On Sat, 22 Jun 2002, Dan Nelson wrote:

>In the last episode (Jun 22), Mike Silbersack said:
>> On Sat, 22 Jun 2002, Lucky Green wrote:
>> > Let me turn my original inquiry into an offer: I volunteer to write
>> > the section for the Handbook or other documentation detailing the
>> > various CPU options in LINT if somebody who fully understands what
>> > these options do is willing to spend 30 minutes on the phone with
>> > me answering questions about the options.
>>
>> Despite your enthusiasm, it's still a rather pointless exercise.  To
>> make explaining the cpu options worthwhile, you must show that only
>> specifying I686 is sufficiently more optimal than specifying
>> I686/I586/I486/I386.
>
>I think he's referring to the flotilla of CPU feature options, mainly
>aimed at AMD and old Cyrix processors.
[snip]

I would argue that any effort put toward documenting this is better
spent documenting something else.  That particular "flotilla" of options
relates entirely to a group of rather old, rather slow, and rather rare
processors.  The incidence of their use or necessity in the general
FreeBSD user base is likely to be quite small.  If you're running on
asome ancient bastard child of Cyrix processor then chances are you're
not running a performance critical application.  At that level I say to
anyone who tries to squeeze every last ounce of performance out of their
CPU, "buy a faster CPU or UTFSL".

Feel free to disagree and work on what interests you.

Brandon D. Valentine
-- 
http://www.geekpunk.net                         [EMAIL PROTECTED]
++[>++++++<-]>[<++++++>-]<.>++++[>+++++<-]>[<+++++>-]<+.+++++++..++
+.>>+++++[<++++++>-]<++.<<+++++++++++++++.>.+++.------.--------.>+.


To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-hackers" in the body of the message

Reply via email to