BSD vs. Linux - Which is better?
Jordan
advice from some people who know about the different UNIX distributions. I
was just wondering why Linux has gotten so much attention. Is the Linux
kernel better than NetBSD and FreeBSD?
If I was to install one of these, or any other UNIX offspring OS which one
should I try? Also which of these three is better and why? Does one offer
more features or support more hardware?
Also, what is your favorite window manager?
Thanks,
Jordan
Tom Keats
>
> Also, what is your favorite window manager?
Open Look (olwm)
--
Why not just drop everything and go fishin'?
remove NO_SPAM. from address to reply
Martin Dieringer
Johan Rudholm
ze...@bawdycaste.org
: I've just started to learn about UNIX recently and I would like a little
: advice from some people who know about the different UNIX distributions. I
: was just wondering why Linux has gotten so much attention. Is the Linux
: kernel better than NetBSD and FreeBSD?
Yes and no.
: If I was to install one of these, or any other UNIX offspring OS which one
: should I try?
Yes.
: Also which of these three is better and why?
The forth one, because it's kEwL.
: Does one offer more features or support more hardware?
Yes.
: Also, what is your favorite window manager?
xterm
--
-Zenin (ze...@archive.rhps.org) Caffeine...for the mind.
Pizza......for the body.
Sushi......for the soul.
-- User Friendly
Jim Ross
Jordan wrote in message <7hl4m9$ig...@dragon.sk.sympatico.ca>...
>I've just started to learn about UNIX recently and I would like a little
>advice from some people who know about the different UNIX distributions. I
>was just wondering why Linux has gotten so much attention. Is the Linux
>kernel better than NetBSD and FreeBSD?
Linux is realy not based on Unix code.
Only Unix compatible.
Linux is free (cost).
Linux is free (liberated code)
Linux code if changed must be given back.
That has improved it fast.
Really the open nature and radically different way Linux has developed
is why people care about Linux I think.
This link explains that very well
http://www.tuxedo.org/~esr/writings/cathedral-bazaar/
Linux to *BSD is comparible I think, but different.
*BSD is based on Unix code I believe.
>
>If I was to install one of these, or any other UNIX offspring OS which one
>should I try? Also which of these three is better and why? Does one offer
>more features or support more hardware?
Linux is nice.
>
>Also, what is your favorite window manager?
>
> Thanks,
> Jordan
KWM with KDE.
Jim
patrick
<dale.s...@sk.sympatico.ca> wrote:
>I've just started to learn about UNIX recently and I would like a little
>advice from some people who know about the different UNIX distributions. I
>was just wondering why Linux has gotten so much attention. Is the Linux
>kernel better than NetBSD and FreeBSD?
>
>If I was to install one of these, or any other UNIX offspring OS which one
>should I try? Also which of these three is better and why? Does one offer
>more features or support more hardware?
>
>Also, what is your favorite window manager?
>
> Thanks,
> Jordan
>
>
You should try both.. linux and freebsd...
find out what's your flavor..
and use kde stuff voor your X enviroment.. it;s pretty cool...
patrick
Gergo Barany
>Linux code if changed must be given back.
Really? Please quote the section of the GPL that states such a thing.
>>Also, what is your favorite window manager?
I use fvwm2.
Gergo
--
Several years ago, some smart businessmen had an idea: Why not build a
big store where a do-it-yourselfer could get everything he needed at
reasonable prices? Then they decided, nah, the hell with that, let's
build a home center. And before long home centers were springing up
like crabgrass all over the United States.
-- Dave Barry, "The Taming of the Screw"
GU d- s:+ a--- C++>$ UL+++ P>++ L+++ E>++ W+ N++ o? K- w--- !O !M !V
PS+ PE+ Y+ PGP+ t* 5+ X- R>+ tv++ b+>+++ DI+ D+ G>++ e* h! !r !y+
Robert Heine
Jordan <dale.s...@sk.sympatico.ca> schrieb in im Newsbeitrag:
7hl4m9$ig...@dragon.sk.sympatico.ca...
[...]
> Also, what is your favorite window manager?
AfterStep
MfG
Robert Heine
Martin Dieringer
> In article <877lq9a...@ThinkPad.katz.de>, Martin Dieringer wrote:
> >"Jordan" <dale.s...@sk.sympatico.ca> writes:
> >
> >> Also, what is your favorite window manager?
> >
> >wmx!
>
> http://www.wmx.com/
>
> //Johan
never mind... it's in the ports anyway
martin
Christopher Uy
>I've just started to learn about UNIX recently and I would like a little
>advice from some people who know about the different UNIX distributions. I
>was just wondering why Linux has gotten so much attention. Is the Linux
>kernel better than NetBSD and FreeBSD?
>
>If I was to install one of these, or any other UNIX offspring OS which one
>should I try? Also which of these three is better and why? Does one offer
>more features or support more hardware?
>
As they're both free, why not try them both and see which one you like
for yourself?
>Also, what is your favorite window manager?
>
WindowMaker. :)
- chris
--
(O)
<M
o <M Christopher Uy
/| ...... /:M\------------------------------------------------,,,,,,
(O)[]XXXXXX[]I:K+}=====<{H}>================================------------>
\| ^^^^^^ \:W/------------------------------------------------''''''
o <W c...@chivalry.net
<W http://www.chivalry.net/
(O)
Remove the XYZ from my email address to reply.
Juergen Heinzl
>>advice from some people who know about the different UNIX distributions. I
>>was just wondering why Linux has gotten so much attention. Is the Linux
>>kernel better than NetBSD and FreeBSD?
>
>Only Unix compatible.
>Linux is free (cost).
>Linux is free (liberated code)
>Really the open nature and radically different way Linux has developed
>is why people care about Linux I think.
>This link explains that very well
>http://www.tuxedo.org/~esr/writings/cathedral-bazaar/
>
>Linux to *BSD is comparible I think, but different.
>*BSD is based on Unix code I believe.
No.
Cheers,
Juergen
--
\ Real name : Jürgen Heinzl \ no flames /
\ EMail Private : jue...@monocerus.demon.co.uk \ send money instead /
Vernon
<dale.s...@sk.sympatico.ca> wrote:
>I've just started to learn about UNIX recently and I would like a little
>advice from some people who know about the different UNIX distributions. I
>was just wondering why Linux has gotten so much attention. Is the Linux
>kernel better than NetBSD and FreeBSD?
>
>If I was to install one of these, or any other UNIX offspring OS which one
>should I try? Also which of these three is better and why? Does one offer
>more features or support more hardware?
>
>Also, what is your favorite window manager?
>
> Thanks,
> Jordan
>
>
FWIW....
I was asking similar questions a few months back on a Linux list. I
was curious about the opinions of folks who had experience with both
Debian Linux and FreeBSD. There was no criticism aimed at FreeBSD and
responders had high regard for FreeBSD. The bottom line seems to come
down to personal preference.
At the time I had more experience with Linux, but went ahead and tried
FreeBSD. I was pretty impressed and have since all but forgotten
about Linux. The ports collection smooths a lot of the bumps out of
the road for a new user. I have also noted a few posts as of late on
the FreeBSD lists from sysadmin types transitioning over to FreeBSD
from the Linux camp, citing frustration with recent Linux kernel
"games".
The one area that I personally feel Linux may have an edge is support.
The Debian lists, for example, are very helpful and supportive. By
contrast, I feel that the FreeBSD lists have been some of the least
helpful I'v experienced. Not to start a flame war or offend anyone,
but this is in fact my personal experience, limited though it may be.
Posts I've made to lists like MYSQL, Debian, PHP, etc. generally yield
several helpful and informative responses. Unfortnately, this has not
been the case with FreeBSD, which can be very frustrating.
As far as Linux flavors go, I would look hard at Debian or SuSE.
(Slackware if you enjoy hacking.) Although it's quite popular, I
personally am not particularly fond of Red Hat.
So there's my 2 cents worth...
Regards, Vernon
Chris Wolfe
"You must cause any work that you distribute or publish, that in whole
or in part contains or is derived from the Program or any part thereof,
to be licensed as a whole at no charge to all third parties under the
terms of this License."
The code is not given back to the author, it is given back to the
community.
Chris
Gergo Barany wrote:
>
> In article <jwt%2.1330$Ce.3...@news.itd.umich.edu>, Jim Ross wrote:
> >Linux code if changed must be given back.
>
> Really? Please quote the section of the GPL that states such a thing.
>
> >>Also, what is your favorite window manager?
>
> I use fvwm2.
>
> Gergo
James Carlson
> In article <877lq9a...@ThinkPad.katz.de>, Martin Dieringer wrote:
> >"Jordan" <dale.s...@sk.sympatico.ca> writes:
> >
> >
>
> http://www.wmx.com/
I could be wrong about this, but I don't think Waste Management,
Inc. is going to sell you a window manager.
Maybe you mean:
http://www.all-day-breakfast.com/wmx/
--
James Carlson, Software Architect <car...@ibnets.com>
IronBridge Networks / 55 Hayden Avenue 71.246W Vox: +1 781 372 8132
Lexington MA 02421-7996 / USA 42.423N Fax: +1 781 372 8090
"PPP Design and Debugging" --- http://people.ne.mediaone.net/carlson/ppp
Gergo Barany
>I think he was referring to section 2. b)
>
>"You must cause any work that you distribute or publish, that in whole
>or in part contains or is derived from the Program or any part thereof,
>to be licensed as a whole at no charge to all third parties under the
>terms of this License."
>
>The code is not given back to the author, it is given back to the
>community.
The point I was trying to make is that the GPL states that *if* you give
back to the community, the product has to be under the GPL, but you're
not *required* to give back, you can keep everything for yourself, so
his statement that (paraphrased) "Linux code must be given back" is not
correct.
Gergo
--
An idea is an eye given by God for the seeing of God. Some of these
eyes we cannot bear to look out of, we blind them as quickly as
possible.
-- Russell Hoban, "Pilgermann"
Dave Barr
Jim Ross <jkt...@cw-f1.umd.umich.edu> wrote:
>Linux is realy not based on Unix code.
>Only Unix compatible.
What makes code "Unix code"? What makes it better than "non-Unix" code?
In addition to making little sense, your statement makes little difference.
It's also quite wrong.
Yes, the *BSD code had its start with Net/2 and other code which came
out of the Berkeley UNIX Software Development group. However it
contains none of the commercial (proprietary) "UNIX" code from AT&T.
(Remember the 1-800-ITS-UNIX phone number BSD Inc had? Remember
The Lawsuit?)
>Linux is free (cost).
>Linux is free (liberated code)
True, more or less. Not all of the stuff in a Linux distribution
is under a GNU copyleft-like license. (XFree86 is the biggest example
of stuff that is not)
>Linux code if changed must be given back.
No. Copylefted code if changed and distributed in binary form must be
published. There's nothing to prevent anyone from modifying a
copylefted software and using it for your own use without distributing
changes.
>Linux to *BSD is comparible I think, but different.
Why are you posting if you just "think" you know the difference?
>*BSD is based on Unix code I believe.
See above.
--Dave
--
http://www.cis.ohio-state.edu/~barr/
ba...@cis.ohio-state.edu
ze...@bawdycaste.org
>snip<
: The one area that I personally feel Linux may have an edge is support. The
: Debian lists, for example, are very helpful and supportive. By contrast,
: I feel that the FreeBSD lists have been some of the least helpful I'v
: experienced.
By "lists" do you refer to the FreeBSD USENET newsgroups or the
mailing lists?
FreeBSD is "officially supported" through its mailing lists only and
so one should not expect good or any feedback from the USENET
newsgroups. Please don't judge FreeBSD support by what you'll find
in USENET.
Christopher B. Browne
>In article <jwt%2.1330$Ce.3...@news.itd.umich.edu>,
>Jim Ross <jkt...@cw-f1.umd.umich.edu> wrote:
>>Linux is realy not based on Unix code.
>>Only Unix compatible.
>
>What makes code "Unix code"? What makes it better than "non-Unix" code?
>In addition to making little sense, your statement makes little difference.
>It's also quite wrong.
>
>Yes, the *BSD code had its start with Net/2 and other code which came
>out of the Berkeley UNIX Software Development group. However it
>contains none of the commercial (proprietary) "UNIX" code from AT&T.
>(Remember the 1-800-ITS-UNIX phone number BSD Inc had? Remember
>The Lawsuit?)
However, the BSD "community" most definitely takes the stance that the
BSD-based systems are "truly UNIX-like" whereas Linux isn't, the reasoning
seeming to be that the BSDs represent a code base written by people that
built "real UNIX," whereas Linux certainly doesn't.
>>Linux is free (cost).
>>Linux is free (liberated code)
>
>True, more or less. Not all of the stuff in a Linux distribution
>is under a GNU copyleft-like license. (XFree86 is the biggest example
>of stuff that is not)
Linux as kernel uses GPL. LIBC on top of that uses GPL, as does GCC and
fileutils.
A common "BSD position" seems to be that software that goes much beyond that
isn't really part of "the system." Which would go along with the notion
that "a Linux system" substantially *is* GPLed, whilst there may be added
components in a distribution that aren't.
The crux of the matter is the somewhat philosophical issue of where "Linux
system" ends. It's been in the interests of many to define it as broadly as
possible...
>>Linux code if changed must be given back.
>
>No. Copylefted code if changed and distributed in binary form must be
>published. There's nothing to prevent anyone from modifying a
>copylefted software and using it for your own use without distributing
>changes.
>
>>Linux to *BSD is comparible I think, but different.
>
>Why are you posting if you just "think" you know the difference?
Most GNU software can run on both Linux and *BSD. The same is probably true
for BSD-licensed software.
>>*BSD is based on Unix code I believe.
>
>See above.
The connection is that people that built the BSD code base did development
on "true UNIXes." Which establishes that similarities/parallels are likely
to be closer for *BSD than for Linux, whose developers have been largely
disconnected from *formal* development work on *officially branded* UNIXes.
There is a valid stance that *BSD "inherits some UNIXness" that Linux
doesn't; there is an equally valid stance that neither are branded UNIXes,
and thus are both equally "not UNIX." There is very little value to pursuing
those stances; they establish little of ongoing value, except ample
opportunities for people to hurl abuse at one another...
--
Those who do not understand Unix are condemned to reinvent it, poorly.
-- Henry Spencer <http://www.hex.net/~cbbrowne/lsf.html>
cbbr...@hex.net - "What have you contributed to free software today?..."
John S. Dyson
Chris Wolfe <cwo...@globetrotter.qc.ca> writes:
> I think he was referring to section 2. b)
>
> "You must cause any work that you distribute or publish, that in whole
> or in part contains or is derived from the Program or any part thereof,
> to be licensed as a whole at no charge to all third parties under the
> terms of this License."
>
> The code is not given back to the author, it is given back to the
> community.
>
the user that you provide binaries. That user can destroy your
profitability by distributing the source to your competitors.
--
John | Never try to teach a pig to sing,
dy...@iquest.net | it makes one look stupid
jdy...@nc.com | and it irritates the pig.
Jim Ross
>In article <jwt%2.1330$Ce.3...@news.itd.umich.edu>, Jim Ross wrote:
>>
>>Jordan wrote in message <7hl4m9$ig...@dragon.sk.sympatico.ca>...
>>>advice from some people who know about the different UNIX distributions. I
>>>was just wondering why Linux has gotten so much attention. Is the Linux
>>>kernel better than NetBSD and FreeBSD?
>>
>>Only Unix compatible.
>>Linux is free (liberated code)
>>Really the open nature and radically different way Linux has developed
>>is why people care about Linux I think.
