The state of FreeBSD
John Nielsen
> I've been teetering on the edge for some time now. I almost switched a
> couple weeks ago, but I wasn't sure if I'd be able to my purchased
> software running. Namely Crossover Plugin (not sure), Quanta Gold (works)
> and StarOffice 6 (which I bought from SuSE, thus I wasn't sure if it would
> work with FreeBSD - http://makeashorterlink.com/?G3F0523A1).
>
> But I think I'm close to being pushed to using it as my main desktop. I've
> used it for some time on a second computer and I really like it so far.
> The only sticking point being the software mentioned. This week, though,
> I've had a problem with SuSE that has me near the edge. Namely that I find
> out that updates such as Mozilla 1.0, Gnome 2, etc. won't be available via
> SuSE's standard update method. Now I have people trying to convince me to
> install Apt4RPM. If I'm going to have to update from here on out with
> apt4rpm, I'm just going to use FreeBSD and ports, I think.
>
> So the nature of my query is this. Given that one of the biggest
> challenges in the Linux/FreeBSD world seems to be keeping your system up
> to date, I'm wondering what the state of the FreeBSD community is right
> now. I've heard a lot about defections and grumblings of problems with the
> organization and these things concern me, because above all else whatever
> Unix-like OS I choose to use as my desktop, the OS has to be easy to keep
> up to date and secure.
>
> Any thoughts? This may be too broad of a topic, but hopefully someone can
> help me. My impression thus far is that the technology is great, but I
> fear that with the rising popularity of Linux, that there might not be as
> many committed people around to make sure that the ports are in a state
> that makes them a better choice than going with a Linux distro.
Contrary to whatever the trolls have been saying, FreeBSD is not going away
any time soon. It is still being very actively developed, and it is
extremely easy to keep up-to-date and secure. It looks like you already
recognize the value of the ports system. I don't know if your purchased
software will work, but since it'll be worth your time to try FreeBSD and
see how you like it, you can tell us.
JN
--
Remove pig-latin to reply by e-mail
Mark Hittinger
> there might not be as
>many committed people around to make sure that the ports are in a state
>that makes them a better choice than going with a Linux distro.
Quantity does not mean quality :-)
Later
Mark Hittinger
bu...@pu.net
John S. Dyson
"Preston Crawford" <pres...@serpentor.cobrala> wrote in message news:slrnamt1c7....@serpentor.cobrala...
>
> Any thoughts? This may be too broad of a topic, but hopefully someone can
> help me. My impression thus far is that the technology is great, but I
> that makes them a better choice than going with a Linux distro.
>
disagreement. It will continue in the future. Even though I disagree with
many of the strategic decisions of the past, and in a way, the project has
been hurt by some past decisions, the project will prevail. FreeBSD is
too big for a few problems to destroy it, and it is too good (technically)
to be dismissed EVEN in a politically correct environment (both of the
Microsoft or Linux style of political correctness.)
John
Bill Vermillion
Preston Crawford <pres...@serpentor.cobrala> wrote:
>So the nature of my query is this. Given that one of the biggest
>challenges in the Linux/FreeBSD world seems to be keeping your
>system up to date, I'm wondering what the state of the FreeBSD
>community is right now. I've heard a lot about defections and
>grumblings of problems with the organization and these things
>concern me, because above all else whatever Unix-like OS I
>choose to use as my desktop, the OS has to be easy to keep up to
>date and secure.
>Any thoughts?
Compare the 70-100 globally located sites that have ftp sources
for the complete system and almost as many cvs sites I think you'll
find the number of sites dedicated to BSD is far more than any
single Linux distribution you'll find.
Look at the sites that are providing the support. The master
site is owned and operated by the Danish Telephone Company for
example.
>This may be too broad of a topic, but hopefully someone can help
>me. My impression thus far is that the technology is great, but
>I fear that with the rising popularity of Linux, that there
>might not be as many committed people around to make sure that
>the ports are in a state that makes them a better choice than
>going with a Linux distro.
Sound like someone has been filling you with FUD. The number of
ports are increasing daily. As of yesterday there are 7522 in
my /usr/ports tree. That's one more than the day before :-)
I just happened to have the last two days emails to root
from my portsdb program which is run by cron nightly.
Bill
--
Bill Vermillion - bv @ wjv . com
Donn Miller
> FreeBSD (like any other growing entity) has had various kinds of
> disagreement. It will continue in the future. Even though I disagree with
> many of the strategic decisions of the past, and in a way, the project has
> been hurt by some past decisions, the project will prevail. FreeBSD is
> too big for a few problems to destroy it, and it is too good (technically)
> to be dismissed EVEN in a politically correct environment (both of the
> Microsoft or Linux style of political correctness.)
Yeah, that's what I thought as well. When I tried Gentoo, I was kind
of disappointed. The multi-tasking was nowhere as good as FreeBSD's,
and the way Linux caches a whole gob of memory and then flushes giant
chunks of it to swap really was a downer. I saw on kerneltrap that
there was consideration of a dynamically-shrinking cache, and this
sounds like something FreeBSD has been doing all along.
There does seem to be a sort of political correctness about Linux,
more than anything else. It seems like it is very self-promoting in
the way it has all these people contributing patches to the Linux
kernel. It makes Linux seem more "open", as there is an image of
FreeBSD being too "closed" with a smaller core team of developers.
Then, there is the GPL, which is supposed to be "truly open",
according to Linux-GPL advocates, because it "guarentees" that the
source is always available due to the GPL. According to what some Linux-GPL
advocates were saying, FreeBSD is actually "closed" because of the
smaller core team and the fact that the average person is not allowed
to have patches committed to the tree as in Linux. Those people point
to the fact that Linux has more developers (essentially everyone), and
because of this, Linux has better support for threading.
I think the real issue is mindshare, which in turn is due to the
popular icon-like status of RMS and the GPL, which sort of implies and
purports to "save computing-kind" of the likes of crass
commercialization and dominance of large software companies such as
Microsoft. It's also very popular amoung Linux advocates to joke that
the Win* kernels have "plenty of BSD code in them". One other joke
states that Microsoft is a development playground for BSD.
But then, when you cut through all the cruft, you realize that Linux
really doesn't have a lot of the tight quality control of FreeBSD.
For example, there is a tight relationship between the FreeBSD kernel
guys and the people maintaining moused and syscons. Moused has been
very excellent in terms of stability. GPM, OTOH, tends to be very
quirky, erratic, and unstable. The reason for this is that the Linux
kernel team is a little too isolated from the rest of the Linux
developers, for example the developers of GPM. Also, Linux tends to
use a lot of GNU software all over the place in the userland. The
result, it seems, is that each development group is isolated from each
other by really thick walls, and there's little communication between
each group. It's just too fragmented. It seems as though Linux in
general is too fragmented: the development groups are fragmented, and
the distros are fragmented. Also, why does Linux have so much
information exported to /proc? Maybe it's to reduce the dependance of
the Linux userland on the kernel? I've seen some cases where you can
even set Linux system parameters by writing to certain files on /proc.
