Return to BSD News archive
Newsgroups: comp.unix.bsd
Path: sserve!manuel.anu.edu.au!munnari.oz.au!sgiblab!zaphod.mps.ohio-state.edu!caen!hellgate.utah.edu!fcom.cc.utah.edu!cs.weber.edu!terry
From: terry@cs.weber.edu (A Wizard of Earth C)
Subject: 386BSD PATCH KIT: WHAT WAS LEFT OUT AND WHY
Message-ID: <1992Oct6.181017.21831@fcom.cc.utah.edu>
Sender: news@fcom.cc.utah.edu
Reply-To: terry@icarus.weber.edu
Organization: Weber State University (Ogden, UT)
Date: Tue, 6 Oct 92 18:10:17 GMT
Lines: 84
I have received enough questions regarding what was left out of the
386BSD patch kit (mostly drivers) and why, that I felt it merits a post.
I was tempted to include the following (sorry about those without names):
o Garret Wollman's StarLAN card ethernet driver
o Francis Hitchens' DEC EtherWORKSTurbo Ethernet Controller DE200 driver
o Julian Elischer's SCSI system
o The "Norway" wd driver replacement
o The "non-interrupt" lpt driver.
o One of the Logitech Bus-Mouse drivers
I didn't. The drivers were omitted both because I didn't have permission
for the ethernet drivers and didn't seek it for the Logitech or lpt drivers.
I am of the opinion that they constitute full subsystems on their own, and
thus require approval.
I didn't pursue permission for the lpt driver and didn't pick a Logitech
driver for the simple reason that I had no way to make a qualatative
judgement as to which Logitech driver was the best or whether a
non-interrupt lpt driver was superior to the existing lpt driver.
I didn't get permission on the StarLAN or DEC drivers because both authors
felt that the code was still in testing (pre-release, although Garret's
code had been posted). I can definitely agree with their reasons.
The idea of patching new drivers into the system is kind of moot, anyway,
since there isn't a context free mechanism for modifying the files.i386
and Makefile.1386 files, nor a mechanism to change or coercively rebuild a
kernel frmo within a shell script.
I was tempted to put the replacement wd driver (what I have taken to
calling "the Norway code" -- sorry to the author whose name I can't remember
right now; I don't have the code in front of me) into a patch; it is
*vastly* superior and supports scanning for bad sectors and other important
operations stubbed out in the current one. Julian's SCSI code is in the
same boat.
The reason I didn't do the disk drivers was that I couldn't see how to
maintain proper revision levels (ie: patch levels) with a full replacement
for an existing file which was to be overwritten without releasing some
form of installation/package management system. These "patches" wouldn't
really increment a patch level; they would increment revision level, which
I couldn't cope with without putting out installation software as well as
patch software. It would be possible for me to do this, since I already
have a basic installation system protyped.
This doesn't mean that you can't install the drivers or disk systems as
the authors suggest *and* install the patches in the patch kit at the same
time; you can. You'll just have to do it by hand according to the authors
instructions.
I finally decided that I would leave the installation software and full
subsystem revisions to the Jolitz's 0.2 release, since Lynne was supposedly
working on new installation tools, Chris Torek on configuration tools,
and both Jolitz's on the 0.2 release, which I suspect will include the
SCSI and wd systems, and probably the drivers as well. (Lynne, Chris,
Bill: correct me here if I'm wrong, and I'll do something about it).
This incidently leaves me free to work on what I consider more interesting
projects by having the patch kit out the door, since that ate a lot of the
time I had budgeted to 386BSD. The patch kit is basically a code integration
tool, and in my mind, fairly unexciting. I did it because I was frustrated
three times in one hour trying to install conflicting patches. Hopefully
my other projects will useful enough that you'll agree "dropping the ball
on the drivers" was warranted.
I hope this answers the majority of questions I have received, since I just
have too much mail on these topics to answer them all personally. (I know,
I know; I asked for it 8-)).
Terry Lambert
terry@icarus.weber.edu
terry_lambert@novell.com
---
Any opinions in this posting are my own and not those of my present
or previous employers.
--
-------------------------------------------------------------------------------
"I have an 8 user poetic license" - me
Get the 386bsd FAQ from agate.berkeley.edu:/pub/386BSD/386bsd-0.1/unofficial
-------------------------------------------------------------------------------