>>This link explains that very well
>>http://www.tuxedo.org/~esr/writings/cathedral-bazaar/
>>
>
>
>Cheers,
>Juergen
Thanks for the correction.
Unix design sounds closer to what I should have said.
Jim
Jim Ross
>>Linux is realy not based on Unix code.
>>Only Unix compatible.
>
What makes it better than "non-Unix" code?
It doesn't. I didn't say it is. It is just "different".
>In addition to making little sense, your statement makes little difference.
Just your opinion. If fact judging from what you write next you understand
perfectly.
As a break from this discussion.
If the original poster is reading this.
I recommend you buy both and try them.
Also surf to http://www.deja.com and type in BSD vs Linux.
You will get tons of info on opinions.
>It's also quite wrong.
>
>Yes, the *BSD code had its start with Net/2 and other code which came
>out of the Berkeley UNIX Software Development group. However it
>contains none of the commercial (proprietary) "UNIX" code from AT&T.
>(Remember the 1-800-ITS-UNIX phone number BSD Inc had? Remember
>The Lawsuit?)
OK. I read that BSD is faster on I/O, disk operations.
I think that is outweighed by the extra momentum behind Linux and
how it is more accepted.
>
>>Linux is free (cost).
>>Linux is free (liberated code)
>
>True, more or less. Not all of the stuff in a Linux distribution
>is under a GNU copyleft-like license. (XFree86 is the biggest example
>of stuff that is not)
When I said Linux I meant Linux proper. The kernel. It is GPL.
I didn't say Redhat.
>
>>Linux code if changed must be given back.
>
>No. Copylefted code if changed and distributed in binary form must be
>published. There's nothing to prevent anyone from modifying a
>copylefted software and using it for your own use without distributing
>changes.
You're right.
>
>>Linux to *BSD is comparible I think, but different.
>
>Why are you posting if you just "think" you know the difference?
I know more than the guy asking. It is all I need to know.
In reality it doesn't really matter which is better does it?
Apples to oranges. Does it work for the guy asking which is better or not.
He should try both. It is opinion mostly anyway.
>
>>*BSD is based on Unix code I believe.
>
>See above.
You go Dave.
It's your birthday.
Jim
Jim Ross
>In article <jwt%2.1330$Ce.3...@news.itd.umich.edu>, Jim Ross wrote:
>>Linux code if changed must be given back.
>
You're right.
For private use you don't.
If you redistribute changed Linux code you do.
http://www.fsf.org/copyleft/gpl.html
Jim
>
>>>Also, what is your favorite window manager?
>
>I use fvwm2.
>
>Gergo
>
>--
>Several years ago, some smart businessmen had an idea: Why not build a
>big store where a do-it-yourselfer could get everything he needed at
>reasonable prices? Then they decided, nah, the hell with that, let's
>build a home center. And before long home centers were springing up
>like crabgrass all over the United States.
> -- Dave Barry, "The Taming of the Screw"
>
Jim Ross
>In article <373EF6EA...@globetrotter.qc.ca>,
> Chris Wolfe <cwo...@globetrotter.qc.ca> writes:
>> I think he was referring to section 2. b)
>>
>> "You must cause any work that you distribute or publish, that in whole
>> or in part contains or is derived from the Program or any part thereof,
>> to be licensed as a whole at no charge to all third parties under the
>> terms of this License."
>>
>> The code is not given back to the author, it is given back to the
>> community.
>>
>The code is not necessarily given back to the community, but given to
>the user that you provide binaries. That user can destroy your
>profitability by distributing the source to your competitors.
Two points to reply to what you just wrote here.
1. Users and business are used to MS style products where they don't have
access to source and can't change it. They are used to that. So I don't
think the are going to think of doing that first. Also most people shouldn't
be mucking around with the kernel much anyway. It is a very low level piece
of software. User level stuff and even some level are what should be written.
We already have Linux. If that isn't what you need then why use Linux. It
doesn't fit your need.
2. Writing source once and selling binaries sounds like an outdated concept.
Strategic advance is being able to keep turning out new source, building
solutions, and selling support wrapped around that.
A side thought. Even if you gave that user the changes you made to the kernel
why would that destroy your profitability? Surely you are building a solution
with a program. That program/solution you keep the code to.
I'm sure you wouldn't build and base your profitability on a simple kernel
tweak which you would turn over source to. Come on. You only have to give away
Linux changes, not anything else. Plus it isn't really the problem of giving
back Linux changes people make it out to be. Linux is GPL. You can just
change code and keep it. Everyone should know going in Linux is not yours but
GPL. Everyone owns it and end of story.
Jim
Chris Wolfe
Chris
"John S. Dyson" wrote:
>
> In article <373EF6EA...@globetrotter.qc.ca>,
> Chris Wolfe <cwo...@globetrotter.qc.ca> writes:
> > I think he was referring to section 2. b)
> >
> > "You must cause any work that you distribute or publish, that in whole
> > or in part contains or is derived from the Program or any part thereof,
> > to be licensed as a whole at no charge to all third parties under the
> > terms of this License."
> >
> > The code is not given back to the author, it is given back to the
> > community.
> >
> The code is not necessarily given back to the community, but given to
> the user that you provide binaries. That user can destroy your
> profitability by distributing the source to your competitors.
>
Vernon
Sorry if I wasn't clear, but yes, I was refererencing the FreeBSD
mailing lists: e.g. ports, questions, and stable. I don't mean to say
that these lists are bad, it just seems like there are lots of
questions that go unanswered compared to other lists I am subscribed
to.
Regards, Vernon
John S. Dyson
Chris Wolfe <cwo...@globetrotter.qc.ca> writes:
>
> I do not want to even think about starting that argument in here...
>
>> the user that you provide binaries. That user can destroy your
>> profitability by distributing the source to your competitors.
>>
license knows that:
1) You don't have to give the source away to a party that you haven't
given binaries to.
2) Your customer can easily put you on the hook by giving the sources
away, including to a competitor who might compete with you for
"support" services.
This certainly disadvantages the developer who adds significant value
to a piece of software under the GPL.
Richard Steiner
spake unto us, saying:
>This certainly disadvantages the developer who adds significant value
>to a piece of software under the GPL.
Assuming, of course, that they don't agree with the conditions laid out
in the GPL. Not all software developers disgree with the GPL's stated
conditions, though, as you so often seem to assume.
Don't you folks have better things to do than turn the technical groups
into advocacy groups? This anti-GPL whining gets old rather fast.
--
-Rich Steiner >>>---> rste...@visi.com >>>---> Bloomington, MN
OS/2 + Linux (Slackware+RedHat+SuSE) + FreeBSD + Solaris + BeOS +
WinNT4 + Win95 + PC/GEOS + MacOS + Executor = PC Hobbyist Heaven!
The universe is a spheroid region 705 meters in diameter
Chris Costello
> It is certainly closer to BSD than SYSV... It is LOTS AND LOTS
> more than POSIX -- please note NT complying with POSIX... If
> you had to compare them all, the difference between BSD and Linux
> is much smaller than between Linux and SYSVr4, or even Linux and
> many POSIX implementations. IMO, Linux is more like SYSVr3 + BSD
> networking + better VM... Hmmm... That almost makes it the same
> as BSD!!!
Better VM and networking than what now?
> > However, while I have been a Linux user for 5+ years and having using
> > FreeBSD for much less (only about 1 year), I am beginning to wonder why
> > Linux has become the "de facto free Unix" standard. Basically nearly
> > everything in Linux is a re-implementation of something that existed
> > in BSD already, suggesting that there is a great deal of wasted effort
> > going on here.
> Yes, exactly!!!
I agree here in part. Linux is clearly an interesting effort,
but I still believe that its best work is being done on the
desktop, because, as jedi noted, the work being done on the
desktop is much easier than making it perform as a high-end
server.
> > It makes one wonder: what if timing and luck had been slightly different,
> > and instead of Linux catching on, {Free|Open|Net}BSD had.
> > Or, if a Linux kernel was placed on to of the BSD toolbase, instead of
> > the GNU toolbase?
> It seems to me that Linux is more a creature of NIH (not invented here),
> and "I can do it better than the real professionals" attitude of many
> young folk. The bottom line is that Linux and BSD are asymtotically
> the same, with the difference being that Linux has the unfortunately
> restrictive license. Since BSD has been done without GPL (and adopting
> GPLed tools along the line), and Linux has been done with GPL, arguments
> that the GPL is necessary are quite specious.
The arguments are not specious, they are of absolute
incorrectness.
> GPL simply restricts the freedom of the developers who add innovation.
I'm tempted to add that to my tag file.
> --
> John | Never try to teach a pig to sing,
> dy...@iquest.net | it makes one look stupid
> jdy...@nc.com | and it irritates the pig.
--
Chris Costello
Congratulations! You are the one-millionth user to log into our system.
Peter Mutsaers
DK> Having tried both Linux (from .99 to 2.2.7) and FreeBSD (2.2,
DK> not 3.0) I'd have to say they are awfully similar in
Same here, only FreeBSD also for the past for years until today
(4.0 which is now -current).
DK> However, while I have been a Linux user for 5+ years and
DK> having using FreeBSD for much less (only about 1 year), I am
DK> beginning to wonder why Linux has become the "de facto free
DK> Unix" standard. Basically nearly everything in Linux is a
So did I after I started with FBSD several years ago.
DK> re-implementation of something that existed in BSD already,
DK> suggesting that there is a great deal of wasted effort going
DK> on here.
Also in some parts FBSD really outperforms Linux, especially disk I/O
and its general system impact. When I do a find / in the background on
FBSD, I hardly notice any interactive perfomance decay. On Linux it
kills interactive performance (also in 2.2.7 which I'm currently
running in Slackware 4.0beta).
DK> It makes one wonder: what if timing and luck had been slightly
DK> different, and instead of Linux catching on,
DK> {Free|Open|Net}BSD had. Or, if a Linux kernel was placed on
DK> to of the BSD toolbase, instead of the GNU toolbase?
*BSD was having some licensing troubles at the time Linux was taking
off. Also for a long time I just found Linux and its culture more
sympathetic which made me use Linux for the first 2 years. Only after
a while I noticed that FBSD really was better.
But: I think that current Linux popularity can be very advantegeous
for *BSD. It attracts many new people to UNIX in general, and many of
them find their way to BSD.
I'm still not too positive about Linux's future. The development is
quite chaotic, and some of the (commercial) software developers get
fed up with rapid change in non-compatible libraries and general chaos
and difference between distributions. I expect those to discover FBSD
too and to see its beaty and stability; then we'll see more and more
native FBSD software. In the meantime the Linux emulator does a good
job.
Note that all software necessary to get Linux to the desktop (GNOME,
KDE or whatever, and all the tools, filemanagers and stuff that come
with it) runs just fine on FBSD too, so once there is a large
installed base of Linux it'll be really easy to switch over to
FBSD. Now that UNIX at the desktop is still small and exotic, all
media attention goes to the biggest and most mythical one; but once it
gets normal and generally accepted I expect that others such as FBSD
get their fair share of attention too.
--
Peter Mutsaers | Abcoude (Utrecht), | Trust me, I know
p...@xs4all.nl | the Netherlands | what I'm doing.
---------------+---------------------+------------------
Powered by FreeBSD (-current). See http://www.freebsd.org
Dave Barr
Christopher B. Browne <cbbr...@hex.net> wrote:
>However, the BSD "community" most definitely takes the stance that the
>BSD-based systems are "truly UNIX-like" whereas Linux isn't, the reasoning
>seeming to be that the BSDs represent a code base written by people that
>built "real UNIX," whereas Linux certainly doesn't.
Sure, I just question highly the value of such discussions.
Linux has proven itself to be UNIX. To say that Linux is somehow less
of a UNIX because of its heritage is the true definition of elitism.
Sure, the fact that Linux is a reinvention has been a historical
stumbling block (for example in differences in the networking code, and
how it historically was way behind in NFS stability).
It seems to me to be more of a question of people having a too narrow
view of what "UNIX" means, and confusing code differences with functional
differences.
The measure of UNIX is how close it is to UNIX in action, not whose
parents it had.
>>>Linux to *BSD is comparible I think, but different.
>>
>>Why are you posting if you just "think" you know the difference?
>
>Most GNU software can run on both Linux and *BSD. The same is probably true
>for BSD-licensed software.
A total non-sequitor. Most GNU software can run on NT (with help), but
it's not UNIX.
>The connection is that people that built the BSD code base did development
>on "true UNIXes." Which establishes that similarities/parallels are likely
>to be closer for *BSD than for Linux, whose developers have been largely
>disconnected from *formal* development work on *officially branded* UNIXes.
Linux and BSD I believe are closer to each other than to
"offically branded" UNIXes like AIX or HP/UX.
Unfortunately it's easier to have an argument about how pure-blooded
Linux is or isn't compared to *BSD than it is to have a discussion
comparing their technical merits. The latter requires more knowledge.
People also forget the fact that BSD historically has been NOT UNIX,
but a parallel development track. From time to time there have been
spurts of cross fertilization between the CSRG and the AT&T folks, but
they've remained largely separate. Whole parts of the kernel have been
rewritten on both sides, diverging them even more. Linux is about as
different in functionality to *BSD as BSD is AT&T UNIX, even though
it shares no common code with either of them.
Dave Barr
Jim Ross <jkt...@cw-f1.umd.umich.edu> wrote:
>ba...@cis.ohio-state.edu (Dave Barr) wrote:
>>What makes code "Unix code"?
>From the original Unix code from way back.
Ah, but you forget (or maybe you don't know) that "the original Unix
code" from way back WAS NOT BSD!
John S. Dyson
rste...@visi.com (Richard Steiner) writes:
> Here in comp.unix.bsd.freebsd.misc, to...@y.dyson.net (John S. Dyson)
> spake unto us, saying:
>
>>This certainly disadvantages the developer who adds significant value
>>to a piece of software under the GPL.
>
> Assuming, of course, that they don't agree with the conditions laid out
> in the GPL. Not all software developers disgree with the GPL's stated
> conditions, though, as you so often seem to assume.
>
>
> Don't you folks have better things to do than turn the technical groups
> into advocacy groups? This anti-GPL whining gets old rather fast.
>
It is *very* important to keep reminding new users as to the
disadvantages of GPL. Until the defacto GPL advocacy stops,
then the other side has to be presented.
Note that I am not "whining", but providing another viewpoint. I
sometimes perceive others who don't agree with me as "whining", but
that is *sometimes* not accurate either. I suspect that the use
of that word "whining" in this case is due to a choice of the wrong
term. Perhaps the correct term is "warning", or "elucidating."
John S. Dyson
>> Chris Wolfe <cwo...@globetrotter.qc.ca> writes:
>>> I think he was referring to section 2. b)
>>>
>>> "You must cause any work that you distribute or publish, that in whole
>>> or in part contains or is derived from the Program or any part thereof,
>>> to be licensed as a whole at no charge to all third parties under the
>>> terms of this License."
>>>
>>> The code is not given back to the author, it is given back to the
>>> community.
>>>
>>the user that you provide binaries. That user can destroy your
>>profitability by distributing the source to your competitors.
>
> 1. Users and business are used to MS style products where they don't have
> access to source and can't change it. They are used to that. So I don't
>
(e.g. libraries.) Geesh, with the movement towards GPL'ing libraries,
it seems that MS and GPL world is on more and more even footing.
>
> Also most people shouldn't
> be mucking around with the kernel much anyway. It is a very low level piece
> of software. User level stuff and even some level are what should be written.