OTOH, Linux IS a free unix, and as such, you can do many things with
it that wouldn't be possible with other OSes (such as Windows). Also,
the fragmentation might actually be a plus in certain cases. For
example, there's not as great a dependance of the userland on the Linux
kernel, as opposed to FreeBSD, where the userland does depend on the
kernel to a greater extent.
-----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----
http://www.newsfeeds.com - The #1 Newsgroup Service in the World!
-----== Over 80,000 Newsgroups - 16 Different Servers! =-----
Mike
<pres...@serpentor.cobrala> wrote:
>So the nature of my query is this. Given that one of the biggest
>challenges in the Linux/FreeBSD world seems to be keeping your system up
>to date, ...
In actuality, I had that exact problem with RH Linux (i.e., difficulty
keeping the system up to date), which is why I switched to FreeBSD.
I find it rather easy to keep my FreeBSD system up to date.
Perhaps you've been handed an incorrect "Given"?
TomFC
> I've been teetering on the edge for some time now. I almost switched a
> couple weeks ago, but ...
[snippage]
While I wouldn't say I have "switched" to FreeBSD, I did install it on a
spare hard drive for my laptop a few weeks ago. Surprisingly easy. Not as
wiz-bang slam-dunk easy an install as RedHat, certainly, but not far from
that.
I am fairly sure StarOffice 6 or OpenOffice is available as a precompiled
package. I've seen it mentioned somewhere in one of the FreeBSD newsgroups.
I have Crossover Plugin and Crossover Office but I have not yet tried to
install them on FreeBSD (the installers are sitting on my other hard drive
which, obviously, isn't in the laptop right now ;-)
> So the nature of my query is this. Given that one of the biggest
> challenges in the Linux/FreeBSD world seems to be keeping your system up
> to date, I'm wondering what the state of the FreeBSD community is right
> now.
I wish I could speak authoritatively on that. But I am new to the FreeBSD
game, as well.
All I can say is what my general impressions are. And that is that Linux is
going to "win" since it has enormous momentum, the GPL is very popular, and
it's in the board rooms now. But -- so what? FreeBSD will run virtually all
that cool Linux software and then some. FreeBSD isn't going to disappear,
either. It's too well put together and the people working on it are doing
too good a job.
But ... that's just my opinion.
I will say this, though. Lots of nice people in the FreeBSD world. Very
helpful, none of that "stop trolling" bullshit I see CONSTANTLY in the
Linux newsgroups, and I have yet to see someone jump down the throat of a
newbie.
:-)
Seems a substantially more mature camp.
Donn Miller
> All I can say is what my general impressions are. And that is that Linux is
> going to "win" since it has enormous momentum, the GPL is very popular, and
> it's in the board rooms now. But -- so what? FreeBSD will run virtually all
> that cool Linux software and then some. FreeBSD isn't going to disappear,
> either. It's too well put together and the people working on it are doing
> too good a job.
It's still a win for FreeBSD, no matter what, because most of the Linux
software compiles on FreeBSD, and a lot of it is already in ports.
Also, the binary closed source software for Linux runs extremely well
under Linux emulation. So, FreeBSD still benefits, except for
board-room discussions. I think that a lot of companies are getting the
idea that they should be "making the switch to Linux" because they are
tired of paying license fees and such. Here, FreeBSD doesn't even get
mentioned a lot of times because Linux has the lead in OSS mindshare.
If you ask the average user about open source software, he almost
automatically brings up Linux, not BSD. Remind the port maintainer for
opensource_mindshare to fix that.
Donn Miller
> Sound like someone has been filling you with FUD. The number of
> ports are increasing daily. As of yesterday there are 7522 in
> my /usr/ports tree. That's one more than the day before :-)
I dunno. The number of ports in Gentoo's portage collection is
increasing as well. And Gentoo has an extremely spiffy init system to
boot. Someone in COLA said that Gentoo makes FreeBSD obsolete, and that
FreeBSD will lose developer support because the GPL guarantees that the
source will live on forever until the end of time. Yet another Linux
advocate claimed that BSD will be "dead and gone" in 20 or so years, but
Linux will live forever because of the GPL. And, of course, don't even
dare try to criticize Richard Stallman, GNU, or the GPL in COLA, because
"he's the guy that's responsible for giving the world open source".
Nick Hilliard
> Linux emulation. So, FreeBSD still benefits, except for board-room
> discussions. I think that a lot of companies are getting the idea that
> they should be "making the switch to Linux" because they are tired of
> paying license fees and such. Here, FreeBSD doesn't even get mentioned
> a lot of times because Linux has the lead in OSS mindshare.
While this is the case in the US and Western Europe, FreeBSD is proportionally
much more popular in places like Eastern Europe, the ex-Soviet areas and Japan.
Nick
Rainer Duffner
> other by really thick walls, and there's little communication between
> each group. It's just too fragmented. It seems as though Linux in
> general is too fragmented: the development groups are fragmented, and
> the distros are fragmented.
Hence: United Linux.
:-)
Positive things about Linux:
- runs 3d-games
- runs vmware 3
- runs some commercial enterprise sw (oracle, notes, novell, SAP)
that otherwhise would end up on an nice but expensive SPARC or
on a Wintel-Box.
For the rest, FreeBSD is much better.
It's commercial-grade (though that might actually be an insult,
judging from the state of some commercial software...).
Rainer
Donn Miller
> Positive things about Linux:
> - runs 3d-games
> - runs vmware 3
> - runs some commercial enterprise sw (oracle, notes, novell, SAP)
> that otherwhise would end up on an nice but expensive SPARC or
> on a Wintel-Box.
Well, Linux might be better at threading. Xine worked very well for me
on Linux, but on FreeBSD, doesn't work very well at all. It may be a
problem with threads, and I think Linux is more thread-safe overall.
Also, hmm... well, there's a lot more kernel options with Linux, I suppose.
> For the rest, FreeBSD is much better.
> It's commercial-grade (though that might actually be an insult,
> judging from the state of some commercial software...).
Yeah. I've seen some Linux people complain about HP-UX, although I've
never used it myself. So I can't comment. I'm guessing Solaris is
pretty decent, though.
Dave Pimlott
>
> Yeah. I've seen some Linux people complain about HP-UX, although I've
> never used it myself. So I can't comment. I'm guessing Solaris is
> pretty decent, though.
I have used FreeBSD, Solaris and Linux (in that order) and lately I have
been inflicted with HP-UX 11. it is the biggest pile of crap I have ever
had the displeasure to work with...
For example: it's sysV based (which I have no problems with) but logs are
in /var/adm/syslog/<log_files_here> but the startup logs are else where
(somewhere in /etc IIRC). /etc has symlinks to binaries in /sbin. Init
scripts (the equivalent to Solaris /etc/init.d/*) can be found in
/sbin/init.d/. HP-UX is just plain *wrong*.
It also takes ages to boot up - the startup time can be counted in ice
ages!!
If anyone has stories or reasons why HP-UX is good, I would be very
interested in hearing about them (probably best off-list).
Dave Pimlott.
---
to reply direct, just hit reply.
