Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!news.Hawaii.Edu!ames!elroy.jpl.nasa.gov!usc!cs.utexas.edu!uunet!vtserf.cc.vt.edu!csugrad.cs.vt.edu!csugrad.cs.vt.edu!not-for-mail
From: briggs@csugrad.cs.vt.edu (Allen `Xext' Briggs)
Newsgroups: comp.os.386bsd.misc
Subject: Re: So you say you want an interim release of 386bsd? (What to do?)
Date: 1 May 1993 21:55:57 -0400
Organization: Virginia Tech Computer Science Dept, Blacksburg, VA
Lines: 63
Message-ID: <1rv9nd$s6f@csugrad.cs.vt.edu>
References: <2109@hcshh.hcs.de> <1r8qnp$1dc@csugrad.cs.vt.edu> <DERAADT.93Apr24135253@newt.fsa.ca>
NNTP-Posting-Host: csugrad.cs.vt.edu
In article <DERAADT.93Apr24135253@newt.fsa.ca> deraadt@fsa.ca (Theo de Raadt) writes:
>In article <1r8qnp$1dc@csugrad.cs.vt.edu> briggs@csugrad.cs.vt.edu (Allen `Xext' Briggs) writes:
>> The Jolitz' appear to be sticking to the CSRG release system.
>
>This is irrelevant, misleading, etc. The CSRG works cooperatively with
>many people, not in a vacuum, but the Jolitz' obviously want to.
It's not obvious to me. From what I've seen, they haven't been working
in any more of a vacuum than anyone swamped with work. They're working
hard on both "386BSD" 0.2 and "regular" jobs. Just about every post
from Lynne says that they're too busy to keep up with usenet news.
Working away from usenet does not constitute working in a vacuum.
>> There are some good reasons for revamping the kernel interfaces now as
>> opposed to five years from now.
>
>I'm a kernel person. I don't see any reason to revamp stuff.
Oh, really? Have you tried being a kernel person on multiprocessor
machines? Haven't you seen problems with the current kernel device
driver support? Have you considered portability to other platforms?
How about dynamically loadable device drivers? I'd much rather have
a kernel that was designed for them rather than kludging them in.
>> 0.2 is a somewhat different entity in that many of its changes sound
>> like they will not be very compatible with 0.1. I beleive Chris
>> mentioned in his announcement that NetBSD would pick pieces from 0.2
>> after it's available. Following the assumptions in the previous
>> paragraphs, these will bring the 0.1+pk and NetBSD paths converging
>> with 0._3_ (stable 0.2) when that's available. The process then starts
>> anew as 0.4 starts exploring new ground.
>
>What a wonderful plan. It gives people a useable and stable operating
>386BSD operating system around the turn of the century.
You're welcome to your opinions. What makes you think it'll take that
long?
>A very rich system inconsistant with many of the ways that standard BSD
>unix works, it appears.
What version of BSD are you considering to be "standard?" Is not one of
BSD's features that it evolves to meet the computing community's needs,
not the marketing wizard's?
It would be useful to have more net.input into what features are
important to work on, and there is a lot of development talent on
the net that could be applied toward achieving those features.
This seems to be where NetBSD comes in--it's much more net.driven,
and has a host of people working on it, directly or indirectly.
BSD 0.2 may suffer from the lack of net.input. NetBSD may suffer from
too much. Both have advantages and disadvantages. We'll see better
what those are when 0.2 is released. Wasn't it supposed to be around
March of this year???
-allen
--
Allen Briggs \ Swift has sailed into his rest;/Savage indignation
briggs@csugrad.cs.vt.edu | there/Cannot lacerate his breast./Imitate him
- end / if you dare,/World-besotted traveller; he/
killing - / Served human liberty. -- W.B.Yeats, 1931