> We already have Linux. If that isn't what you need then why use Linux. It
> doesn't fit your need.
>
People shouldn't be 1) mucking around with kernels, 2) mucking around
with libraries, 3) mucking around with applications, 4) mucking around
with user interfaces, 5) mucking around with ANY software?
Note that you are making a value judgment that shows that you might
think that software in general shouldn't be "mucked" with? A kernel
is no different than any other piece of software, and not necessarily
the most complex piece. It isn't beyond improvement for some kind of
competitive advantage.
>
> 2. Writing source once and selling binaries sounds like an outdated concept.
>
That is a statement that is simply not true. By not allowing that, you
cause all enhancement for a sector to have to be funded by one customer.
(Or secondly, mis-sell support as access.)
>
> You can just
> change code and keep it. Everyone should know going in Linux is not yours but
> GPL. Everyone owns it and end of story.
>
In cases where everyone owns an expensive investment, there will be
stagnation. Linux will forever be a '80's monolithic kernel, which
is totally unacceptable. The sad thing is that the short term thinking
of a portion of the userbase is accepting Linux, mostly because of advocacy,
timing, and a short-term marginal benefit over commercial software.
Some companies are even becoming fearful of any kind of forward looking
investment, and will eventually become the mediocre that they chose
to be by fear. GPL is a kind of license that preys on fear and the
notion of software being "cheap" to get. Hey you programmers out there:
if the software is "cheap", you are going to be "cheap" also. When
you have families to support, and the economy normalizes, then you'll
start becoming more sensitive to these facts.
Geesh!!!
David Burgess
> In article <926903067....@news.itd.umich.edu>,
> Jim Ross <jkt...@cw-f1.umd.umich.edu> wrote:
> >ba...@cis.ohio-state.edu (Dave Barr) wrote:
> >>What makes code "Unix code"?
> >From the original Unix code from way back.
>
> Ah, but you forget (or maybe you don't know) that "the original Unix
> code" from way back WAS NOT BSD!
No, but the original BSD was "the original UNIX code" and for the most
part the current 'UNIX' code (whether BSD or not) is certainly no
longer. You needed a source code license from Bell Labs to provide the
couple of dozen files that were still copyrighted by Bell Labs/AT&T
until 1992 (or so).
BSD and 'modern UNIX' both trace their roots back to System III (IIRC).
From there, Bell Labs decided that putting the source code into
(effectively) the public domain was a good thing. They released this
code, without sufficient copyright (check out 'The History of the
Lawsuit' (if it exists someplace) for more information) to various
Universities and organizations around the globe.
UC - Berkeley (the B in BSD) started making some good modifications to
the code. Many of these were under contract to DARPA, others were parts
of various Masters and Ph.D work, and still others just kind of
happened. One by one, components of the system (UNIX) were replaced by
Berkeley code. Lots of these (like the networking code) were pulled
into the commercial UNIX systems and used as a basis for the current
UNIXen of today.
So, claiming that any UNIX today is the 'true UNIX' because it is based
on the original System III code is like claiming that you have a 500
year old ax, except that the handle's been replaced twelve times and the
head's been replaced twice.
At the end of 'The Lawsuit' there were only 24 files that were of
'questionable progenity' (none of the lawyers could easily identify
where they came from) and there were 'encumbered' file but there were
several files in AT&T UNIX that were definitely Copyright UC -
Berkeley. The failure of AT&T to properly attribute these files, for
the history majors in the crowd, was the leverage that was finally used
to settle the lawsuit in a favorable light for everyone.
Not that any of this actually makes any difference, by the way.
John S. Dyson
ba...@cis.ohio-state.edu (Dave Barr) writes:
> In article <slrn7juhuf....@knuth.brownes.org>,
> Christopher B. Browne <cbbr...@hex.net> wrote:
>>However, the BSD "community" most definitely takes the stance that the
>>BSD-based systems are "truly UNIX-like" whereas Linux isn't, the reasoning
>>seeming to be that the BSDs represent a code base written by people that
>>built "real UNIX," whereas Linux certainly doesn't.
>
> Sure, I just question highly the value of such discussions.
>
is a gratuitiously rewritten U**X clone, that has a MORE restrictive
license than the previously available versions. There were some minor
bumps in the BSD road, but those have been resolved for many years.
The technically superior (in one way or another) choice isn't always
the most successful. What the BSD folks probably lament is that a
better choice is being overwhelmed by a vast marketing engine. It
is very *easy* to be communist (in the specific economic sense), but
requires strength and belief in oneself to trust one's own innovative
skills. GPL often takes advantage of the fear that the "secrets" of
software development are hidden from the masses. Such secrets are NOT
hidden.
GPL is quite advantageous to those with the name recognition, and the
marketeers. However the value add developers in the middle are cut out
of the stream, not able to effectively *capitalize* on their specific
area of expertise (innovation.) Of course, such developers can lie
and sell their work (or access) as "support", but that seems to
be *dishonest*.
Brad Knowles
> The measure of UNIX is how close it is to UNIX in action, not whose
> parents it had.
I must respectfully disagree. Part of the long history of Unix is in
the proven code that has not been thrown out and reinvented. If you throw
everything out and start over, then no matter how similar to Unix you
might be, you simply are not (and cannot ever be) Unix.
You could certainly be POSIX (assuming you met all the criteria), but
you could never be Unix -- you don't have the decades of experience and
"burn-in" in that code which has not been thrown out.
No matter how similar in appearance and feel a reproduction might be
to the original, it simply is not the original. A reproduction can, in
some ways, be even better than the original (have you ever driven an
original E-type Jag, and then compared that to how nicely a more modern
car drives?), but it can never *BE* the original.
And there's nothing in the Universe that can change that.
Frankly, I don't think we should be discussing Linux in this context
at all. The developers of Linux should attempt to make it better than
Unix (and sufficiently source code compatible), and people should stop
wasting their time trying to make silly comparisons like this.
However, it is my personal opinion that, today, Linux still has a ways
to go before it has the necessary "fit and finish" to start seriously
replacing Unix in production use on servers.
Maybe it's adequate for use on personal desktop machines (what might
be referred to as a "workstation"), but it is not yet suitable for use as
a server. And even as a personal desktop machine, I believe that there
are other options available that I would consider more suitable than
Linux. But, that's just my personal opinion.
--
Brad Knowles <br...@shub-internet.org> <http://www.shub-internet.org/brad/>
<http://wwwkeys.pgp.net:11371/pks/lookup?op=get&search=0xE38CCEF1>
Are you looking for a news feed from a site in the Freenix Top 130?
If so, contact me via private e-mail for details.
Alexander Viro
David Burgess <bur...@mitre.org> wrote:
>No, but the original BSD was "the original UNIX code" and for the most
>part the current 'UNIX' code (whether BSD or not) is certainly no
>longer. You needed a source code license from Bell Labs to provide the
>couple of dozen files that were still copyrighted by Bell Labs/AT&T
>until 1992 (or so).
>
>BSD and 'modern UNIX' both trace their roots back to System III (IIRC).
YHRI. BSD predates System III.
Folks, what about retromoderation on such crosspostings? It's not a net.*,
indeed, but rule of 3 would stop those flamewars for good and I've *never*
seen a thread crossposted between c.u.b.f.* and c.o.l.* that would not turn
into the content-free flamewar (or should go to c.u.misc or a.f.c).
--
"You're one of those condescending Unix computer users!"
"Here's a nickel, kid. Get yourself a better computer" - Dilbert.
Peter da Silva
are using. Make your own decision.
--
In hoc signo hack, Peter da Silva <pe...@baileynm.com>
`-_-' Ar rug tú barróg ar do mhactíre inniu?
'U` "Twenty-two points, plus triple-word-score, plus fifty points for
using all my letters. Game's over. I'm outta here."
John S. Dyson
ba...@cis.ohio-state.edu (Dave Barr) writes:
> In article <926903067....@news.itd.umich.edu>,
> Jim Ross <jkt...@cw-f1.umd.umich.edu> wrote:
>>ba...@cis.ohio-state.edu (Dave Barr) wrote:
>>>What makes code "Unix code"?
>>From the original Unix code from way back.
>
> Ah, but you forget (or maybe you don't know) that "the original Unix
> code" from way back WAS NOT BSD!
>
from BSD in one way or another. The original Unix predates BSD, but
BSD predates (partially) the current Unix. Practically everything
predates Linux, and Linux is a re-coding of the previously created
research and innovation. (Sometimes better, sometimes worse, but
a restrictively licensed pseudo-clone of BSD anyway.)
David Konerding
>In article <7hp771$cds$1...@news.cis.ohio-state.edu>,
>The Unix that most people see as Unix is more than partially derived
>from BSD in one way or another. The original Unix predates BSD, but
>BSD predates (partially) the current Unix. Practically everything
>predates Linux, and Linux is a re-coding of the previously created
>research and innovation. (Sometimes better, sometimes worse, but
>a restrictively licensed pseudo-clone of BSD anyway.)
I would call it a "pseudo-clone of POSIX", rather than BSD, since
it was never written to be a clone of BSD. pseudo of course because there
is still a bit more work to be done to make Linux a truly first-class
Unix(clone) (in my case, there are some mmap implementation issues
that affect performance of my code).
Having tried both Linux (from .99 to 2.2.7) and FreeBSD (2.2, not 3.0)
I'd have to say they are awfully similar in functionality, there isn't really
a tremendous difference between the quality of performance (for example
NFS) or in the quality of the implementations (NFS, TCP, mm, paging, filesystem)
especially if one is talking about Linux 2.2, not 2.0.
However, while I have been a Linux user for 5+ years and having using
FreeBSD for much less (only about 1 year), I am beginning to wonder why
Linux has become the "de facto free Unix" standard. Basically nearly
everything in Linux is a re-implementation of something that existed
in BSD already, suggesting that there is a great deal of wasted effort
going on here.
It makes one wonder: what if timing and luck had been slightly different,
and instead of Linux catching on, {Free|Open|Net}BSD had.
Or, if a Linux kernel was placed on to of the BSD toolbase, instead of
the GNU toolbase?
--
--------------------------------------------------------------------------------
Email: d...@cgl.ucsf.edu David Konerding WWW: http://picasso.ucsf.edu/~dek
--------------------------------------------------------------------------------
Snail: Graduate Group in Biophysics
Medical Sciences 926, Box 0446
University of California
San Francisco, CA 94143
Joseph T. Adams
John S. Dyson (to...@y.dyson.net) wrote:
: Actually, there are large parts of the MS code that are source distributed.
: (e.g. libraries.)
What libraries? (MFC doesn't count because it's a wrapper around
Win32, which most assuredly is not available in source form.)
: Geesh, with the movement towards GPL'ing libraries,
: it seems that MS and GPL world is on more and more even footing.
I understand that you don't like the GPL license and that's fine. But
what would possess any seemingly bright human being to suggest that an
open-source model (albeit one you obviously don't like) can even be
compared to a source-not-available-at-any-price model like the one
Microsoft espouses?
: > Also most people shouldn't
: > be mucking around with the kernel much anyway. It is a very low level piece
: > of software. User level stuff and even some level are what should be written.
: > We already have Linux. If that isn't what you need then why use Linux. It
: > doesn't fit your need.
: >
: People shouldn't be 1) mucking around with kernels, 2) mucking around
: with libraries, 3) mucking around with applications, 4) mucking around
: with user interfaces, 5) mucking around with ANY software?
Isn't one of FreeBSD's selling points over Linux that there is a
relatively greater extent of central control - you can still do what
you want with your source, to an even greater extent than you could
under the GPL, but there is only one "official" release of FreeBSD?
: Note that you are making a value judgment that shows that you might
: think that software in general shouldn't be "mucked" with? A kernel
: is no different than any other piece of software, and not necessarily
: the most complex piece. It isn't beyond improvement for some kind of
: competitive advantage.
Agreed here.
: > 2. Writing source once and selling binaries sounds like an outdated concept.
: >
: That is a statement that is simply not true. By not allowing that, you
: cause all enhancement for a sector to have to be funded by one customer.
: (Or secondly, mis-sell support as access.)
:
: >
: > You can just
: > change code and keep it. Everyone should know going in Linux is not yours but
: > GPL. Everyone owns it and end of story.
: >
: In cases where everyone owns an expensive investment, there will be
: stagnation. Linux will forever be a '80's monolithic kernel, which
: is totally unacceptable.
It's perfectly acceptable now. Nothing drastically better exists yet,
at least not on PC-class hardware.
If, down the road, it proves not to be, then almost all free software
written for Linux, assuming that it isn't broken, will continue to be
able to be compiled and run on FreeBSD or any other POSIX-conforming
OS.
The same should be true of proprietary software as well, although if
you don't at least have access to source, you'll be at the mercy of
your vendor who may or may not choose to do so.
: The sad thing is that the short term thinking
: of a portion of the userbase is accepting Linux, mostly because of advocacy,
: timing, and a short-term marginal benefit over commercial software.
What people love about Linux, as far as I can see, is not that it
isn't FreeBSD. (I like FreeBSD too.) What they love about it is that
it is free - at least, more free than what they are used to - and that
because it is free in the sense of liberty, it's generally of much
higher quality.
The shift away from proprietary, mediocre, and highly vendor-bound
bloatware should benefit the *entire* free and open-source software
communities, and as far as I can see, it does.
: Some companies are even becoming fearful of any kind of forward looking
: investment, and will eventually become the mediocre that they chose
: to be by fear. GPL is a kind of license that preys on fear and the
: notion of software being "cheap" to get. Hey you programmers out there:
: if the software is "cheap", you are going to be "cheap" also. When
: you have families to support, and the economy normalizes, then you'll
: start becoming more sensitive to these facts.
Software intended for the mass market will always be cheap. The
question is not price, but liberty and quality.
The market will determine what programming and programmers are worth.
My guess is that crappy programmers will eventually be paid what they
are worth, which is very little, but good programmers will be paid
what they are worth, which is a hell of a lot. I see no reason to
believe that this will be different on the free software side on one
hand versus the proprietary crapware side on the other.
Joe
John S. Dyson
br...@shub-internet.org (Brad Knowles) writes:
> In article <7hp725$bp3$1...@news.cis.ohio-state.edu>, ba...@cis.ohio-state.edu
> (Dave Barr) wrote:
>
>> The measure of UNIX is how close it is to UNIX in action, not whose
>> parents it had.
>
> I must respectfully disagree. Part of the long history of Unix is in
> the proven code that has not been thrown out and reinvented. If you throw
> everything out and start over, then no matter how similar to Unix you
> might be, you simply are not (and cannot ever be) Unix.
>
paragraph is *exactly* one of my positions, and the fact that Linux is
a gratuitious reimplementation is a macro-level example of the
reimplementation that U**X developers have tried to avoid.
Peter da Silva
David Burgess <bur...@mitre.org> wrote:
>BSD and 'modern UNIX' both trace their roots back to System III (IIRC).
V7. Actually, SIII was Programmer's Work Bench (PWB) with some stuff pulled
in from V7 and a new tty interface. PWB was V6 plus the SCCS suite.
2BSD was based on V7 on the PDP-11.
3BSD and 4BSD were based on VAX ports of V7... I don't recall the exact
interrelationship between V6, V7, and Unix/32V, and the last direct
experience I had with them was 4.1C.
Dave Barr
>is a gratuitiously rewritten U**X clone, that has a MORE restrictive
>license than the previously available versions. There were some minor
>bumps in the BSD road, but those have been resolved for many years.