David Schultz
> There does seem to be a sort of political correctness about Linux,
> more than anything else. It seems like it is very self-promoting in
> the way it has all these people contributing patches to the Linux
> kernel. It makes Linux seem more "open", as there is an image of
> FreeBSD being too "closed" with a smaller core team of developers.
> Then, there is the GPL, which is supposed to be "truly open",
> according to Linux-GPL advocates, because it "guarentees" that the
> source is always available due to the GPL. According to what some Linux-GPL
> advocates were saying, FreeBSD is actually "closed" because of the
> smaller core team and the fact that the average person is not allowed
> to have patches committed to the tree as in Linux. Those people point
> to the fact that Linux has more developers (essentially everyone), and
> because of this, Linux has better support for threading.
The ``openness'' comment you make originates with some FUD preached by
Eric Raymond. Although he has some good insights into the Linux
development model, his understanding of the FreeBSD model is flat out
wrong. In particular, he likes to make distinctions that don't exist.
In the most simplistic terms, you develop for Linux by sending a patch
to one person, whereas you develop for FreeBSD by submitting a patch
to the developer community as a whole. Despite what ESR says, the
average person *can* contribute, as many people have. Grep the CVS
logs for 'Submitted by'.
Alexander Viro
David Schultz <das...@HAL9000.wox.org> wrote:
>
>The ``openness'' comment you make originates with some FUD preached by
>Eric Raymond. Although he has some good insights into the Linux
>development model, his understanding of the FreeBSD model is flat out
No, he hasn't. Witness his, er, lack of success with getting his
configurator toy merged. That was bloody spectacular - especially
the arguments he had used and style of his advocacy. And utter
lack of clue regarding the effect he was making...
He'd managed to piss off pretty much everybody in the core team and
convince all of us that he was a loon. Then he proceeded to vague
threats - with very predictable result. "Hacker of social systems",
my arse...
Believe me, his "insights" into Linux development are worth exactly
the same as his opinion on *BSD. He's occasionally amusing, but
everything he says should be taken about as seriously as sayings of
Zippy The Pinhead. Be glad that usually he chooses to grace Linux
with his, ahem, attention - at least that takes him off your backs...
--
"You're one of those condescending Unix computer users!"
"Here's a nickel, kid. Get yourself a better computer" - Dilbert.
Bill Vermillion
Alexander Viro <vi...@weyl.math.psu.edu> wrote:
>In article <akmln9$1581$1...@agate.berkeley.edu>,
>David Schultz <das...@HAL9000.wox.org> wrote:
>>The ``openness'' comment you make originates with some FUD preached by
>>Eric Raymond. Although he has some good insights into the Linux
>>development model, his understanding of the FreeBSD model is flat out
...
>Believe me, his "insights" into Linux development are worth exactly
>the same as his opinion on *BSD. He's occasionally amusing, but
>everything he says should be taken about as seriously as sayings of
>Zippy The Pinhead. Be glad that usually he chooses to grace Linux
>with his, ahem, attention - at least that takes him off your backs...
He hasn't changed much since I first saw his postings back in '85.
There is something to be said for consistancy :-)
david parsons
>> There does seem to be a sort of political correctness about Linux,
>> more than anything else. It seems like it is very self-promoting in
>> the way it has all these people contributing patches to the Linux
>> kernel. It makes Linux seem more "open", as there is an image of
>> FreeBSD being too "closed" with a smaller core team of developers.
>The ``openness'' comment you make originates with some FUD preached by
>Eric Raymond. Although he has some good insights into the Linux
>development model,
No, he got that wrong too.
But he got a good soundbite, which turned out to be good for
marketing the OS.
____
david parsons \bi/ It would have been a more compelling argument if
\/ 99% of the Linux distributions weren't all using
the Latest! And! Greatest! GNU bloatware.
John S. Dyson
"David Schultz" <das...@HAL9000.wox.org> wrote in message news:akmln9$1581$1...@agate.berkeley.edu...
> Thus spake Donn Miller <dmmi...@acs-24-154-77-97.zoominternet.net>:
> > There does seem to be a sort of political correctness about Linux,
> > more than anything else. It seems like it is very self-promoting in
> > the way it has all these people contributing patches to the Linux
> > kernel. It makes Linux seem more "open", as there is an image of
> > FreeBSD being too "closed" with a smaller core team of developers.
> > Then, there is the GPL, which is supposed to be "truly open",
> > according to Linux-GPL advocates, because it "guarentees" that the
> > source is always available due to the GPL. According to what some Linux-GPL
> > advocates were saying, FreeBSD is actually "closed" because of the
> > smaller core team and the fact that the average person is not allowed
> > to have patches committed to the tree as in Linux. Those people point
> > to the fact that Linux has more developers (essentially everyone), and
> > because of this, Linux has better support for threading.
>
> The ``openness'' comment you make originates with some FUD preached by
> Eric Raymond. Although he has some good insights into the Linux
> development model, his understanding of the FreeBSD model is flat out
> wrong. In particular, he likes to make distinctions that don't exist.
>
nonsense coming from certain parts of the 'open' software movement, they
ended up becoming the 'darker' side of the force. FreeBSD was very inept
at countering the nonsense about the disinformation against it by the
'darker' side. When standing back away from FreeBSD, and the whole
'open' software 'movement', I had realized that too much of the GPL advocacy
was being purely dishonest, either by ignorance or by intent. The
whole thing about calling GPLed works 'free software' is another bit
of disinformation. Perhaps in poetic use, it is 'freer' than some kinds of
commercial software, but redistribution encumberances associated with
the so called 'free' software is a very clear case of double speak at best,
and at worst, absolutely dishonest.
All too often the GPL advocacy (not the users, but the ADVOCACY) would
continually try to explain away the reasons why their own choices of restrictions
made software a better form of free. For all of their protestations, the clear
thinking people can see that they were seriously attempting to define away
freedom/freeness so that they could use the term 'free.'
Using the GPL as a commercial, restricted license is very consistent with
being an honest person. Being deceived by GPL advocacy is still consistant
with being an honest person. Knowing the fact that GPL does apply some
restrictions on the redistribution of software, and continuing to claim that
GPLed works are generally 'free' is rather dishonest. This dishonesty
(or short-sightedness) is the kind of problem that the truer free software
developments have had to fight, when showing justification for their free
software licenses. As soon as someone starts trying to discriminate
against groups (or companies), in the use(reuse) of software, then that
shows a very selective kind of freedom and freeness.
So, the mischaracterization of Linux or FreeBSD development isn't really
a new thing -- it has been happening against FreeBSD for a long time
now, by those who are more associated with the 'darker' side of open
software.
John
Donn Miller
> All too often the GPL advocacy (not the users, but the ADVOCACY) would
> continually try to explain away the reasons why their own choices of restrictions
> made software a better form of free. For all of their protestations, the clear
> thinking people can see that they were seriously attempting to define away
> freedom/freeness so that they could use the term 'free.'
>
> Using the GPL as a commercial, restricted license is very consistent with
> being an honest person. Being deceived by GPL advocacy is still consistant
> with being an honest person. Knowing the fact that GPL does apply some
> restrictions on the redistribution of software, and continuing to claim that
> GPLed works are generally 'free' is rather dishonest.
