NetNews Usenet Archive 1993 #1/NN_1993_1.iso/spool/comp/os/linux/23131

Internet Message Format | 1993-01-07 | 5.4 KB

Xref: sparky comp.os.linux:23131 comp.unix.bsd:11015
Path: sparky!uunet!zaphod.mps.ohio-state.edu!darwin.sura.net!spool.mu.edu!agate!soda.berkeley.edu!wjolitz
From: [e-mail] (William F. Jolitz)
Newsgroups: comp.os.linux,comp.unix.bsd
Subject: Re: 386 BSD
Date: 7 Jan 1993 22:56:21 GMT
Organization: U.C. Berkeley, CS Undergraduate Association
Lines: 98
Message-ID: <[e-mail]>
References: <[e-mail]> <[e-mail]> <[e-mail]>
NNTP-Posting-Host: soda.berkeley.edu



In article <[e-mail]> [e-mail] (Joseph R. Pannon) writes:
>In article <[e-mail]> [e-mail] (Tuomas J Lukka) writes:
>
>One of the reasons that originally steered me toward Linux instead of
>386BSD was the latter's low version number (0.1) compared to Linux (0.97
>at the time).  From what I hear from you guys, that big difference may not be
>justified by the features offered by both.  What may be behind Jolitz's
>thinking by assigning such an early version number to it?  What does he
>think he should have in addition to existing features to qualify for
>version 1.0?
>
>Any ideas?
>
>Joe Pannon

There appears to be some confusion as to how to view 386BSD and
the work associated with it. While BSD goals and procedures are
well-known among the educational/research sites which have formed
the core support for BSD releases over the last 12 years, they
may not be familiar (rightly) to the broader audiences now reached
by 386BSD and Linux.

First of all, 386BSD is intended for research and educational purposes.
There are a number of email Special Interest Groups (SIGs) which are
chartered to do not only short-term work (such as enhancements to
drivers, kernel fixes, new drivers and the like) but also examine
some longer-term items, such as unicode implemention, new approaches
to driver design, very high-speed networking, multithreading, and
so forth. These items are slow and tedious to do correctly, but very
rewarding in the long-term. Some of these items which are in-progress
even now will not appear until 1.0 and later, but they must be designed
now, if they are ever to be a part of the future.

For immediate work and test, the /unofficial directory was established
as a base of operation for new work to be tested/used as needed. Hence,
new work to be included in future releases is thoroughly tested before
release engineering by the 386BSD user community first.  In addition,
it is hoped that the call for votes (announced yesterday in news.groups)
will be favorable for the establishment of the 386bsd newsgroups, thus
channeling people to appropriate people for problems, solutions, new work,
announcements, and so forth, and eliminating the confusion which
currently and understandably exists.

For example, with respect to shared libraries, just as with Linux,
there already exist several shared library implementations
which work with 386BSD and are available for test/use. These versions
were announced a month ago in comp.unix.bsd, and are among the many
new items done by the 386BSD user community and made available to any
interested party.

There is also a research group which is attempting to complete
a more extensive and difficult shared library implementation which
will have serious ramifications on other areas of the system.

Should anyone wish shared libraries now, one can get a version which
works and is quite suitable for most purposes. However, that is
not the end of the matter for 386BSD.

Many of the people who work on 386BSD do so in order to continue the
tradition of innovation which is the hallmark of BSD over the decade.
BSD has been the basis for much exciting new work which has been incorporated
into research and commercial projects, and we hope to encourage
challanging new ideas to create a future, and not just accommodate the present.

Regarding 386BSD version numbering. All versions of 386BSD follow
BSD release numbering guidelines. Release 0.0 was the base 386BSD
release. Generally, even numbered releases (i.e. 2.8BSD, 4.0BSD, 4.2BSD
and 386BSD 0.2) incorporate new research or work. As a consequence,
the release may not be as robust as desired, since establishment of
new paradigms can often perturb other areas of the system. Odd numbered
releases (2.9BSD, 4.1BSD, 4.3BSD in it's various incarnations,
and 386BSD 0.1) on the other hand are generally more comprehensive
and fixed versions of the previous releases. However, as more "things"
are added to these releases, the paradigms and implementations tend
to sag under the weight the additional items, thus requiring redesign
or new design of the system.

Each release of 386BSD has a planned area of technology to be addressed.
This will take some time. But we hope that overall the correct
design decisions will be the basis for the next generation of
technology as well, and not just another "unix" system. This is
a very ambitious goal, but then again, BSD has always been a
very ambitious (and audacious) program from the beginning.

I hope this clarifies matters. I urge you to vote to establish the
386bsd newsgroups, so we can better meet people's needs and interest.

We are always open to feedback from all users of both 386BSD and Linux,
in the hopes that both groups can encourage creativity and exploration
of new ideas.

Thank you for your time.

Lynne Greer Jolitz
386BSD Development
[e-mail]