You forget your history lesson.
The *BSD code base was basically locked up in legal fights for *2 years*.
That's an incredibly long time -- just look at where Linux was 2 years ago.
That's not gratuitous rewriting. Linus *wanted* a good UNIX, and given
there was no available code base from which to work he had to start
from scratch.
By the time the 4.4 code base went public, Linux was already quite far
along.
David Burgess
David Konerding wrote:
>
> It makes one wonder: what if timing and luck had been slightly different,
> and instead of Linux catching on, {Free|Open|Net}BSD had.
> Or, if a Linux kernel was placed on to of the BSD toolbase, instead of
> the GNU toolbase?
Here is my own personal opinion on this:
None of the free Unix clones (and especially Linux) would have existed were it not
for the pioneering work that was done with Minix. There were two different camps
there, the 'pure minix' crowd, and the group that was supporting Linus and his new
kernel implementation. Early versions of Linux were 'minix compatible' to allow the
work that Andrew Tanenbaum had done to leverage the new kernel. At the same time,
in an apparently unrelated effort, 386BSD version 0.1 was released. Note that both
of these were (IMHO) driven by the same 'market demand' for a Unix-like system for
the PC.
Shortly after the release of 386BSD 0.1, UC Berkeley was slapped with a lawsuit
saying that they had released unspecified copyrighted and 'trade secret' materials.
This pretty effectively froze development of 386BSD. There were personality issues
as well, but those are better left to private E-Mail.
Linux, unencumbered by the negative press of the AT&T lawsuit and being led by a
relatively charismatic project manager, was able to catch up with the significant
lead that 386BSD had at that time. Part of their success was the adoption of as
many GNU tools as they could convert. By the time the lawsuit was settled, enough
people had decided which 'camp' they were going to be in that there was little else
to do.
That brings us up to the last couple of years. One of the things that has continued
to hinder the acceptance of the BSD systems is the perception that there were three
different BSD systems. Each had it's own strategic advantages, and are outwardly
different. So, given that, it seems inevitable that the people who were in a
position to decide which OS to support (Linux, or the three BSDs) would choose the
'single' OS.
So here we are today. The major BSD systems (FreeBSD, NetBSD, and OpenBSD) are all
available and are doing some pretty amazing stuff. The major Linux systems (SuSe,
Red Hat, Slackware, etc.) are all available and are doing some pretty amazing
stuff. Linux now has the same flamewars the BSD folks have about which BSD
'distribution' they are going to use. Unlike BSD which was fragmented on technical
issues, Linux is fragmented by vendor support issues.
All seven systems run Linux executables (whch is good for the software publishers).
All seven run on the PC, a few of them run on more than one platform. Since they
all work with Linux exectuables, it's pretty safe to say that they are all pretty
much the same.
So, I guess the bottom line on this is the opinion I promised you earlier.
The reason Linux is so much more popular than any of the BSD systems has nothing to
do with timing, luck, or technical superiority. AT&T attacked the distribution of
386BSD since they feared the loss of it's market dominance in the Unix market, and
acted to protect it by filing a lawsuit against UC Berkeley. They saw the quality
of the work done and recognized a threat. Once this was done, the adoption of Linux
as an alternative operating system was practically assured. Whether they intended
it or not, that led to the inevitable adoption of Linux by the majority of the
developers, which in turn led to the adoption of the system by other groups. So, my
view is that the reason everyone today uses Linux is because AT&T thought so little
of Linux that they didn't bother killing it; a strategic mistake on the part of
AT&T.
Dave Barr
Brad Knowles <br...@shub-internet.org> wrote:
> I must respectfully disagree. Part of the long history of Unix is in
>the proven code that has not been thrown out and reinvented.
Part of the long history of UNIX is precisely the throwing out of
proven code and reinventing it. Every single major piece of UNIX has
been overhauled many times over the years. Every vendor has overhauled
major portions of the code too.
I simply don't understand this preoccupation with blood lines.
>If you throw
>everything out and start over, then no matter how similar to Unix you
>might be, you simply are not (and cannot ever be) Unix.
Utter bull. That's like saying a Honda is not a car because it's not
made by Ford.
> No matter how similar in appearance and feel a reproduction might be
>to the original, it simply is not the original.
What's so special about the original? The original is a piece of crap
running on a PDP-11!! This is not a painting. This is a living, evolving,
and in some cases reborn piece of work.
David Burgess
Dave Barr wrote:
> In article <7hp8q3$2v...@enews1.newsguy.com>,
> John S. Dyson <to...@y.dyson.net> wrote:
> >I tend to question them also, mostly because of mis-statement. Linux
> >is a gratuitiously rewritten U**X clone, that has a MORE restrictive
> >license than the previously available versions. There were some minor
> >bumps in the BSD road, but those have been resolved for many years.
>
> You forget your history lesson.
>
> The *BSD code base was basically locked up in legal fights for *2 years*.
> That's an incredibly long time -- just look at where Linux was 2 years ago.
>
It was about where NetBSD and FreeBSD were two years ago. There has been
relative parity for most of that two years. The question, of course, would be
"where would Linux be today if Andrew Tanenbaum had filed a restraining order
against Linux for 'gratuitous borrowing with or without attribution'?"
That's pretty much what AT&T did to us. Prentice Hall would have bankrolled
the suit, and Linus would not have had the resources of the state of California
to back him. There would have been grounds. Linux originally was Minix
executable compatible and used all of the Minix file systems. I suppose you
could 'just stumble onto' all of that, but I doubt it was seriously
re-engineered.
> That's not gratuitous rewriting. Linus *wanted* a good UNIX, and given
> there was no available code base from which to work he had to start
> from scratch.
Linus wasn't working in a vacuum either. Remember that he made the original
announcement in "comp.os.minix". Linus wanted a good Unix, but rather than
support one of the on-going efforts (386BSD and Minix-386 were available by
then) he decided to strike out on his own.
> By the time the 4.4 code base went public, Linux was already quite far
> along.
386BSD and Linux were released within weeks of each other. The 4.4 code base
was the response to the AT&T lawsuit; 386BSD was based on 4.3 which was
released in source code form. That's what started the lawsuit in the first
place.
o r c e l l . p o r t l a n d . o r . u s
>I tend to question them also, mostly because of mis-statement. Linux
>is a gratuitiously rewritten U**X clone,
If Linux was written for the explicit purpose of killing off *BSD,
then it would have been gratuitiously rewritten. It was not.
It's possible that Linus would have joined a *BSD development effort
when he started writing operating systems if there wasn't for the
teeny little detail that the million-pound hammer of an AT&T lawsuit
was poised over *BSD's head (I can't speak for anyone else, but
that lawsuit was directly responsible for my choosing to run Linux;
I wanted a free Unix and I didn't want to have to either abandon
ship or go pirate if AT&T pounded the CSRG flat) but certainly by
the time AT&T sold Unix to Novell that moment had passed.
>that has a MORE restrictive
>license than the previously available versions.
It's free to the end user. That's a lot less restrictive than the
traditional AT&T/Sun/IBM/Digital/Silicon Graphics/etc "pay us lots
of money and you can use this software on one machine" approach.
At at the time Linux went GPL, there was a little matter of a
still-unresolved lawsuit between a multi-billion dollar company
and the CSRG; the BSD license would have not been worth the
paper it was on if that lawsuit had gone the wrong way.
>The technically superior (in one way or another) choice isn't always
>the most successful. What the BSD folks probably lament is that a
>better choice is being overwhelmed by a vast marketing engine.
But it's not. In the free software world, a rising tide DOES
lift all boats, and once the user has tasted Unix it's easy
for them to switch between Unices. Don't quibble about who
brings them into the boat, just be happy that they're running
Unix.
____
david parsons \bi/ Don't keep fighting a war that was over in 1993.
\/
John S. Dyson
ba...@cis.ohio-state.edu (Dave Barr) writes:
> In article <7hp8q3$2v...@enews1.newsguy.com>,
> John S. Dyson <to...@y.dyson.net> wrote:
>>I tend to question them also, mostly because of mis-statement. Linux
>>license than the previously available versions. There were some minor
>>bumps in the BSD road, but those have been resolved for many years.
>
> You forget your history lesson.
>
> The *BSD code base was basically locked up in legal fights for *2 years*.
> That's an incredibly long time -- just look at where Linux was 2 years ago.
>
except for marketing. It was GPL idealogy that has been trying to
destroy free software.
>
> That's not gratuitous rewriting. Linus *wanted* a good UNIX, and given
> there was no available code base from which to work he had to start
> from scratch.
>
I wanted a good UNIX, and decided (and still know) that the BSD codebas
is simply better (note the VM hackery going on: can you say tweak and
tune? :-)).
>
> By the time the 4.4 code base went public, Linux was already quite far
> along.
>
It was always public, and in no way were users going to be enjoined.
John S. Dyson
o r c @ p e l l . p o r t l a n d . o r . u s (david parsons) writes:
> In article <7hp8q3$2v...@enews1.newsguy.com>,
> John S. Dyson <to...@y.dyson.net> wrote:
>
>>I tend to question them also, mostly because of mis-statement. Linux
>>is a gratuitiously rewritten U**X clone,
>
> then it would have been gratuitiously rewritten. It was not.
> It's possible that Linus would have joined a *BSD development effort
> when he started writing operating systems if there wasn't for the
> teeny little detail that the million-pound hammer of an AT&T lawsuit
> was poised over *BSD's head (I can't speak for anyone else, but
> that lawsuit was directly responsible for my choosing to run Linux;
> I wanted a free Unix and I didn't want to have to either abandon
> ship or go pirate if AT&T pounded the CSRG flat) but certainly by
> the time AT&T sold Unix to Novell that moment had passed.
>
have worried (too much.) I know, I was there. I guess that there was
just alot of uninformed speculation (I worked at AT&T and worked on FreeBSD
AT THE SAME TIME.)
>>that has a MORE restrictive
>>license than the previously available versions.
>
> It's free to the end user.
>
It isn't free to all end users, by virtue of the restrictions against
developers.
>
> That's a lot less restrictive than the
> traditional AT&T/Sun/IBM/Digital/Silicon Graphics/etc "pay us lots
> of money and you can use this software on one machine" approach.
>
Geesh, so GPL is somewhere between free and commercial? So what, it isn't
free!!!
>
> But it's not. In the free software world, a rising tide DOES
> lift all boats, and once the user has tasted Unix it's easy
> for them to switch between Unices. Don't quibble about who
> brings them into the boat, just be happy that they're running
> Unix.
>
The rising tide is tieing up alot of the economy, thereby destroying
innovation. GPLers tend not to respect the developers where the innovation
is created, while GPL certainly does help (RMS, Linus, Cygnus, Marketeers
such as Red-Hat), but the value added developer in the middle is
screwed (without lieing about selling software as support.)
John S. Dyson
d...@nano.ucsf.edu (David Konerding) writes:
> On 17 May 1999 15:07:47 GMT, John S. Dyson <to...@y.dyson.net> wrote:
>>In article <7hp771$cds$1...@news.cis.ohio-state.edu>,
>>The Unix that most people see as Unix is more than partially derived
>>from BSD in one way or another. The original Unix predates BSD, but
>>BSD predates (partially) the current Unix. Practically everything
>>predates Linux, and Linux is a re-coding of the previously created
>>research and innovation. (Sometimes better, sometimes worse, but
>>a restrictively licensed pseudo-clone of BSD anyway.)
>
> I would call it a "pseudo-clone of POSIX", rather than BSD, since
> it was never written to be a clone of BSD.
>
more than POSIX -- please note NT complying with POSIX... If
you had to compare them all, the difference between BSD and Linux
is much smaller than between Linux and SYSVr4, or even Linux and
many POSIX implementations. IMO, Linux is more like SYSVr3 + BSD
networking + better VM... Hmmm... That almost makes it the same
as BSD!!!
>
> pseudo of course because there
> is still a bit more work to be done to make Linux a truly first-class
> Unix(clone) (in my case, there are some mmap implementation issues
> that affect performance of my code).
>
Well, BSD was already first class, and still mostly out-performing.
>
> However, while I have been a Linux user for 5+ years and having using
> FreeBSD for much less (only about 1 year), I am beginning to wonder why
> Linux has become the "de facto free Unix" standard. Basically nearly
> everything in Linux is a re-implementation of something that existed
> in BSD already, suggesting that there is a great deal of wasted effort
> going on here.
>
Yes, exactly!!!
>
> It makes one wonder: what if timing and luck had been slightly different,
> and instead of Linux catching on, {Free|Open|Net}BSD had.
> Or, if a Linux kernel was placed on to of the BSD toolbase, instead of
> the GNU toolbase?
>
It seems to me that Linux is more a creature of NIH (not invented here),
and "I can do it better than the real professionals" attitude of many
young folk. The bottom line is that Linux and BSD are asymtotically
the same, with the difference being that Linux has the unfortunately
restrictive license. Since BSD has been done without GPL (and adopting
GPLed tools along the line), and Linux has been done with GPL, arguments
that the GPL is necessary are quite specious.
GPL simply restricts the freedom of the developers who add innovation.
--
Leslie Mikesell
>
>Isn't one of FreeBSD's selling points over Linux that there is a
>relatively greater extent of central control - you can still do what
>you want with your source, to an even greater extent than you could
>under the GPL, but there is only one "official" release of FreeBSD?
Hmmm, if you think that being unable to get a distribution
configured approximately the way you want is a 'selling point'
for FreeBSD, what do you consider the bad points to be?
Les Mikesell
l...@mcs.com
G. Sumner Hayes
> It is certainly closer to BSD than SYSV... It is LOTS AND LOTS
> more than POSIX -- please note NT complying with POSIX...
NT complies with POSIX.1, Linux complies with many more POSIX
specifications. For instance, select() and poll() are POSIX.1g.
Realtime signals and scheduling are parts of POSIX.1b. NT supports
none of those standards. (well, it has a broken, non-POSIX
select() that sort of works on sockets only). But yes, Linux
does try to support common BSDisms where possible. e.g. the GNU C
library has changed signal restart behaviorto be BSDish by default.
But it also tries to support common sysVisms by default. It even
steals NT features on occasion.
> Well, BSD was already first class, and still mostly out-performing.
I believe strongly that BSD would not be as good as it is today if
Linux didn't exist. Nor would Linux be as good as it is without
BSD. You seem to feel that competing projects are wasted effort;
that's possible in some cases, but in many cases it also gives good
incentive for improvements and added features. Seeing the other
guy trounce you on the benchmark of the week makes you anxious to
come back and bury him next week. Witness the statement on Terry
Lambert's devel page about how it would be nice to beat Linux to a
feature for once. Or Trond's NFSv3 work, one goal of which is
to beat out BSD's NFS performance. The sense of competition is a
real motivator, especially when you have at least a small amount of
respect for the competitor. Trouncing NT on context-switching tests
is just too easy.
> > It makes one wonder: what if timing and luck had been slightly
> > different, and instead of Linux catching on, {Free|Open|Net}BSD had.
> > Or, if a Linux kernel was placed on to of the BSD toolbase, instead of
> > the GNU toolbase?
> >
> It seems to me that Linux is more a creature of NIH (not invented here),
> and "I can do it better than the real professionals" attitude of many
> young folk.
Well, Linus had two major goals when he started.
1. Get a good, free Unix for his PC. He (and HJ Lu, and a bunch of
other core Linux folk) have basically said that if BSD were available
and didn't have a suit facing it at the time he went to work, there
would be no Linux. Whether it was a real threat or not, the
suit did have enough of a chilling effect to drive some to Linux.