The biggest justification I've heard recently is that Microsoft can suck
in all the BSD code they want without contributing anything back to
*BSD. The GPL is supposed to prevent this, they claim. But I countered
by stating that why make MS an example for what you can and can't do
with BSD vs. GPL code? Why punish everyone just because GPL advocates
are afraid Microsoft is going to wind up using the code? Who cares?
Like I pointed out earlier, it's a joke among Linux/GPL advocates and/or
RMS worshippers that the "Win* kernel has plenty of *BSD code". And
this Linux is supposed to set everyone free from MS' code theft because
of the GPL.
Also, a few Linux advocates who FUD FreeBSD often quote Eric Raymond.
pat
Help is always there most of the time for newbie's and for the experts as
well. There is a fair amount of documentation and the OS is stable and stood
the test of time I think. There is very little ranting at newbie's.
Criticism and pointing out the short comings of FBSD does not launch flame
wars in most cases but starts constructive conversations. I think as someone
else stated "mature" user for the most part.
IMHO as a newbie myself that the process of updating is not as easy as some
state, if you do it correctly. It's the merge process that might trip me up
a bit. Gathering an updated source and compiling it I found to be quite
simple. It was when I tried to complete the merge process that was extremely
difficult, complicated and time consuming to do correctly. It requires a
fair amount of knowledge of Unix as well as FBSD was my experience.
I love the ports collection however it's not always paradise. For the simple
install it works great most of the time. I have at times found yourself
involved in complicated installation of different apps with
interdependencies among each other that became "Technical Challenging" but
that might be true on any OS. I found myself having to learn how the ports
process works in detail to over come this in some cases.
In short , it's not peaches and cream but I am very happy with it. I have
not burned my other OS's nor will I. It is my experience that different OS's
soot some of my needs better for various reasons.
Very Happy FBSD User
"Preston Crawford" <pres...@serpentor.cobrala> wrote in message
news:slrnamt1c7....@serpentor.cobrala...
> I've been teetering on the edge for some time now. I almost switched a
> couple weeks ago, but I wasn't sure if I'd be able to my purchased
> software running. Namely Crossover Plugin (not sure), Quanta Gold (works)
> and StarOffice 6 (which I bought from SuSE, thus I wasn't sure if it would
> work with FreeBSD - http://makeashorterlink.com/?G3F0523A1).
>
> But I think I'm close to being pushed to using it as my main desktop. I've
> used it for some time on a second computer and I really like it so far.
> The only sticking point being the software mentioned. This week, though,
> I've had a problem with SuSE that has me near the edge. Namely that I find
> out that updates such as Mozilla 1.0, Gnome 2, etc. won't be available via
> SuSE's standard update method. Now I have people trying to convince me to
> install Apt4RPM. If I'm going to have to update from here on out with
> apt4rpm, I'm just going to use FreeBSD and ports, I think.
>
> So the nature of my query is this. Given that one of the biggest
> challenges in the Linux/FreeBSD world seems to be keeping your system up
> to date, I'm wondering what the state of the FreeBSD community is right
> now. I've heard a lot about defections and grumblings of problems with the
> organization and these things concern me, because above all else whatever
> Unix-like OS I choose to use as my desktop, the OS has to be easy to keep
> up to date and secure.
>
> Any thoughts? This may be too broad of a topic, but hopefully someone can
> help me. My impression thus far is that the technology is great, but I
> fear that with the rising popularity of Linux, that there might not be as
> many committed people around to make sure that the ports are in a state
> that makes them a better choice than going with a Linux distro.
>
> Preston
>
Mark Hittinger
>The biggest justification I've heard recently is that Microsoft can suck
>in all the BSD code they want without contributing anything back to
>*BSD. The GPL is supposed to prevent this, they claim.
The true reason why Micro$oft can (and does) do this is that the BSD-4.4 lite
code is recognized as "unencumbered" by any ATT/USL/SYSV trade secret or
patent claims. This is one of *BSD's great assets and is a huge long term
advantage that the Linux side of the house does not have.
Linux has no such protection against a trade secret or patent claim from the
surviving SYSV cabal (hp/sco).
One day the commercial Linux houses will have to face the lawyers. As revenue
for the proprietary SYSV systems continues to decline this day fast approaches.
Later
Mark Hittinger
bu...@pu.net
Richard Tobin
John S. Dyson <dy...@iquest.net> wrote:
>Perhaps in poetic use, it is 'freer' than some kinds of
>commercial software, but redistribution encumberances associated with
>the so called 'free' software is a very clear case of double speak at best,
>and at worst, absolutely dishonest.
It's neither double speak nor dishonest. GPL supporters just disagree
with you about what constitutes freedom. What it seems to boil down
to is that they think it's an important aspect of freedom that their
code can't be made proprietary, and you think it's important that it
*can* be made proprietary. No dishonesty, just a disagreement.
-- Richard
--
Spam filter: to mail me from a .com/.net site, put my surname in the headers.
FreeBSD rules!
Michael Sierchio
> ... All too often the GPL advocacy (not the users, but the ADVOCACY) would
> continually try to explain away the reasons why their own choices of restrictions
> made software a better form of free. ...
Having read George Orwell, you were not taken in. ;-)
Donn Miller
> Linux has no such protection against a trade secret or patent claim from the
> surviving SYSV cabal (hp/sco).
>
> One day the commercial Linux houses will have to face the lawyers. As revenue
> for the proprietary SYSV systems continues to decline this day fast approaches.
Uh huh uh huh. And, as Linux's popularity starts to skyrocket among
various commercial vendors, software patents are going to become common.
RedHat has already patented some of Ingo Molinar's work. AFAIK, he's
doing some very good work. However, it can't be considered free if
patents are going to be applied towards parts of the Linux kernel. It
may be GPL'd, but if it's GPL'd AND patented, well then that's something
entirely different altogether.
http://kerneltrap.org/node.php?id=388
I can see where the GPL can foster software patents. Normally, when
software companies take the BSD code, and incorporate it into their own
products with changes, they don't have to make the changes available to
everyone, as the GPL stipulates. So, there's no need for a patent. But
since the GPL states that any modifications to the GPL'd code be made
public, well, I can see the need for additional protection by patenting
the code. Too bad the GPL doesn't state anything about patents. Here,
the complexity of the GPL works against itself, and the simplicity of
the BSDL is the "winner".
John S. Dyson
"Richard Tobin" <ric...@cogsci.ed.ac.uk> wrote in message news:ako737$1i0p$1...@pc-news.cogsci.ed.ac.uk...
> In article <3MMb9.1207$I4.2...@news.iquest.net>,
> John S. Dyson <dy...@iquest.net> wrote:
>
> >Perhaps in poetic use, it is 'freer' than some kinds of
> >commercial software, but redistribution encumberances associated with
> >the so called 'free' software is a very clear case of double speak at best,
> >and at worst, absolutely dishonest.
