Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 13 Mar 1996 17:54:13 +1000
From:      Stephen McKay <syssgm@devetir.qld.gov.au>
To:        Jon Loeliger <jdl@jdl.com>
Cc:        freebsd-chat@freebsd.org, syssgm@devetir.qld.gov.au
Subject:   Re: historical note..
Message-ID:  <199603130754.RAA20124@orion.devetir.qld.gov.au>

next in thread | raw e-mail | index | archive | help
Jon Loeliger <jdl@jdl.com> wrote:

>So, like "Jonathan M. Bresler" was saying to me just the other day:
>> Julian Elischer wrote:
>> > 
>> > I just found this in the BSD4.3 kernel I'm trying to debug..
>> > it's from vmparam.h..
>> 
>> 	HEY! somebody took this out fo the FreeBSD sources!  not fair.
>> 
>> 	i want it back.  please submit a diff.		;^)
>
>I was curious too.  So I fired off a piece of mail to
>our eternally good buddy Keith.  With all due respect,
>I forward his private mail to me.  He says:
>
>    To:      jdl@jdl.com
>    From:    Keith Bostic
>    Subject: Re: Historical note...
>    Date:    Mon, 11 Mar 1996 15:58:45 EST
>    ----------------------------------------------------------------
>    > [ You don't just happen to know who wrote this, do you? :-)
>    >   But, to be a '81 era comment, it's gotta be more like 4.1,
>    >   (or 2.9?) or earlier, right?  -- jdl ]
>
>    Bill Joy checked in the SCCS delta, presumably he wrote the text.
>    And, yes, this comment disappeared before the 4.2 release, so it's
>    not a 4.3BSD kernel.
>
>    --keith
>
>Oh wow!

For contrast, here is the appropriate bit from 386BSD 0.1 in
/sys/sys.386bsd/i386/include/vmparam.h (yes, I keep some weird old cruft):

/*
 * Just for fun: current memory prices are 4600$ a megabyte on VAX (4/22/81),
 * so we loan each swapped in process memory worth 100$, or just admit
 * that we don't consider it worthwhile and swap it out to disk which costs
 * $30/mb or about $0.75.
 * { wfj 6/16/89: Retail AT memory expansion $800/megabyte, loan of $17
 *   on disk costing $7/mb or $0.18 (in memory still 100:1 in cost!) }
 */
#define SAFERSS		8		/* nominal ``small'' resident set size
					   protected against replacement */

Bill got this from net/2, didn't he?  Net/2 was post-4.3, true?  What I'm
building to is the near-to-impossible suggestion that Keith might be wrong.

Still, that's not my main point.  I find this sort of buried historical
trivia fascinating, and would love to have it re-inserted, together with
a 1996 data point.  That is, of course, if we can find an appropriate bit
of code to attach it to, since SAFERSS no longer exists.

Stephen.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199603130754.RAA20124>