2. Learn how to program the 386. Writing your own is a great way to
learn a system from the inside out.
Bear in mind, too, that the vast majority of a Linux system
wasn't written for Linux. The GNU utilities, X, TeX, and a whole
lot of the rest of a typical Linux distro are shared with plenty
of other platforms and predate Linux. BSD even uses some of them.
The kernel is NIH by design, the rest of the system is often happy
to steal other people's work.
> the difference being that Linux has the unfortunately restrictive
> license.
Ooo, flamebait. Unfortunate from your perspective. Fortunate from
other people's perspectives. Obviously people feel strongly about
this issue. Thankfully, we have choices; you can use BSD if you
want that license, and Linux if you want GPL. Most users don't
really care. At the moment, they're mostly using Linux. I wouldn't
be so anxious to get them if I were you. There's something to be
said for a system that is used primarily by people with a certain
amount of technical background, especially when there's a similar
platform out there gathering the market share needed to keep the
basic standards open. One can concentrate solely on technical issues,
the other has to balance that with evangelization and keeping
nontechnical users happy.
Indeed, I believe that if BSD had become as popular as Linux is
today, something would have emerged as a place where certain hackers
could live in peace and make technical decisions without putting
up with the bizarre misfeature requests of all the newbies. Heck,
some Debian users heckle Red Hat without end, in part just because
it's popular.
--Sumner
Colin R. Day
> In article <brad-17059...@brad.techos.skynet.be>,
> br...@shub-internet.org (Brad Knowles) writes:
> > In article <7hp725$bp3$1...@news.cis.ohio-state.edu>, ba...@cis.ohio-state.edu
> > (Dave Barr) wrote:
> >
> >> The measure of UNIX is how close it is to UNIX in action, not whose
> >> parents it had.
> >
> > I must respectfully disagree. Part of the long history of Unix is in
> > the proven code that has not been thrown out and reinvented. If you throw
> > everything out and start over, then no matter how similar to Unix you
> > might be, you simply are not (and cannot ever be) Unix.
> >
> Maybe I don't agree with everything in your statement, however the above
> paragraph is *exactly* one of my positions, and the fact that Linux is
> a gratuitious reimplementation is a macro-level example of the
> reimplementation that U**X developers have tried to avoid.
>
Linux is not a gratuitous reimplementation of UNIX
>
> --
> John | Never try to teach a pig to sing,
> dy...@iquest.net | it makes one look stupid
> jdy...@nc.com | and it irritates the pig.
--
Colin R. Day cd...@ix.netcom.com alt.atheist #1500
EAC Cheerleader RAH! RAH! RAH! Go, team, go! (of course, there
is no EAC team)
o r c e l l . p o r t l a n d . o r . u s
> o r c @ p e l l . p o r t l a n d . o r . u s (david parsons) writes:
>> In article <7hp8q3$2v...@enews1.newsguy.com>,
>>>is a gratuitiously rewritten U**X clone,
>>
>> If Linux was written for the explicit purpose of killing off *BSD,
>> then it would have been gratuitiously rewritten. It was not.
>> It's possible that Linus would have joined a *BSD development effort
>> when he started writing operating systems if there wasn't for the
>> teeny little detail that the million-pound hammer of an AT&T lawsuit
>> was poised over *BSD's head (I can't speak for anyone else, but
>> that lawsuit was directly responsible for my choosing to run Linux;
>> I wanted a free Unix and I didn't want to have to either abandon
>> ship or go pirate if AT&T pounded the CSRG flat) but certainly by
>> the time AT&T sold Unix to Novell that moment had passed.
>>
>That lawsuit begs the issue. Anyone knowing what was going on, wouldn't
>have worried (too much.) I know, I was there.
So was I. Your interpretation of what was going on is rather different
than mine (and, alas, a large part of the rest of the computing world);
it is truely unfortunate that your opinions were not both more widely
published and paid attention to.
Implying that people who were worried about AT&T didn't know what was
going on, though it might be good for your ego, does absolutely nothing
for the standing of FreeBSD TODAY, in 1999, except making it look like
the FreeBSD community contains a large number of sore losers who can't
help but whine about the past.
>>>that has a MORE restrictive
>>>license than the previously available versions.
>>
>> It's free to the end user.
>>
>It isn't free to all end users, by virtue of the restrictions against
>developers.
No, it's equally free to all end users; any user can get a copy of
it and copy it around and redistribute it. This sucks for the
class of developers who can't take advantage of the loopholes to
write free (as in freedom of speech, not free beer) code, but
*you don't have to use it if you don't like it*
>> That's a lot less restrictive than the
>> traditional AT&T/Sun/IBM/Digital/Silicon Graphics/etc "pay us lots
>> of money and you can use this software on one machine" approach.
>>
>Geesh, so GPL is somewhere between free and commercial? So what, it isn't
>free!!!
It's free like free beer. It's not free as in freedom of speech.
That's why I ship all my code (except for the few kernel hacks I
do to the Linux kernel) with a BSD-style copyright. BUT as an
end user the GPLed stuff is still free, because I don't have to
pay to use it.
>> But it's not. In the free software world, a rising tide DOES
>> lift all boats, and once the user has tasted Unix it's easy
>> for them to switch between Unices. Don't quibble about who
>> brings them into the boat, just be happy that they're running
>> Unix.
>>
>The rising tide is tieing up alot of the economy, thereby destroying
>innovation.
No it doesn't; about the only way I can see it destroying
innovation is that GCC is so dominant in the free software community
that it has killed every free (speech, not beer) compiler project
out there. Certainly it hasn't killed any other projects -- the
ongoing GNOME jihad against KDE is one GPLed project trying to
kill another one, not someone trying to reinvent a GPLed project
under a BSD license.
>GPLers tend not to respect the developers where the innovation
>is created,
So don't use it; once you get away from GCC and the desktops there
are freely licensed developments everywhere, many of which predate
the second wind of the FSF. The existance of the GPL hasn't
stopped me from moving almost the entire userland of Mastodon Linux
over to OpenBSD tools, nor has it stopped me from continuing to use
free-licensed versions of software that has been `improved' by being
converted to the GPL (like ncurses, which I'd love to contribute
code to but won't now that the license has been changed.)
____
david parsons \bi/ If you hate the GPL, drop a clause into your licenses
\/ forbidding use of your code with GPLed code. If you'd
rather the GPL wither and vanish, write better code.
Jim Ross
>> Also most people shouldn't
>> be mucking around with the kernel much anyway. It is a very low level piece
>> of software. User level stuff and even some level are what should be written.
>> We already have Linux. If that isn't what you need then why use Linux. It
>> doesn't fit your need.
>>
>People shouldn't be 1) mucking around with kernels, 2) mucking around
>with libraries, 3) mucking around with applications, 4) mucking around
>with user interfaces, 5) mucking around with ANY software?
>
>think that software in general shouldn't be "mucked" with? A kernel
>is no different than any other piece of software, and not necessarily
>the most complex piece. It isn't beyond improvement for some kind of
>competitive advantage.
Yes. I don't want the average joe with a software program mucking with the
kernel. I want them so send the fixes for people who know what they are doing
the incorporate the changes unless you some software like clustering which
must update the kernel. I hope others agree.
> >>
>> 2. Writing source once and selling binaries sounds like an outdated concept.
>>
>That is a statement that is simply not true. By not allowing that, you
>cause all enhancement for a sector to have to be funded by one customer.
>(Or secondly, mis-sell support as access.)
>
>>
>> You can just
>> change code and keep it. Everyone should know going in Linux is not yours but
>> GPL. Everyone owns it and end of story.
>>
>In cases where everyone owns an expensive investment, there will be
>stagnation. Linux will forever be a '80's monolithic kernel, which
>is totally unacceptable. The sad thing is that the short term thinking
>of a portion of the userbase is accepting Linux, mostly because of advocacy,
>timing, and a short-term marginal benefit over commercial software.
>
>Some companies are even becoming fearful of any kind of forward looking
>investment, and will eventually become the mediocre that they chose
>to be by fear. GPL is a kind of license that preys on fear and the
>notion of software being "cheap" to get. Hey you programmers out there:
>if the software is "cheap", you are going to be "cheap" also. When
>you have families to support, and the economy normalizes, then you'll
>start becoming more sensitive to these facts.
>
>Geesh!!!
>
>--
>John | Never try to teach a pig to sing,
>dy...@iquest.net | it makes one look stupid
>jdy...@nc.com | and it irritates the pig.
Ok ok take a stress pill.
I don't see GPL taking over the world or MS going anywhere.
Jim
Richard Pennington
devil its due, Linux should be refered to GNU/Linux. Or should
I send the thought police after you?
Long live Big Brother!
-Rich
> It isn't? Remember, it is a reimplementation of UNIX, and
> the difference between it and free UNIX clones is the encumbered
> GPL. No net gain in reimplementation...
>
> --
> John | Never try to teach a pig to sing,
> dy...@iquest.net | it makes one look stupid
> jdy...@nc.com | and it irritates the pig.
--
Richard Pennington Introl Corporation, Milwaukee, WI USA
Email: ri...@introl.com Phone: +1 414-273-6100 Fax: +1 414-273-6106
US and Canada: 1 800-327-7171
Cross development tools for the 68HC05, 68HC08, 6809, 68HC11, 68HC12,
68HC16,
and 68XXX: http://www.introl.com ftp://ftp.introl.com
Tom Keats
David Burgess <bur...@mitre.org> writes:
[snip]
>
> Shortly after the release of 386BSD 0.1, UC Berkeley was slapped with a lawsuit
> saying that they had released unspecified copyrighted and 'trade secret' materials.
> This pretty effectively froze development of 386BSD.
IIRC, BSDi had the jam to continue to distribute their
flagship of the time: BSD/386 v. 1, released in March '93.
(i still have the discs here somewhere...)
Novell sought and failed to obtain an injunction against
further distribution while the court date was pending.
Wish i could remember when v1.1 came out.
A comparison of the time lines of the legal dispute, and the
development/distribution from BSDi might prove interesting.
I believe that v. 1.1 was BSDi's first distro based on 4.4Lite
(the one with the encumbered code stripped-out.)
[snip]
--
--
Why not just drop everything and go fishin'?
remove NO_SPAM. from address to reply
o r c e l l . p o r t l a n d . o r . u s
Richard Pennington <ri...@introl.com> wrote:
>I think this discussion has been interesting. However, to give the
>devil its due, Linux should be refered to GNU/Linux.
If you're referring to Debian GNU/Linux, go right ahead.
If you're referring to other Linux distributions, you're
on somewhat shakier ground. eg:
665 files in /Mastodon/cd/basic:
174 say 'Regents' somewhere in them (26.17%)
56 say 'GNU' somewhere in them ( 8.42%)
1 mentions both 'GNU' and 'Regents'
34 say 'Mastodon' somewhere in them ( 5.11%)
2 are ELF-format ( 0.30%)
161 files in /Mastodon/cd/x:
2 say 'GNU' somewhere in them ( 1.24%)
30 say 'XConsortium' somewhere in them (18.63%)
50 files in /Mastodon/cd/games:
38 say 'Regents' somewhere in them (76.00%)
____
david parsons \bi/ It's hard to programmatically tell the extent of the
\/ GNU code, because the FSF doesn't embed copyright
notices in their binaries; looking for ``GNU''
also picks up GNU-worship in non-FSF binaries.
o r c e l l . p o r t l a n d . o r . u s
>The restrictions placed by the GPL are designed to guarantee that
>GPL'd software and derivative works will remain free, preventing, one
>would hope, things like the attempted hijacking of X by the Open
>Group, and the theft of Mosaic by Microsoft.
Neither of which prove anything. UI always reserved the right to
make profits from Mosaic, and their licensing terms are such that
it's not worth the trouble of tweaking Mosaic because you can't
even redistribute it, and when TOG attempted to make X11R6v4 into a
proprietary product the XFree86 project shrugged, said "fine, it's
your funeral" and prepared to continue with X11R6v3. Note that TOG
has since retreated from their new license.
In both of these cases, the copyright holders were perfectly within
their rights to do what they wished with their code, just as the
FSF would be completely within THEIR rights to re-release the GNU
project's software base under a different license.
____
david parsons \bi/ Hope that helps!
\/
Tim Smith
>The restrictions placed by the GPL are designed to guarantee that
>GPL'd software and derivative works will remain free, preventing, one
>would hope, things like the attempted hijacking of X by the Open
>Group, and the theft of Mosaic by Microsoft.
Mosaic is still available if you want it.
--Tim Smith
Navindra Umanee
> Finally, a UNIX that no-one can own. FreeBSD doesn't fit
> that need (technically wonderful though it is).
Since you're around... Have you benchmarked Samba on FreeBSD vs
Linux? :-)
-N.
--
"These download files are in Microsoft Word 6.0 format. After unzipping,
these files can be viewed in any text editor, including all versions of
Microsoft Word, WordPad, and Microsoft Word Viewer." [Microsoft website]
< http://www.cs.mcgill.ca/~navindra/editors/ >
Chris Costello
>
> > Linux should be refered to GNU/Linux
>
I don't think NetBSD is a GNU system. I don't think NetBSD is
GPL'd.
--
Chris Costello
In /dev/null no one can hear you scream
Graffiti
David Konerding <d...@cgl.ucsf.edu> wrote:
[snip]
>I would call it a "pseudo-clone of POSIX", rather than BSD, since
Sure it was. Linus based his inital work on SunOS 4.
(Hint: SunOS 4 is a BSD-ish UNIX)
>However, while I have been a Linux user for 5+ years and having using
>FreeBSD for much less (only about 1 year), I am beginning to wonder why
>Linux has become the "de facto free Unix" standard. Basically nearly
>everything in Linux is a re-implementation of something that existed
>in BSD already, suggesting that there is a great deal of wasted effort
>going on here.
Wasted effort? Coding because you enjoy coding is not "wasting effort".
If it was, otoh, a group that had to get a working system up and running
*fast* and refused to use existing code, then yes, it'll be "wasted effort".
>It makes one wonder: what if timing and luck had been slightly different,
>and instead of Linux catching on, {Free|Open|Net}BSD had.
>Or, if a Linux kernel was placed on to of the BSD toolbase, instead of
>the GNU toolbase?
Wonder all you want. Only the die-hard fanatics will take offense at
*that*. :-)
-- DN
Graffiti
david parsons <o r c @ p e l l . p o r t l a n d . o r . u s> wrote:
[snip]
> 665 files in /Mastodon/cd/basic:
> 174 say 'Regents' somewhere in them (26.17%)
> 56 say 'GNU' somewhere in them ( 8.42%)
> 1 mentions both 'GNU' and 'Regents'
> 34 say 'Mastodon' somewhere in them ( 5.11%)
> 2 are ELF-format ( 0.30%)
Huh?!
I can see the kernel being ELF, and the modules (assuming you're
using a kernel that has to be ELF), but I assume these are user-land?
What do you use ELF for?
> david parsons \bi/ It's hard to programmatically tell the extent of the
> \/ GNU code, because the FSF doesn't embed copyright
> notices in their binaries; looking for ``GNU''
> also picks up GNU-worship in non-FSF binaries.
Hmm... I assume you have some kind of 'make world'-like script for your
distribution? Then you can probably look for the COPYING file in the
source dir to find most of the GPLd stuff, since people tend to toss the
GPL in that file.
-- DN
Jeremy Allison
>Jeremy Allison <jer...@netcom.com> wrote:
>> Finally, a UNIX that no-one can own. FreeBSD doesn't fit
>> that need (technically wonderful though it is).