>
> It's neither double speak nor dishonest. GPL supporters just disagree
> with you about what constitutes freedom.
>
doublespeak. It is as nearly a self-evident proof as possible.
John
Marc Spitzer
> In article <3MMb9.1207$I4.2...@news.iquest.net>,
> John S. Dyson <dy...@iquest.net> wrote:
>
>>Perhaps in poetic use, it is 'freer' than some kinds of
>>commercial software, but redistribution encumberances associated with
>>the so called 'free' software is a very clear case of double speak at best,
>>and at worst, absolutely dishonest.
>
> It's neither double speak nor dishonest. GPL supporters just disagree
> with you about what constitutes freedom. What it seems to boil down
> to is that they think it's an important aspect of freedom that their
> code can't be made proprietary, and you think it's important that it
> *can* be made proprietary. No dishonesty, just a disagreement.
When you use a definition of free that is not in the dictionary and
conflicts with those that are in the dicionary it is hard to give
credit for good motives.
marc
Paulie
Donn Miller <dmmi...@cvzoom.net> wrote:
> Uh huh uh huh. And, as Linux's popularity starts to skyrocket among
> various commercial vendors, software patents are going to become common.
> RedHat has already patented some of Ingo Molinar's work. AFAIK, he's
> doing some very good work. However, it can't be considered free if
> patents are going to be applied towards parts of the Linux kernel.
Why not? If you are allowed to use it for GPL'ed s/ware for nothing,
but have to pay for commercial use, then what's the problem?
> It may be GPL'd, but if it's GPL'd AND patented, well then that's something
> entirely different altogether.
> http://kerneltrap.org/node.php?id=388
Where does MySQL fit in here then? They charge for MySQL for
commercial stuff, but not for open stuff?
> I can see where the GPL can foster software patents. Normally, when
> software companies take the BSD code, and incorporate it into their own
> products with changes, they don't have to make the changes available to
> everyone, as the GPL stipulates. So, there's no need for a patent. But
> since the GPL states that any modifications to the GPL'd code be made
> public, well, I can see the need for additional protection by patenting
> the code. Too bad the GPL doesn't state anything about patents. Here,
> the complexity of the GPL works against itself, and the simplicity of
> the BSDL is the "winner".
I don't see what's wrong with a commercial company (or contributors to
the Linux project or whoever) putting a patent out on their GPL'd
work, as long as it can continue in GPL'd stuff, but not closed
commercial s/ware.
Has this been done?
Paul...
--
plinehan__AT__yahoo__DOT__com
John S. Dyson
"Paulie" <plin...@yahoo.com> wrote in message news:3d6ff1bb...@127.0.0.1...
>
>
> Donn Miller <dmmi...@cvzoom.net> wrote:
>
>
> > Uh huh uh huh. And, as Linux's popularity starts to skyrocket among
> > various commercial vendors, software patents are going to become common.
> > RedHat has already patented some of Ingo Molinar's work. AFAIK, he's
> > doing some very good work. However, it can't be considered free if
> > patents are going to be applied towards parts of the Linux kernel.
>
>
> Why not? If you are allowed to use it for GPL'ed s/ware for nothing,
> but have to pay for commercial use, then what's the problem?
>
and not in use. This is very similar to the normal commercial encumberances,
where you can usually use software, but redistributing it can carry legal
consequences.
John
Richard Tobin
John S. Dyson <dy...@iquest.net> wrote:
>Proclaiming that it is a superior free by being restrictive is indeed CLEARLY
>doublespeak.
Not at all - freedoms often conflict. Your freedom to remain alive is
enhanced by removing my freedom to kill you. You may disagree that
the GPL issue is analogous, but that's all it is - a disagreement.
JD
"Richard Tobin" <ric...@cogsci.ed.ac.uk> wrote in message news:akp63g$25bg$1...@pc-news.cogsci.ed.ac.uk...
> In article <DQOb9.1215$I4.2...@news.iquest.net>,
> John S. Dyson <dy...@iquest.net> wrote:
>
> >Proclaiming that it is a superior free by being restrictive is indeed CLEARLY
> >doublespeak.
>
> Not at all - freedoms often conflict. Your freedom to remain alive is
> enhanced by removing my freedom to kill you. You may disagree that
> the GPL issue is analogous, but that's all it is - a disagreement.
>
actions, or giving away work product of other (but potentially related) work
extending beyond the original code that has the license,
then that redistribution is encumbered, and the software cannot
reasonably be considered to be 'free.' Costs can be incurred by license
limitations that are
relieved by purchase, or relieved by accepting encumberance of
other work-product that wasn't contained in the original works.
In essense, to redistribute GPLed works, especially if you modify
them, you have certain requirements that apply to works beyond
the original code. The GPL kind of license cost is pretty much
akin to the social aspect of finding ways of taxing everyone
but yourself -- it is advocated to a certain group as 'free', yet
it only appears free to those who don't have to comply with the
costs. The fact is that something that is claimed to be 'free'
or 'free software', stretching the truth when challenged (even
to the extent of giving software a human characteristic to
stretch the truth in some of the more pathetic cases) isn't
a matter of a 'different' perspective, but perhaps a matter of
self delusion for the most honest (but true believer) advocates.)
Justifying these limitations by claiming that such redistribution
encumberances are good, because they are socially better (which
one might or might not agree with), doesn't justify the misleading
claim that the software is 'free.' The word 'free' is not the same
word as the word 'good'. The word 'good' is much easier to justify
a personal intepretation, but 'free' is mostly just being misused
by the GPL advocacy.
Remember: free is NOT a synonym for 'socially responsible'...
Even the term 'socially responsible' is subject to interpretation,
but misusing the term 'free' knowing that the word is CLEARLY
being selectively used, where certain freedoms of certain
classes of users/redistributors are being prejudicially discriminated
against. In this case it is probably legal and moral to discriminate
in this fashion, but it is being VERY DISHONEST to claim the
use of a term that suggests that limitations aren't important.
Perhaps those limitations aren't important to YOU, but the
thing about freedom is that you let those who don't cause you
material harm (not harm you by ideas, but actual damage) do
what they want to do. The thought police, that deem certain
freedoms as not being important, and redefine usage based upon
those ideas (e.g. 'free') are imposing pressure that isn't consistant
with the notion of 'free' use, reuse and redistribution.
The term 'free' doesn't imply limitations or restrictions on your
choice to redistribute something in whole, or in part. It also
doesn't imply rules or restrictions on work that you might add
on to that 'free' work. It is certainly legally, morally okay to add
those restrictions to add-on work, but the term 'free' doesn't really
apply.
So again, you can add restrictions, create licenses that have
redistribution rules. It MIGHT be socially good, but the resulting
encumberances dont' make the software 'free.'
One perfect example of the GPL mindset is the notion that some
advocates are so fearful of their 'family jewels' being used by
Microsoft. Alas, that is exactly the mentality that begats GPL
advocacy, and it might even be socially good to try to discriminate
against companies like Microsoft. However, such software with
licenses that are tuned against 'exploitation' are certainly not free,
and also, the selective notion of 'right and wrong' has collateral
effects in different developer communities. The term 'free' is
NOT a synonym with 'good'.
