high as twelve or even
fifteen kilobytes when i'm done. it'll vary, too, from port to port,
and from kernel to kernel, as various options are included or not,
etc.
but it's still an option.
--
|-----< "CODE WARRIOR" >-----|
codewarrior@daemon.org * "ah! i see you have the internet
twofsonet@graffiti.com (Andrew Brown) that goes *ping*!"
werdna@squooshy.com * "information is power -- share the wealth."