>Since you're around... Have you benchmarked Samba on FreeBSD vs
>Linux? :-)
I've asked Julian Elisher (one of my old FreeBSD mates
at Whistle) to see if he can get Jorden & crew to
come into the SGI lab and do just that.
As we now know we're TCP limited when serving Win95 clients
on Linux this will be a very interesting test. Personally
I'm happy for either to win. They're both Open Source
UNIX and they both use Samba (as did Sun in the PC Week
benchmark - I was *so* happy to see that, it means
they know their own SMB server products don't cut it
speed-wise :-) :-).
Cheers,
Jeremy.
Leslie Mikesell
Peter da Silva <pe...@baileynm.com> wrote:
>>Hmmm, if you think that being unable to get a distribution
>>configured approximately the way you want is a 'selling point'
>>for FreeBSD, what do you consider the bad points to be?
>
>Where do you get the idea that you can't get a distribution configured
>approximately the way you want with FreeBSD? It's pretty easy to build a
>custom distribution of FreeBSD: tune the source tree the way you want and
>run a "make" in /usr/src/release.
By definition, having to build it yourself isn't getting it configured
the way you want. Is there a freeBSD distribution where I can
say 'install everything', and have it fill 2 gigs or less and
come up running KDE, postgresql, samba, xntpd, apache and a nice
collection of utilities when it is done? Are there an assortment
of such distributions so other people can pick the ones they
want too?
Les Mikesell
l...@mcs.com
Jeremy Allison
>It isn't? Remember, it is a reimplementation of UNIX, and
>the difference between it and free UNIX clones is the encumbered
>GPL. No net gain in reimplementation...
Seriously John, from my (and many other's) points of view
the GPL license on Linux *IS* the net gain in the reimplementation.
Finally, a UNIX that no-one can own. FreeBSD doesn't fit
that need (technically wonderful though it is).
Regards,
Jeremy Allison,
Samba Team.
Christopher B. Browne
>Unfortunately it's easier to have an argument about how pure-blooded
>Linux is or isn't compared to *BSD than it is to have a discussion
>comparing their technical merits. The latter requires more knowledge.
Historically, it has often enough been easier for people to have killing
hatred towards near "cousins" than for those that ought to have been
enemies, particularly when it comes to religious matters. Some of the
deadliest conflicts have been between not-too-distant sects of Christian,
Jewish, and Islamic faiths.
And I don't think it's too much of a stretch to suggest that the hatred that
Linux-versus-BSD elicits is closer to representing a religious reaction than
a technical reaction...
--
Those who do not understand Unix are condemned to reinvent it, poorly.
-- Henry Spencer <http://www.hex.net/~cbbrowne/lsf.html>
cbbr...@hex.net - "What have you contributed to free software today?..."
Christopher B. Browne
>It is *very* important to keep reminding new users as to the
>disadvantages of GPL. Until the defacto GPL advocacy stops,
>then the other side has to be presented.
>
>Note that I am not "whining", but providing another viewpoint. I
>sometimes perceive others who don't agree with me as "whining", but
>that is *sometimes* not accurate either. I suspect that the use
>of that word "whining" in this case is due to a choice of the wrong
>term. Perhaps the correct term is "warning", or "elucidating."
When you spend your time "bashing" the GPL, you come off looking like a
jerk, in much the way that politicians that engage in "smear" politics look
bad.
People feel sympathy for those that are under attack, and seldom for those
that are the aggressor, and your attacks on the GPL certainly make you look
like an aggressor.
I suggest that you try spending some time and words suggesting what is GOOD
about BSD-related licenses. That gives you the opportunity to sit on the
moral "high ground."
Note that if the best you can do is to say what's bad about the GPL, people
have little reason to think that you have a useful alternative.
Christopher B. Browne
posted:
>John S. Dyson <to...@y.dyson.net> wrote:
>>The rising tide is tieing up alot of the economy, thereby destroying
>>innovation.
>
> No it doesn't; about the only way I can see it destroying
> innovation is that GCC is so dominant in the free software community
> that it has killed every free (speech, not beer) compiler project
> out there. Certainly it hasn't killed any other projects -- the
> ongoing GNOME jihad against KDE is one GPLed project trying to
> kill another one, not someone trying to reinvent a GPLed project
> under a BSD license.
Actually, on the C compiler front, it could be argued that the somewhat
mediocre success of GCC has "starved" labour for other possible efforts for
C compilers such as TENDRA, LCC, or other alternatives that have never even
gotten started since GCC was "good enough."
I say "mediocre success" in that there have been considerable periods of
time during which there has been little improvement, and certainly not to
say that GCC is "horrible/bad/evil and that the developers are morons."
The point is that there have been some possible alternatives to GCC over the
years, and innovations that aren't compatible with the GCC architecture *do*
tend to get stifled.
>>GPLers tend not to respect the developers where the innovation
>>is created,
>
> So don't use it; once you get away from GCC and the desktops there
> are freely licensed developments everywhere, many of which predate
> the second wind of the FSF. The existance of the GPL hasn't
> stopped me from moving almost the entire userland of Mastodon Linux
> over to OpenBSD tools, nor has it stopped me from continuing to use
> free-licensed versions of software that has been `improved' by being
> converted to the GPL (like ncurses, which I'd love to contribute
> code to but won't now that the license has been changed.)
>
> ____
> david parsons \bi/ If you hate the GPL, drop a clause into your licenses
> \/ forbidding use of your code with GPLed code. If you'd
> rather the GPL wither and vanish, write better code.
Indeed.
The best way to make the GPL "wither and vanish" is not to wail at its
ascendancy, but to produce code under other licenses that is so good and so
eloquently promoted that the code and user base grows with such fecundity
that it chokes out all that "bad old GPLed code."
Christopher Browne
>On 17 May 1999 15:07:47 GMT, John S. Dyson <to...@y.dyson.net> wrote:
>>In article <7hp771$cds$1...@news.cis.ohio-state.edu>,
>>The Unix that most people see as Unix is more than partially derived
>>from BSD in one way or another. The original Unix predates BSD, but
>>BSD predates (partially) the current Unix. Practically everything
>>predates Linux, and Linux is a re-coding of the previously created
>>research and innovation. (Sometimes better, sometimes worse, but
>>a restrictively licensed pseudo-clone of BSD anyway.)
>
>it was never written to be a clone of BSD.
That's definitely a better comparison.
>pseudo of course because there is still a bit more work to be done to
>make Linux a truly first-class Unix(clone) (in my case, there are some
>mmap implementation issues that affect performance of my code).
That's more arguable, of course; there are a lot of opinions out there
on the matter...
>Having tried both Linux (from .99 to 2.2.7) and FreeBSD (2.2, not 3.0)
>I'd have to say they are awfully similar in functionality, there isn't
>really a tremendous difference between the quality of performance (for
>example NFS) or in the quality of the implementations (NFS, TCP, mm,
>paging, filesystem) especially if one is talking about Linux 2.2, not
>2.0.
Of course, FreeBSD 3.0 might leapfrog ahead :-).
But for a whole lot of purposes, "awfully similar" is pretty fair
statement.
>However, while I have been a Linux user for 5+ years and having using
>FreeBSD for much less (only about 1 year), I am beginning to wonder why
>Linux has become the "de facto free Unix" standard. Basically nearly
>everything in Linux is a re-implementation of something that existed
>in BSD already, suggesting that there is a great deal of wasted effort
>going on here.
Of course, there are not one, but *four* BSDs, which implies that
there's probably some wasted effort within the BSD "camps" too.
>It makes one wonder: what if timing and luck had been slightly
>different, and instead of Linux catching on, {Free|Open|Net}BSD had.
>Or, if a Linux kernel was placed on to of the BSD toolbase, instead of
>the GNU toolbase?
Linus Torvalds has been on record as saying that if there had been a
good BSD implementation available at the time he started working on
Linux, he probably wouldn't have taken it past "toy-hood."
That being said, a big chunk of what has happened in the progressions
of the various projects {*BSD, FSF, Linux} has come as a result of the
combinations of personalities involved.
There is something of a "Linus personality cult" (those being Linus'
direct words); while it perhaps "has got to go," I think it is
nonetheless fair to say that his personality has had a fair bit to do
with the willingness of people to participate. It has apparently
"rubbed enough people the right way."
If the lawsuit situation had happened differently, BSD, in one or
another of the forms, might nonetheless have "ruled the roost" for those
working on free software.
I suspect that if BSD folk didn't so head to gnu.misc.discuss to bash
the GPL quite as often that there might be better attitudes towards
BSD...
--
"It's obvious that the ``Linus personality cult'' has got to go."
-- Linus Torvalds, May 5, 1999, ABC News
cbbr...@hex.net- <http://www.hex.net/~cbbrowne/lsf.html>
Christopher Browne
>John S. Dyson (to...@y.dyson.net) wrote:
>: Actually, there are large parts of the MS code that are source
>: distributed. (e.g. libraries.)
>
>What libraries? (MFC doesn't count because it's a wrapper around
>Win32, which most assuredly is not available in source form.)
Apparently "wrappers" are the same thing as "libraries."
>: Geesh, with the movement towards GPL'ing libraries,
>: it seems that MS and GPL world is on more and more even footing.
>
>I understand that you don't like the GPL license and that's fine. But
>what would possess any seemingly bright human being to suggest that an
>open-source model (albeit one you obviously don't like) can even be
>compared to a source-not-available-at-any-price model like the one
>Microsoft espouses?
The only "movement" I am aware of is that RMS is deprecating the use of
the LGPL. That is not the same thing as anyone actually changing the
licenses for libraries.
Although I could be entirely wrong; perhaps John can name some libraries
that have been released under the GPL lately. Otherwise I have to
assume that the comment merely represents a bash against RMS.
>: In cases where everyone owns an expensive investment, there will be
>: stagnation. Linux will forever be a '80's monolithic kernel, which
>: is totally unacceptable.
>
>It's perfectly acceptable now. Nothing drastically better exists yet,
>at least not on PC-class hardware.
>
>If, down the road, it proves not to be, then almost all free software
>written for Linux, assuming that it isn't broken, will continue to be
>able to be compiled and run on FreeBSD or any other POSIX-conforming
>OS.
>
>The same should be true of proprietary software as well, although if
>you don't at least have access to source, you'll be at the mercy of
>your vendor who may or may not choose to do so.
The Debian/Hurd effort is, I would suggest, the most critical "chink" in
this regard. Those that despise the GPL and the FSF may feel free to
consider this, a system more closely tied to GPL and FSF than Linux,
something that is worthless to think about.
I suggest the contrasting thought that this effort, which is striving to
take Debian packages and put them on Hurd, is a step towards making
Debian into a more kernel-independent distribution. That then reduces
the forcible ties between software running on Linux and Linux itself.
Some wags have proposed building "Debian/FreeBSD;" I'd think it more
interesting to build "Debian/XOS" or "Debian/OSKit," as steps towards
"dramatically better" OSes.
And I agree with the "perfectly acceptable" part. Griping that Linux,
as a monolithic kernel, is "totally unacceptable," is devoid of value
unless there is some *acceptable* alternative. And the only systems I
can name offhand that use microkernels now that aren't hopelessly
proprietary are:
a) Hurd,
b) Linux on Mach
c) Linux on L4
I can't find the "Lites" site anymore, otherwise that would have been
d)...
>: The sad thing is that the short term thinking of a portion of the
>: userbase is accepting Linux, mostly because of advocacy,
>: timing, and a short-term marginal benefit over commercial software.
>
>What people love about Linux, as far as I can see, is not that it
>isn't FreeBSD. (I like FreeBSD too.) What they love about it is that
>it is free - at least, more free than what they are used to - and that
>because it is free in the sense of liberty, it's generally of much
>higher quality.
>
>The shift away from proprietary, mediocre, and highly vendor-bound
>bloatware should benefit the *entire* free and open-source software
>communities, and as far as I can see, it does.
... And as the libraries and other tools converge towards POSIX and ANSI
C and other aspects of being 'nearer-to-official UNIX conformance,' the
software being developed atop Linux becomes *less* tied to Linux.
A fair bit of the breakage taking place with the glibc transitions
seems to relate to cases where people wrote code that was tightly coded
either to:
a) Linux, or
b) Internal bits of libc.
Moving to the new library exposes this, and hopefully encourages Doing
It Right, which should make the code more readily usable on other
platforms.
--
"I withdraw my claim that rpm is proprietary -- my objections were based
on the documentation for the version of rpm (2.2.6) that I used as a
documentation source when writing makepkg and xrpm." - david parsons
cbbr...@hex.net- <http://www.hex.net/~cbbrowne/lsf.html>
Colin R. Day
> In article <37409A01...@ix.netcom.com>,
> "Colin R. Day" <cd...@ix.netcom.com> writes:
> >> >
> >> Maybe I don't agree with everything in your statement, however the above
> >> paragraph is *exactly* one of my positions, and the fact that Linux is
> >> a gratuitious reimplementation is a macro-level example of the
> >> reimplementation that U**X developers have tried to avoid.
> >>
> >
> > Linux is not a gratuitous reimplementation of UNIX
> >
> It isn't? Remember, it is a reimplementation of UNIX, and
> the difference between it and free UNIX clones is the encumbered
> GPL. No net gain in reimplementation...
No net gain for whom? Also, what version(s) of BSD has
been ported to a Palm Pilot?
>
>
> --
> John | Never try to teach a pig to sing,
> dy...@iquest.net | it makes one look stupid
> jdy...@nc.com | and it irritates the pig.
--
Lennart Augustsson
> > > Linux is not a gratuitous reimplementation of UNIX
> > >
> > It isn't? Remember, it is a reimplementation of UNIX, and
> > the difference between it and free UNIX clones is the encumbered
> > GPL. No net gain in reimplementation...
>
> No net gain for whom? Also, what version(s) of BSD has
> been ported to a Palm Pilot?
What kind of an argument is that? BSD could be ported to
a Palm Pilot if you want to. Also, what version(s) of
Linux has been ported to a Vax?
--
-- Lennart Augustsson
[This non-signature is unintentionally not left unblank.]
te...@toybox.placo.com
>
> However, it is my personal opinion that, today, Linux still has a ways
>to go before it has the necessary "fit and finish" to start seriously
>replacing Unix in production use on servers.
>
Ah, but your missing something - people aren't out there comparing Linux against
Solaris for production systems, at least the majority of them aren't.
The only logical way to account for the explosive growth of Linux is to
understand that people are comparing Linux on servers against Windows NT
on servers!!!
If you say that Linux isn't quite ready for prime time, then can you imagine
how far behind Windows NT is?!? Compared to Windows NT and Novell NetWare,
Linux is far and away better for use on servers. At least, this is the conclusion
that I believe the majority of new Linux users are coming to. There is simply
no other server user group other than NetWare and NT that is large enough to
account for the explosive growth of Linux. Where else are they getting the
people from? I don't see hordes of people migrading from FreeBSD to Linux,
or Solaris to Linux, or some such. It's the NT and NetWare users that are
doing it.
This I believe is one thing that Linux has over FreeBSD - somehow they have
been able to tap into that subculture of Apple-Macintosh-using/VW-New-Beatle-
driving/Birkenstock-wearing computer revolutionaries and fire them up with
the ideals of smaller/faster/more flexible that is the hallmark of Unix. Now, it's
only a matter of time before I predict that we will have a major division in the
industry with the Linux/open source users and their GPL lining up on one side,
and the NT/binary-distribution-only no-source users on the other.