John
Paulie
"John S. Dyson" <dy...@iquest.net> wrote:
> > Why not? If you are allowed to use it for GPL'ed s/ware for nothing,
> > but have to pay for commercial use, then what's the problem?
> Where GPL encumberances start rearing their ugly heads are in redistribution
> and not in use.
I believe that I understand this.
> This is very similar to the normal commercial encumberances,
> where you can usually use software, but redistributing it can carry legal
> consequences.
I'm no expert on this, but here's my take (my excuses if I've missed
something important - please let me know if I have).
I'm programmer X (brilliant, dashing, handsome...), I devise algorithm
or code snippet Y (brilliant, dashing, handsome).
I patent my work, but I also allow it to be released under the GPL,
with the stipulation that people who use my algorithm in a commercial
situation (i.e. to make money out of it) must negotiate with me for
the right to use my algorithm and/or snippet in their product.
If programmer Z (brilliant, comely, beautiful) wishes to use my
algorithm in her home spun RDBMS which she uses to keep details of her
cookery recipes or whatever, she's perfectly entitled to do that.
If however, she wants to sell a product to the Wicked Witch BillZilla,
which contains software using my algorithm or code, then she has to
negoiate a licence with me.
What is the problem with that scenario?
Paul...
> John
--
plinehan__AT__yahoo__DOT__com
John S. Dyson
"Paulie" <plin...@yahoo.com> wrote in message news:3d71262b...@127.0.0.1...
>
> If however, she wants to sell a product to the Wicked Witch BillZilla,
> which contains software using my algorithm or code, then she has to
> negoiate a licence with me.
>
>
> What is the problem with that scenario?
>
will morph the discussion from 'GPL isn't free' (which it isn't free), to a claim
that 'GPL isn't good' (which may or may not be good, depending upon
viewpoint.)
Just because someone clarifies the GPL as not being a license of free software,
doesn't mean that claim is the same as 'the GPL isn't good'.
John
Donn Miller
> Nothing, except the software isn't 'free'. All too often, the GPL proponents
> will morph the discussion from 'GPL isn't free' (which it isn't free), to a claim
> that 'GPL isn't good' (which may or may not be good, depending upon
> viewpoint.)
Another big misconception is that people seem to think that the GPL is
going to magically turn a bad code into diamonds simply because so many
coders and banging on it and throwing code at it. I think the GPL and
the bazaar development style leads to bloat.
John S. Dyson
"Donn Miller" <dmmi...@cvzoom.net> wrote in message news:3d717...@corp.newsgroups.com...
> John S. Dyson wrote:
>
> > Nothing, except the software isn't 'free'. All too often, the GPL proponents
> > will morph the discussion from 'GPL isn't free' (which it isn't free), to a claim
> > that 'GPL isn't good' (which may or may not be good, depending upon
> > viewpoint.)
>
> Another big misconception is that people seem to think that the GPL is
> going to magically turn a bad code into diamonds simply because so many
> coders and banging on it and throwing code at it. I think the GPL and
> the bazaar development style leads to bloat.
>
the onion layers of rhetoric, the GPL is essentially a very crafty license for
developers who want to release their software/source code, yet maintain
a competitive advantage (or parity) with others who might decide to invest
either minor amount of time, or major improvements in the
works. This is certainly not the character of a license of 'free' software.
When I give my work away, or my work product away, my minimal criteria
is that it shouldn't be used for explicitly illegal or explicitly unethical
activities. Very useful tools, like 'hammers' can be used for murder,
but also can be used for home construction. The kind of thing that would
cause me to place conditions on reuse of my explicitly free work
would be to avoid obvious misuse for explicit damage to other people.
For example, if I produced a work of the nature of a large destructive
weapon, I certainly would NOT make it free -- and would restrict access
for ethical and moral reasons.
I certainly wouldn't try to prejudicially restrict my *free* software work product
against those that have different political opinion, or have a business interest
that makes me feel 'bad.' However, if I thought that I wanted to restrict my work
against a certain usage, in a discriminatory fashion, like the GPL is meant to do for
specific kinds of situations, then I wouldn't misleadingly call the software 'free.'
Alot of my work on FreeBSD was various levels of modification of existing
files, however, when the issue of the advertising clause became significant,
I was one of the first to relax my licensing terms (when I was the full author
of the code.) Even considering that
relatively high-profile issue with the BSD licensing terms, with careful (legal
and ethical) ethical behavior, there was no reason that the advertising
clause would affect or encumber other people's work product. The cost
of some of the inconvienient old BSD license terms was actually minor
for a developer who earns their keep by writing code, or gaining some
'advertisement' from code authorship. When complex and costly code
is encumbered by a license, CERTAINLY doesn't make that license
immoral or unethical, because following and respecting those license
terms were a choice of that developer... However, calling a license
with such restrictions: "FREE" is deceitful.
Without the 'rhetoric', the GPL is a crafty license that can be used to
suppress competition and maintain control by a small developer. As a license
of 'free' software, it is somewhat contradictory, at least, until absurd
anthropomorphic attributes are applied to 'software' -- which is a further
stretch than applying such attributes to a person's pet dog :-). (e.g.
information or software don't really want to be free, do they?) :-). Even
then, the GPL imposes restrictions even beyond the software's lack of
free will to allow it to hop from machine to machine :-).
John
Donn Miller
> Without the 'rhetoric', the GPL is a crafty license that can be used to
> suppress competition and maintain control by a small developer. As a license
> of 'free' software, it is somewhat contradictory, at least, until absurd
> anthropomorphic attributes are applied to 'software' -- which is a further
> stretch than applying such attributes to a person's pet dog :-). (e.g.
> information or software don't really want to be free, do they?) :-). Even
> then, the GPL imposes restrictions even beyond the software's lack of
> free will to allow it to hop from machine to machine :-).
I was thinking that the Linux kernel would benefit if it were placed
under a much less restrictive license. For example, I think the
LGPL would be a good choice. But I think that the downside is that
the GPL prevents forks from ocurring in the code, which sort of makes
that particular GPL'd codebase function as the "one true codebase". Since
there's one codebase, it seems to me that this would cause a lot of
bloating. For example, you've got the "one true" fileutils (find, ls,
awk, etc.), the "one true" kernel (Linux). There's really no room for
competition. Interestingly enough, I was thinking about writing some
alternative utils to the standard GNU fileutils if I ever get back to
Gentoo, and putting those under the LGPL, BSDL, or MITL. I
actually think I'd prefer to create my own license, though.
But I do think the licensing on Linux should be less restrictive. I
was wondering what effect it would have had it been LGPL'd instead of
GPL'd? Some commercial vendors are being strangled by the GPL's
requirements. For example, Lindows wanted the flexibility of not
having to release the source code of the GPL-derived components, at
least not immediately. The Free Software Foundation was quick to jump
on their backs, demanding to know where the source code was.
In some ways, the GPL is like the open-source equivalent of MS,
although diametrically opposite by pi/2 radians. Also, I think it's
interesting that everyone in the free software industry hates MS so
much, yet RedHat is moving towards a yearly subscription model, not
unlike that which MS is offering. I think Lindows has such a thing as
well.