Ted
Chris Costello
> In article <7hpiba$mk5$1...@plonk.apk.net>, Joseph T. Adams <j...@apk.net> wrote:
> >
> >Isn't one of FreeBSD's selling points over Linux that there is a
> >relatively greater extent of central control - you can still do what
> >you want with your source, to an even greater extent than you could
> >under the GPL, but there is only one "official" release of FreeBSD?
>
> configured approximately the way you want is a 'selling point'
> for FreeBSD, what do you consider the bad points to be?
Is sysinstall not good enough for you?
> Les Mikesell
> l...@mcs.com
--
Chris Costello
And on the seventh day, He exited from append mode.
Leslie Mikesell
John S. Dyson <to...@y.dyson.net> wrote:
>I am NOT anti Linux, even though I lament the wasted effort on
>a GPLed product, but it is that GPL is such a bane on the future
>of innovative developers, that it needs to be isolated from the
>free software evolution. If anything, GPL is a stepping stone
>away from pure commercial software, but with it, additional
>limitations WORSE than commercial ensue.
I agree with this viewpoint, but I am curious why there is not
yet a 'commercially embellished' flavor of *bsd using the
free base with some of the patented and otherwise restricted
technology that will be difficult to embed in Linux, shipping
at mass market prices? Is this not a practical scenario or
hasn't the market been proven yet?
Les Mikesell
l...@mcs.com
Karri Kalpio
> a Palm Pilot if you want to. Also, what version(s) of
> Linux has been ported to a Vax?
Surprisingly, a version called "Linux/VAX".
More @ http://www.mssl.ucl.ac.uk/~atp/linux-vax/
--karri
--
You have moved your mouse, for these : Karri Kalpio
changes to take effect you must shut : ka...@moremagic.com
down and restart your computer. Do you : [+358] (9) 7017131 (home)
want to restart your computer now? : [+358] (9) 43543665 (work)
Leslie Mikesell
>: By definition, having to build it yourself isn't getting it configured the
>: way you want. Is there a freeBSD distribution where I can say 'install
>: everything', and have it fill 2 gigs or less and come up running KDE,
>: postgresql, samba, xntpd, apache and a nice collection of utilities when
>: it is done?
>
> Yes, actually. Sysinstall->Configure->Packages->All
>
> It would be as stupid to do as it is under Linux, but you can do it
> just as easily. -Then why is "All" there? Lint my friend, just
> for lint...
It is not at all stupid to do this with the RedHat 6.0 distribution,
or the Mandrake distribution which added KDE earlier. A very nice
system is created, which is the point I've been trying to make.
I'm curious as to why you think it is stupid to do it under freeBSD
and what you would actually end up with. If someone who is apparently
trying to promote freeBSD doesn't recommend it's default setup, why
would you expect anyone to like it?
>: Are there an assortment of such distributions so other people can pick the
>: ones they want too?
>
> Why? They would fit absolutely no one, ever. You'd still have to
> manually walk down the choice trees to make sure that everything you
> want is selected. -Which, I might add is quite easy in FreeBSD
> sysinstall as everything is in a simple, hierarchal, menu driven
> format.
I expect to have to add a few things that no one else would want, and
to recompile a couple to meet local needs, but beyond that, why
would you want every machine to be so drastically different?
> If you want "everything", you're welcome to it with one simple
> option. If you want to pick and choose, then pick and choose. It's
> really, *really* not that hard.
> You'll have a much easier time attacking something that you have
> actually used.
I have used sysinstall to add things to existing freeBSD systems
originally loaded by someone else. Each choice is approximately
as much effort as the whole RedHat install (yes, not really hard).
Les Mikesell
l...@mcs.com
ze...@bawdycaste.org
: I think this discussion has been interesting. However, to give the
: I send the thought police after you?
You've been reading to many rms rants.
--
-Zenin (ze...@archive.rhps.org) Caffeine...for the mind.
Pizza......for the body.
Sushi......for the soul.
-- User Friendly
Joseph T. Adams
: changed?
Much of it is, but a number of important functions, including printf
and friends and many trig/math functions, are not.
: >: Geesh, with the movement towards GPL'ing libraries,
: >: it seems that MS and GPL world is on more and more even footing.
: >
: > I understand that you don't like the GPL license and that's fine. But
: > what would possess any seemingly bright human being to suggest that an
: > open-source model (albeit one you obviously don't like) can even be
: > compared to a source-not-available-at-any-price model like the one
: > Microsoft espouses?
: >
: Your argument creates a straw position:
:
: BSD: OpenSource, value add unencumbered, free
: GPL: OpenSource, value add encumbered.
: Commercial: restricted, encumbered in many ways, but
: almost never value add encumbered in normal circumstances.
I consider both BSD- and GPL-licensed software to be free software.
The restrictions placed by the GPL are designed to guarantee that
GPL'd software and derivative works will remain free, preventing, one
would hope, things like the attempted hijacking of X by the Open
Group, and the theft of Mosaic by Microsoft.
That's a very old flame war that probably doesn't need to be repeated
here, but I'm stating my position just so it's clear. I think the GPL
is the better license in some cases, BSD in others, and perhaps the
LGPL (Mr. Stallman's desire to deprecate it notwithstanding) in yet
others.
: GPL (esp in the case of libraries) encumbers works linked to them!!!
That's why there is the LGPL, Richard Stallman's recent change of
heart notwithstanding. GPL'd libraries would be significantly less
useful than LGPL'd ones, without being any more free. I agree with
Stallman that free software is a good thing and should be encouraged.
I don't think it can or should be forced, and using the GPL for a
library would indeed give developers only two options: use the library
and release all files linked to it under the GPL, *or* don't use the
library at all. I don't think that is an acceptable choice, or one
that would encourage the development *or* use of significant GPL'd
libraries.
: That is *worse* than commercial. Without any other qualification,
: OpenSource is better than commercial. With the limitations of GPL,
: OpenSource is sometimes (often) worse than commercial.
I guess I'm not following your argument as to how. Are you saying you
can develop more kinds of software using a well-documented commercial
library than a GPL'd one? If so, yes, that is true, but libraries
typically do not use the GPL; they use the BSD, LGPL or some other
license that allows non-GPL'd (and often non-free) software to link to
it, so as to promote greater use, and development, of the library.
: >: People shouldn't be 1) mucking around with kernels, 2) mucking around
: >: with libraries, 3) mucking around with applications, 4) mucking around
: >: with user interfaces, 5) mucking around with ANY software?
: >
: > Isn't one of FreeBSD's selling points over Linux that there is a
: > relatively greater extent of central control - you can still do what
: > you want with your source, to an even greater extent than you could
: > under the GPL, but there is only one "official" release of FreeBSD?
: >
: There can easily be another "official" release of JohnBSD which is
: derived *directly* from FreeBSD. That would be *silly* without alot
: of value added, but theoretically possible. Initially, OpenBSD was
: almost a direct derivative of NetBSD, for example. The name isn't
: important when it comes to *technical* innovation.
I can see the need for a single official or "reference" implementation
of any technology. Both *BSD and Linux have this. So does all GNU
software. There are only a few really bad code forks out there
(gcc/egcs and the various libc/glibc) and those are a problem mainly
because there is some very important code (including the Linux kernel)
that relies on quirky behavior from one particular version of the
compiler or library, and therefore doesn't compile properly using
earlier *or* later versions. This is a big problem and one I hope the
kernel folks (as well as others who are in charge of broken software)
consider a high priority. Fixing a broken compiler or library won't
break good code; it only will break broken code, exposing flaws in it
that can and need to be corrected quickly.
Anyone on the kernel list have any insights/updates on this problem?
I don't follow it myself since much of it is way over my head.
: > It's perfectly acceptable now. Nothing drastically better exists yet,
: > at least not on PC-class hardware.
: >
: I suspect that nothing drastically better exists YET because of the
: giant sucking of mindshare into the Linux camp. I was working on one,
: but am too old to do it (not enough energy anymore), having to deal
: with income issues :-).
I suspect some architectures are more amenable to a microkernel or
nanokernel design than others. However, Linux was designed for, and
still runs primarily on, the Intel x86 architecture. Context switches
are very expensive, and tend to be avoided if you care at all about
performance. I think this might explain why microkernel/nanokernel
architecture is so rare on x86 machines.
It might be an entirely different story in the RISC world, about
which, unfortunately, I know little.
: > The same should be true of proprietary software as well, although if
: > you don't at least have access to source, you'll be at the mercy of
: > your vendor who may or may not choose to do so.
: >
: I like proprietary (to an extent), free, open, GPL (to an extent)
: software also. A problem exists in mistaken beliefs that GPL implies
: free. It hurts *exactly* those people who invest in it.
How?
GPL'd software is free in the same sense that people should be free:
not free to hurt others, but free to do pretty much anything else.
BSD software is free, but like other non-copylefted software, it can
be modified ever so slightly and then made non-free. Should the
non-free version become very popular, then for all intents and
purposes, the software has become non-free. That's what happened with
Mosaic => Internet Exploiter, and almost happened to X11.
People could be burned if they didn't understand the implications of
creating derived works from GPL'd software; for that reason, however,
the GPL itself goes to some lengths to spell out those implications.
I've yet to hear of anyone actually getting hosed through that kind of
misunderstanding. People who are bright enough to write good code are
usually bright enough to read and understand free software licenses.
: > What people love about Linux, as far as I can see, is not that it
: > isn't FreeBSD. (I like FreeBSD too.) What they love about it is that
: > it is free - at least, more free than what they are used to - and that
: > because it is free in the sense of liberty, it's generally of much
: > higher quality.
: >
: FreeBSD is free in the sense of liberty for the developers, marketeers,
: name recog people (we don't have any of those like in Linux though.)
: GPL makes a tradeoff in liberty *away* from those who create, with
: an unfortunately mistaken belief that they are doing as much good
: for humanity (or adding liberty.) Liberty isn't a fixed sized pie,
: but each individual has to demand that liberty once in a while, or
: will loose it. Under GPL, it is the *creative* among us who are
: disadvantaged with more limited liberty.
I just don't see it that way. No one is forced to use the GPL.
People who wish to do so, do so voluntarily. Presumably they
understand the implications, and have reasons for doing it that they
consider worthwhile.
I would generally *prefer* to develop free software under the GPL,
because I would like it to remain free.
: > The shift away from proprietary, mediocre, and highly vendor-bound
: > bloatware should benefit the *entire* free and open-source software
: > communities, and as far as I can see, it does.
: >
: Being somewhat vendor bound isn't bad, if that vendor gives a
: discount price to that customer.
Once you're too dependent, they have much less incentive to (a) keep
prices low, and (b) continue to improve the product.
Competition works wonders.
: If someone funds an entire development,
: then giving source away is much easier. However, GPL takes that
: negotiating point away. As a developer, it is best to avoid basing
: ANYTHING on GPLed works. As a user, if you have a choice between
: supported free software, or GPLed software, you are contributing much
: more to the developers well being by using and supporting free software.
: You will not have diminished your liberty, and added freedom and support
: to the developer class.
:
: REMEMBER: market value has little to do with cost of production, but
: willingness to pay. A developer can choose to make little and give
: away their liberty by developing works based upon GPL, or that developer
: can choose to make little, sell the works to customers, and have
: a chance to resell the works to other customers.
You can do that with the GPL in numerous ways. You just can't make
the derivative work of free software non-free. Anything else, pretty
much, is fair game.
Joe
Leslie Mikesell
Chris Costello <ch...@holly.dyndns.org> wrote:
>>
>> Hmmm, if you think that being unable to get a distribution
>> configured approximately the way you want is a 'selling point'
>> for FreeBSD, what do you consider the bad points to be?
>
> Is sysinstall not good enough for you?
I haven't tried the latest, but the last I saw it was a long way
from checking the '[ ] Everything' box on a RedHat distribution to
get a well-considered selection of useful apps (now including
both GNOME and KDE versions). RedHat has suited my needs pretty well;
others prefer Slackware or Caldera, etc. as a base. You can, of course
tweak things after the base install, but life is too short to assemble
a thousand components yourself or even keep track of the advances in
each of a few hundred from one release to the next. A base install
should get you within a dozen or so programs of what you need, and for
that to happen you need to have a choice of distribution configurations.
That kind of packaging could be done for freebsd too, but it doesn't
seem to mesh with the philosophy. I also don't see any integration of
components like RedHat has done. For example, how long would it take
you to configure a freebsd machine to use a remote postscript printer
served by MS-Windows print sharing? This is a fill-in-the-form setup
on RedHat.
Les Mikesell
l...@mcs.com
Wolfgang Rupprecht
> Linux should be refered to GNU/Linux
And I take it NetBSD should be called GNU/NetBSD for similar reasons???
-wolfgang
--
Wolfgang Rupprecht <wolfga...@dailyplanet.wsrcc.com>
http://www.wsrcc.com/wolfgang/
DGPS signals via the Internet http://www.wsrcc.com/wolfgang/gps/dgps-ip.html
Maury Markowitz
> Actually, on the C compiler front, it could be argued that the somewhat
> mediocre success of GCC has "starved" labour for other possible efforts for
> C compilers such as TENDRA, LCC, or other alternatives that have never even
> gotten started since GCC was "good enough."
My appologies for what is essentially a newbie question, but who do I parse
the above when considering egcs? Is egcs essentially gcc in that regard? I
was under the impression that egcs _is_ gcc3, is that correct?
> I say "mediocre success" in that there have been considerable periods of
> time during which there has been little improvement, and certainly not to
> say that GCC is "horrible/bad/evil and that the developers are morons."
I use it every day, and I can't say I'm terribly happy with it. Apple
moved to egcs over the last few months, so I'm looking forward to trying that
out. As it stands though, I find gcc to be slow and generate less than good
ppc code, so basically I'm not terribly happy with it (2.7.1 IIRC).
> The point is that there have been some possible alternatives to GCC over the
> years, and innovations that aren't compatible with the GCC architecture *do*
> tend to get stifled.
Well there's only so many developers in the world. Let's face it,
developers are a reactionary bunch, and once they gather around something it
takes a LOT to move them on. C++ is a good example.
> The best way to make the GPL "wither and vanish" is not to wail at its
> ascendancy, but to produce code under other licenses that is so good and so
> eloquently promoted that the code and user base grows with such fecundity
> that it chokes out all that "bad old GPLed code."
Wow.
Maury
Leslie Mikesell
>te...@toybox.placo.com wrote:
> >snip<
>: If you say that Linux isn't quite ready for prime time, then can you
>: imagine how far behind Windows NT is?!? Compared to Windows NT and Novell
>: NetWare, Linux is far and away better for use on servers.
>
> what you're talking about.
How much do you have to know to say it costs less and has no extra
charge for nfs and tcp services?
Les Mikesell
l...@mcs.com
Tarjei Vaagstoel
* d...@cgl.ucsf.edu
| I would call it a "pseudo-clone of POSIX", rather than BSD, since
| because there
* ram...@lycaeus.calstatela.edu
| Sure it was. Linus based his inital work on SunOS 4. (Hint: SunOS
| 4 is a BSD-ish UNIX)
I thought that he never tried to make SunOS-clone. From what I've read,
he had experience with SunOS, but he aimed for POSIX-compatibility at a
very early phase. Since he had Minix on is home box, you could just as
well say that Linux was based on Minix.