I was thinking: this is the open source community. Why can't we come
together and draft together what we consider as a whole to be the
ideal open source license? It could have many different versions,
depending on the freeness or restrictedness that is desired. As it is
now, the GPL advocates are forcing themselves to swallow RMS' ideas as
free. Why let one person control what we consider to be free? The
problem, of course, is that there are mountains of existing software
out there that is already GPL'd, such as gcc, the GNU toolchain, and
even Linux. So this isn't feasible, and we'll forever be doomed to
conform to one person's idea of "free".
David Schultz
> When standing back away from FreeBSD, and the whole
> 'open' software 'movement', I had realized that too much of the GPL advocacy
> was being purely dishonest, either by ignorance or by intent. The
> whole thing about calling GPLed works 'free software' is another bit
> of disinformation. Perhaps in poetic use, it is 'freer' than some kinds of
> commercial software, but redistribution encumberances associated with
> the so called 'free' software is a very clear case of double speak at best,
> and at worst, absolutely dishonest.
The advocacy for the GPL and for Linux can be pretty nasty at times.
For example, observe the degenerate cesspool that is comp.os.linux.advocacy.
On the other hand, I have seen a bit of mudslinging on the side of the
various BSDs as well. It seems the most vocal advocates of anything
tend to be more fanatical than their level of competence should allow.
Many people choose the GPL for valid reasons rather than for
popularity, and I respect that. Some of them are idealists who fully
comprehend the intent of the GPL. Others, such as Hans Reiser, blow
off the ``free'' nonsense and take the GPL at face value, using its
restrictions to further their own ideologies. GPL'd software can be
annoying as a practical matter, but I'm willing to tolerate that. But
I can't stand the people who don't understand and wave their batons
nonetheless, *especially* when they criticize other models (including
commercial software).
On the FreeBSD advocacy front, I am impressed by some of the folks on
the Documentation Project. Most of them have enough experience to
know what they're talking about, and they advocate the operating
system the right way, namely, by emphasizing positive aspects of
FreeBSD instead of slandering The Other Side. Of course, it will take
more advocacy and a bit of commercial support to get to, say, the same
place as Linux in terms of popularity.
David Schultz
> The biggest justification I've heard recently is that Microsoft can suck
> in all the BSD code they want without contributing anything back to
> *BSD. The GPL is supposed to prevent this, they claim.
Microsoft *has* used BSDL'd code and their operating system still
doesn't work particularly well. But more to the point, at least the
poor slobs using Windows can have slightly better networking
capabilities due to the code.
For what it's worth, Linux also uses BSDL'd code. The difference is
that Microsoft actually acknowledges the original authors in the
release notes, whereas Linux developers merely remove the copyright
notices. For example, look at the Linux Packet Filter. (It's more
obvious in older versions where the code came from.) There was also
an incident a little while ago in which a FreeBSD SCSI driver was
plagiarized in its entirety and the original author found out.
Erik Nygren
...
> or code snippet Y (brilliant, dashing, handsome).
>
> I patent my work, but I also allow it to be released under the GPL,
> with the stipulation that people who use my algorithm in a commercial
> situation (i.e. to make money out of it) must negotiate with me for
> the right to use my algorithm and/or snippet in their product.
>
> If programmer Z (brilliant, comely, beautiful) wishes to use my
> algorithm in her home spun RDBMS which she uses to keep details of her
> cookery recipes or whatever, she's perfectly entitled to do that.
>
> If however, she wants to sell a product to the Wicked Witch BillZilla,
> which contains software using my algorithm or code, then she has to
> negoiate a licence with me.
>
>
> What is the problem with that scenario?
But what if programmer W adds to your work, making it more useful, and
programmer V takes the work of W one step further and so on?
If Company C wants to create a really clever product that does most of
the things the now quite mature software does, and a little more, they
will have to license work from all of the previous programmers.
I think Company C will have to do a clean-room rewrite, and none of
the programmers will then be able to buy food for their kids.
Company A and B will be inclined to buy the software from Company C, even
though the "free" version might suit them better, because they have every
intention to make money out of it, and it is impossible to find all the
participating programmers and get licenses that permit them to use the
software.
So, we'll end up with a bunch of programmers with very hungry kids, a few
employed programmers that have a very boring job of rewriting working
software, a few companys that are forced to use inferior software and
one company that makes money on software that is worse than the "free"
alternative.
Was that really what you wanted?
--
Erik Nygren
e r i k { a t } s w i p { d o t } n e t
Linux - If you hate Microsoft, FreeBSD - If you love Unix
david parsons
Donn Miller <hac...@yahoo.com> wrote:
>I was thinking that the Linux kernel would benefit if it were placed
>under a much less restrictive license. For example, I think the
>LGPL would be a good choice. But I think that the downside is that
>the GPL prevents forks from ocurring in the code,
Nope. If the interfaces were static, it would be pretty easy to
fork the code regardless of the license. The extreme mutability of
the interfaces makes it, umm, difficult to do a codefork unless you
want to spend much of your life recoding device drivers to work on
your offshoot kernel.
>competition. Interestingly enough, I was thinking about writing some
>alternative utils to the standard GNU fileutils if I ever get back to
>Gentoo,
Why write them? Why not just take the existing body of BSD tools
and port them over to Linux? You will run into some feature
collisions (gnu autoconf, just to name one extremely annoying
example, is lousy with stupid special-case hooks that activate when
it detects that it's running on a Linux box) with these tools
when you encounter unportable code, but the BSD tools just work
[0] as they have for the past 20 or so years [1].
[0: For me, at least.]
[1: At least when people don't write their code assuming that
everything is a fucking GNU [2].]
[2: Or that everybody has perl, python, ruby, or any of a dozen
other stupid vanity languages.]
____
david parsons \bi/ ``df -l'' ? WTF is THAT feature?
\/
Donn Miller
> For what it's worth, Linux also uses BSDL'd code. The difference is
> that Microsoft actually acknowledges the original authors in the
> release notes, whereas Linux developers merely remove the copyright
> notices. For example, look at the Linux Packet Filter. (It's more
> obvious in older versions where the code came from.) There was also
> an incident a little while ago in which a FreeBSD SCSI driver was
> plagiarized in its entirety and the original author found out.
Actually, I've seen some "Regents..." (BSD copyright) strings in the
Linux kernel tree. Of course, maybe the Linux developer did what he did
out of ignorance. Now, if it happens again, sic the Regents on his arse.
Lee Harr
> employed programmers that have a very boring job of rewriting working
> software, a few companys that are forced to use inferior software and
> one company that makes money on software that is worse than the "free"
> alternative.
>
Nice.
This pretty much sums it up.
How come so few people understand this?
Joshua Lee
> Well, Linux might be better at threading. Xine worked very well for me
> on Linux, but on FreeBSD, doesn't work very well at all. It may be a
I use Ogle, it seems to work.