--
tarjei
Tom Wilson
> In article <x7so8u2...@dailyplanet.wsrcc.com>, Wolfgang Rupprecht wrote:
> >
> > > Linux should be refered to GNU/Linux
> >
> > And I take it NetBSD should be called GNU/NetBSD for similar reasons???
>
> I don't think NetBSD is a GNU system. I don't think NetBSD is
> GPL'd.
If I have read the article by RMS right (www.gnu.org/gnu/linux-and-gnu.html)
it has nothing to do with the GPL, but rather that many of the Linux programs
were GNU programs.
The argument about NetBSD may be right (I don't use it), but there are quite a
few important programs under FreeBSD that are GNU (gcc, ld, as) [I use both
Linux and FreeBSD]. Seems to me we should all give credit where credit is
due (which the FSF/BSD/OSS licenses do). For the FreeBSD'ers, look in /sys/gnu
and see how many of those programs that you use regularly, or how difficult
things would be without them (does *BSD have a non-GNU compiler?).
IMHO everyone should acknowledge the contribution the GNU project has made to
the Free Software community. Some may not agree with the licensing terms of
the GPL, but answer this, do you compile your software using the GNU compiler?
>
> --
> Chris Costello
> In /dev/null no one can hear you scream
--
Tom Wilson tmwi...@cyberhighway.net
"Well of course, everything looks bad if you remember it"
- Homer Simpson
Jim Richardson
John S. Dyson, in the persona of <to...@y.dyson.net>,
brought forth the following words...:
>In article <7hpp4v$3...@pell.pell.portland.or.us>,
> o r c @ p e l l . p o r t l a n d . o r . u s (david parsons) writes:
>> In article <7hp8q3$2v...@enews1.newsguy.com>,
>> John S. Dyson <to...@y.dyson.net> wrote:
>>
>>>I tend to question them also, mostly because of mis-statement. Linux
>>>is a gratuitiously rewritten U**X clone,
>>
>> If Linux was written for the explicit purpose of killing off *BSD,
>> then it would have been gratuitiously rewritten. It was not.
>> It's possible that Linus would have joined a *BSD development effort
>> when he started writing operating systems if there wasn't for the
>> teeny little detail that the million-pound hammer of an AT&T lawsuit
>> was poised over *BSD's head (I can't speak for anyone else, but
>> that lawsuit was directly responsible for my choosing to run Linux;
>> I wanted a free Unix and I didn't want to have to either abandon
>> ship or go pirate if AT&T pounded the CSRG flat) but certainly by
>> the time AT&T sold Unix to Novell that moment had passed.
>>
>That lawsuit begs the issue. Anyone knowing what was going on, wouldn't
>have worried (too much.) I know, I was there. I guess that there was
>just alot of uninformed speculation (I worked at AT&T and worked on FreeBSD
>AT THE SAME TIME.)
Perhaps it would be fair to say that you had a perspective that the rest
of us did not?
>
>>>that has a MORE restrictive
>>>license than the previously available versions.
>>
>> It's free to the end user.
>>
>It isn't free to all end users, by virtue of the restrictions against
>developers.
>
They are free to use it, if they modify and distribute it, they must also
distribute the source, it's as simple as that. If you don't like that
restriction, don't use the gpl'd code. What's so hard about that?
>>
>> That's a lot less restrictive than the
>> traditional AT&T/Sun/IBM/Digital/Silicon Graphics/etc "pay us lots
>> of money and you can use this software on one machine" approach.
>>
>Geesh, so GPL is somewhere between free and commercial? So what, it isn't
>free!!!
>
Free to me, free to many others, you might not like it, but that's the way
it is. The whole world isn't made up of people who code for megacorp
inc. Many of us do it for fun, or for projects where the code is a minor
or non-proprietary part. GPL is a great way to use and reuse ideas.
(Nothing wrong with coding for the big boys mind you, but that's
not where I am at at the moment, so I don't worry about the kind of things
they might worry about.)
>>
>> But it's not. In the free software world, a rising tide DOES
>> lift all boats, and once the user has tasted Unix it's easy
>> for them to switch between Unices. Don't quibble about who
>> brings them into the boat, just be happy that they're running
>> Unix.
>>
>The rising tide is tieing up alot of the economy, thereby destroying
>innovation. GPLers tend not to respect the developers where the innovation
>is created, while GPL certainly does help (RMS, Linus, Cygnus, Marketeers
>such as Red-Hat), but the value added developer in the middle is
>screwed (without lieing about selling software as support.)
This makes no real sense. GPL developers have the same (as a group) of
respect for the developers, as any other group. Never mind that respect
has nothing (in this case) to do with economic growth. This statement
of yours makes about as much sense as "Stealing someones culture"
it's a NOP.
--
Jim Richardson
www.eskimo.com/~warlock
All hail Eris
"Linux, where do you want to go tomorrow?"
Peter da Silva
David Konerding <d...@cgl.ucsf.edu> wrote:
>It makes one wonder: what if timing and luck had been slightly different,
>and instead of Linux catching on, {Free|Open|Net}BSD had.
>Or, if a Linux kernel was placed on to of the BSD toolbase, instead of
>the GNU toolbase?
Actually, there's a lot more commonality in the userland side of things. The
camps steal from each other all the time.. compare the tcp tools on Linux
to the BSD ones, then to System V TLI. If they're not ports, someone's
been reimplementing things pretty closely.
The problem I personally have with Linux is the general disorganization of
the core environemnt. There's been a lot of work to standardize the
distribution tree, but I really think the problem is going to be intractible:
it's inherent in the nature of the beast.
For other people that disorganization is an opportunity, a place they can add
value to the system. That's what differentiates Red Hat from Caldera and
Debian from SuSE. It's not just a bazaar, it's the marketplace in Athens
with Socrates and Plato and Aristotle haranguing the multitudes from each
their own soapbox.
I'd rather work on and use something that's got a pretty good basic design
already, where the development is mostly about new functionality rather than
new bells and whistles. But that's my bias showing.
Still, I think that the chances of Free UNIX being seen as "enterprise ready"
by the suits would be a lot beter if it were FreeBSD than Linux. There's
less for the naysayers to shoot at.
--
In hoc signo hack, Peter da Silva <pe...@baileynm.com>
`-_-' Ar rug tú barróg ar do mhactíre inniu?
'U` "Twenty-two points, plus triple-word-score, plus fifty points for
using all my letters. Game's over. I'm outta here."
Peter da Silva
>In article <7hpiba$mk5$1...@plonk.apk.net>, Joseph T. Adams <j...@apk.net> wrote:
>>relatively greater extent of central control - you can still do what
>>you want with your source, to an even greater extent than you could
>>under the GPL, but there is only one "official" release of FreeBSD?
>Hmmm, if you think that being unable to get a distribution
>configured approximately the way you want is a 'selling point'
>for FreeBSD, what do you consider the bad points to be?
Where do you get the idea that you can't get a distribution configured
approximately the way you want with FreeBSD? It's pretty easy to build a
custom distribution of FreeBSD: tune the source tree the way you want and
run a "make" in /usr/src/release.
--
Peter da Silva
>BSD software is free, but like other non-copylefted software, it can
>be modified ever so slightly and then made non-free. Should the
>non-free version become very popular, then for all intents and
>purposes, the software has become non-free. That's what happened with
>Mosaic => Internet Exploiter, and almost happened to X11.
I don't believe that Mosaic was ever "free" in the BSD sense or the GPL sense:
NCSA's licences were encumbered enough to force Netscape to start over from
scratch, for example. And that's what really killed it, not the fact that
Microsoft's specific port beat out the other commercial ports... there wasn't
a free code base to build from.
Peter da Silva
Tim Smith <t...@halcyon.com> wrote:
>In article <7hrman$3f0$1...@plonk.apk.net>, Joseph T. Adams <j...@apk.net> wrote:
>>GPL'd software and derivative works will remain free, preventing, one
>>would hope, things like the attempted hijacking of X by the Open
>>Group, and the theft of Mosaic by Microsoft.
>Mosaic is still available if you want it.
Remember the recent complaints about the Mozilla and Apple releases not
really being open source? They were a lot more open than this:
http://www.ncsa.uiuc.edu/SDG/Software/Mosaic/License/LicenseInfo.html:
The source code for the X Windows System version of Mosaic is available
under copyright for individual personal use, for use at an academic
institution or for internal business use. For those who are using the X
Windows System Mosaic source for these non-commercial purposes, the
following applies: "Licensee may make derivative works. However, if
Licensee distributes any derivative work based on or derived from the
Software, then Licensee will (1) notify NCSA regarding its distributing
of the derivative work, and (2) clearly notify users that such
derivative work is a modified version and not the original NCSA Mosaic
distributed by the UI." As a result of this copyright restriction, if
any product, including a free one like a research version or derivative
of NCSA Mosaic, is created using NCSA Mosaic source code and that
product is released to other users, then the creator of the product
must both (1) notify NCSA and, (2) let the users know that the product
is a derivative and not the original NCSA Mosaic. NCSA cannot support
derivative products and users must be made aware of the distinction
between the original NCSA Mosaic and a product created by someone else.
Any derivative product is also, of course, subject to the restrictions
of the copyright, including distribution and use limitations.
The source code for the Microsoft Windows and Macintosh versions is
also available, but only to academic institutions (universities) and
United States government agencies for internal use. (See information
below entitled "Application for Microsoft Windows and Macintosh Source
Code"). All other uses of the source code requires licensing, which
includes a fee payment. Contact Spyglass at mos...@spyglass.com or
708-505-1010 if interested in licensing.
Christopher Browne
<ch...@holly.dyndns.org> wrote:
>In article <x7so8u2...@dailyplanet.wsrcc.com>, Wolfgang Rupprecht wrote:
>> > Linux should be refered to GNU/Linux
>>
>> And I take it NetBSD should be called GNU/NetBSD for similar reasons???
>
> I don't think NetBSD is a GNU system. I don't think NetBSD is
>GPL'd.
I think Linus said it quite well...
--
(About "LyGNUx") - "Your midwife doesn't select the name of
your babies." -- Linus Torvalds, May 5, 1999, ABC News
cbbr...@hex.net- <http://www.hex.net/~cbbrowne/lsf.html>
Christopher Browne
<maury@remove_this.sympatico.ca.invalid> wrote:
>In <slrn7k1k4f....@knuth.brownes.org> Christopher B. Browne wrote:
>> Actually, on the C compiler front, it could be argued that the somewhat
>> mediocre success of GCC has "starved" labour for other possible efforts for
>> C compilers such as TENDRA, LCC, or other alternatives that have never even
>> gotten started since GCC was "good enough."
>
> My appologies for what is essentially a newbie question, but who do I parse
>the above when considering egcs? Is egcs essentially gcc in that regard? I
>was under the impression that egcs _is_ gcc3, is that correct?
See: <http://egcs.cygnus.com/gcc-2.95/schedule.html>
It indicates that the transition from EGCS to GCC 2.95 is anticipated to
take place in July 1999.
>> I say "mediocre success" in that there have been considerable periods of
>> time during which there has been little improvement, and certainly not to
>> say that GCC is "horrible/bad/evil and that the developers are morons."
>
> I use it every day, and I can't say I'm terribly happy with it. Apple
>moved to egcs over the last few months, so I'm looking forward to trying that
>out. As it stands though, I find gcc to be slow and generate less than good
>ppc code, so basically I'm not terribly happy with it (2.7.1 IIRC).
A *lot* of stuff has been happening with EGCS, which suggests that EGCS
may be quite a bit better.
>> The point is that there have been some possible alternatives to GCC over the
>> years, and innovations that aren't compatible with the GCC architecture *do*
>> tend to get stifled.
>
> Well there's only so many developers in the world. Let's face it,
>developers are a reactionary bunch, and once they gather around something it
>takes a LOT to move them on. C++ is a good example.
A standards-compliant implementation of C++ is a *big project,* and
changing the direction of people in an effort to start something new
"from scratch," as it were, requires a *lot* of pull.
>> The best way to make the GPL "wither and vanish" is not to wail at its
>> ascendancy, but to produce code under other licenses that is so good and so
>> eloquently promoted that the code and user base grows with such fecundity
>> that it chokes out all that "bad old GPLed code."
>
> Wow.
Note that I'm not suggesting that this outcome is *likely,* just that
it's more likely to provide useful outcomes than whining about how bad
the GPL is.
--
Intel engineering seem to have misheard Intel marketing strategy. The
phrase was "Divide and conquer" not "Divide and cock up"
(By iia...@www.linux.org.uk, Alan Cox)
cbbr...@hex.net- <http://www.ntlug.org/~cbbrowne/langc.html>
Christopher B. Browne
License price is only one metric; the fact that NetWare servers *commonly*
run for *months* whilst undergoing heavy file-serving loads indicates that
in the matter of robustness, NetWare is no slouch.
Jesus Monroy, Jr.
Keats) wrote:
>In article <37406690...@mitre.org>,
> David Burgess <bur...@mitre.org> writes:
>......[snip].......
>A comparison of the time lines of the legal dispute, and the
>development/distribution from BSDi might prove interesting.
>I believe that v. 1.1 was BSDi's first distro based on 4.4Lite
>(the one with the encumbered code stripped-out.)
> ^^^
supposedly emcumbered
As this was settled out of court, not one (entity) took
blame or was proven wrong.
--
If you have to read the docs, it's broken.
I hate making mistakes.
You can check my spelling at: http://work.ucsd.edu:5141/cgi-bin/http_webster
Colin R. Day
> "Colin R. Day" <cd...@ix.netcom.com> writes:
>
> > > > Linux is not a gratuitous reimplementation of UNIX
> > > >
> > > It isn't? Remember, it is a reimplementation of UNIX, and
> > > the difference between it and free UNIX clones is the encumbered
> > > GPL. No net gain in reimplementation...
> >
> > No net gain for whom? Also, what version(s) of BSD has
> > been ported to a Palm Pilot?
> What kind of an argument is that? BSD could be ported to
> a Palm Pilot if you want to. Also, what version(s) of
> Linux has been ported to a Vax?
>
Wheteher Linux has been ported to VAX or not doesn't matter.
If Linux is the only UNIX/UNIX clone to run on a Palm Pilot,
then it isn't gratuitous. As long as Linux is the only *NIX to
run on some architecture, it is not gratuitous.
Also, what architectures does BSD run on?
>
> --
>
> -- Lennart Augustsson
> [This non-signature is unintentionally not left unblank.]
--
Michael Maxwell
> On Tue, 18 May 1999 21:42:53 GMT, Chris Costello
> <ch...@holly.dyndns.org> wrote:
> >In article <x7so8u2...@dailyplanet.wsrcc.com>, Wolfgang Rupprecht wrote:
> >> > Linux should be refered to GNU/Linux
> >>
> >> And I take it NetBSD should be called GNU/NetBSD for similar reasons???
> >
> > I don't think NetBSD is a GNU system. I don't think NetBSD is
> >GPL'd.
>
> I think Linus said it quite well...
I think the whole thing is more a matter of a certain someone trying to
foist his philosophy on us again by starting idiotic battles about whether
or not it should be called GNU or Linux or GNU/Linux or GPL-GNU-FSF-RMS-ESR-
Linus-"Cap'n"-Torvalds-Magical-Mystery-Extra-Crispy-O.S.
The bottom line is simply: "Who the HELL gives a damn?!?" I believe credit
should be given where due, and much credit to FSF. But Linux itself was NOT
an FSF creation and as such should not be referred to as such...
and so on...
--
Michael Maxwell <drwho @ xnet.com> | http://www.xnet.com/~drwho/
-- Stop the illegal attacks on Serbia NOW! --