Donn Miller
OK. But why does Xine work on Linux and not FreeBSD? Could it be a
problem with threading? Well, Xine does work occasionally on FreeBSD,
but on Linux it works 100%.
Kristian Rask
On Fri, 30 Aug 2002 12:26:27 -0400, Donn Miller <dmmi...@cvzoom.net>
wrote:
>The biggest justification I've heard recently is that Microsoft can suck
>in all the BSD code they want
*
>without contributing anything back to*BSD.
*
>The GPL is supposed to prevent this, they claim. But I countered
Where is the downside of that argument ? ;-)
Regards
Kristian aka "The eternal newbie"
Donn Miller
> nearly every average user would benefit from MS using _even more_ code
> from BSD, because BSD's code is known to work!
> if MS would base their browser, web-server, ip-stack, etc. on code from
> *BSD or Linux it would surely make a better OS.
No they wouldn't. NT's kernel is known to be based on VMS, and VMS is
super super stable, and compare VMS stability to NT. If MS did suck in
the entirety of the BSD code, they'd just fuck things up by trying to
make it work with their registry. And we all know where that would lead to.
Joshua Lee
> Joshua Lee wrote:
>> I use Ogle, it seems to work.
>
> OK. But why does Xine work on Linux and not FreeBSD? Could it be a
> problem with threading? Well, Xine does work occasionally on FreeBSD,
> but on Linux it works 100%.
Funny, I had problems running Xine on Linux. That's why I switched to Ogle
on that platform and never touched Xine on FreeBSD. (Plus when PCI modem
support and Java was mostly implemented for FreeBSD, I switched my new box
to FreeBSD as well.)
Kristian Rask
On Mon, 02 Sep 2002 14:54:29 -0400, Donn Miller <dmmi...@cvzoom.net>
wrote:
> If MS did suck in
> the entirety of the BSD code, they'd just fuck things up by trying to
> make it work with their registry. And we all know where that would lead to.
There is nothing wrong w. the concept of a registry in terms of
stability..
If FreeBSD allowed users to run user programs that installed them
selves and modified a shit load of stuff in /etc, then you would
quickly discover that < Insert Your Favourite UnixLike Initialization>
is at least as sucky :-)
regards
Kristian
Richard Tobin
Donn Miller <dmmi...@cvzoom.net> wrote:
>OK. But why does Xine work on Linux and not FreeBSD? Could it be a
>problem with threading? Well, Xine does work occasionally on FreeBSD,
>but on Linux it works 100%.
If you're getting "waitq_remove: Not in queue" errors, try changing
if(w->tips_thread)
pthread_cancel(w->tips_thread);
to
if(w->tips_thread) {
pthread_cancel(w->tips_thread);
w->tips_thread = 0;
}
in src/xitk/xine-toolkit/tips.c
A.P.Manners
> I patent my work, but I also allow it to be released under the GPL,
> with the stipulation that people who use my algorithm in a commercial
> situation (i.e. to make money out of it) must negotiate with me for
> the right to use my algorithm and/or snippet in their product.
This is not the GPL license it is your license.
> If however, she wants to sell a product to the Wicked Witch BillZilla,
> which contains software using my algorithm or code, then she has to
> negoiate a licence with me.
Nope. That is how software patents are supposed to work but it is not
what happens in the real world. In the real world Wicked Witch BillZilla
asks his patent officer how many of their patents for trivial algorithms
your code is infringing and tells you to forget your patent. A small
inventor stands no chance of using a patent against a company with a
large portfolio. On the other hand, I suspect large companies find the
whole process very expensive and wish it would get fixed but they have
no option but to build up a large portfolio of patents to defend
themselves against their inevitable infringments. The real losers, as
ever, are the small inventive types who find their patents are worth
very little when it comes to exploiting their ideas (so better to keep
it quiet).
Bart Lateur
>GPL supporters just disagree
>with you about what constitutes freedom. What it seems to boil down
>to is that they think it's an important aspect of freedom that their
>code can't be made proprietary, and you think it's important that it
>*can* be made proprietary.
No... the original GPL'ed code still would be free. It is the wish that
the extra work which enhances the original free stuff significantly,
isn't necessarily free. It looks like a reasonable demand.
The GPL wants to rob you of your rights as a programmer.
--
Bart.
John S. Dyson
"Preston Crawford" <pres...@serpentor.cobrala> wrote in message news:slrnana426....@serpentor.cobrala...
> In article <je4anusap9k097rk7...@4ax.com>, Bart Lateur wrote:
> > No... the original GPL'ed code still would be free. It is the wish that
> > the extra work which enhances the original free stuff significantly,
> > isn't necessarily free. It looks like a reasonable demand.
> >
> > The GPL wants to rob you of your rights as a programmer.
>
>
Of course not!!! Of course, it is still a lie to claim that GPLed code is
'free software.' The lie about GPL isn't the license (everyone is a grown-up
and can choose to avoid grossly encumbred code), the lie is related to
the rhetoric about the GPL being free. If someone 'in the know' makes
a robust claim that GPLed works are 'free software', their honesty and integrity
should be questioned. If 'joe programmer', who doesn't really care about
software licensing, makes a causal comment about GPLed works being
'free', that isn't really a 'lie', and is akin to common misuse of other
terminology.
John
Patrick TJ McPhee
Kristian Rask <krask123SpamM...@mediac.dk> wrote:
% There is nothing wrong w. the concept of a registry in terms of
% stability..
It seems like a significant single point of failure to me. It's one
big file that gets changed all the time and you can't boot if it's
corrupted.
--
Patrick TJ McPhee
East York Canada
pt...@interlog.com
david parsons
Patrick TJ McPhee <pt...@interlog.com> wrote:
>In article <3d746aac...@news.tele.dk>,
>Kristian Rask <krask123SpamM...@mediac.dk> wrote:
>
>% There is nothing wrong w. the concept of a registry in terms of
>% stability..
>
>It seems like a significant single point of failure to me. It's one
>big file that gets changed all the time and you can't boot if it's
>corrupted.
Reminds you of /etc, doesn't it?
____
david parsons \bi/ The only problem with /etc is that everyone and
\/ their sister uses a different file format there.
Bart Lateur
>>% There is nothing wrong w. the concept of a registry in terms of
>>% stability..
>>
>>It seems like a significant single point of failure to me. It's one
>>big file that gets changed all the time and you can't boot if it's
>>corrupted.
>
> Reminds you of /etc, doesn't it?
Except that /etc is a directory... The registry is one file. The impact
of file corruption is much larger in the latter case.
--
Bart.
Kristian Rask
On Wed, 04 Sep 2002 12:06:11 GMT, Bart Lateur <bart....@pandora.be>
wrote:
try cat /etc
etc is a file ... screw it up and see what happens
that the inidividual sections are files and not records does not
guaratee the abscense of a single point of failure
thimk !
jpd
> In article <wVed9.42$qh1....@news.ca.inter.net>,
> Patrick TJ McPhee <pt...@interlog.com> wrote:
>>In article <3d746aac...@news.tele.dk>,
>>Kristian Rask <krask123SpamM...@mediac.dk> wrote:
>>% There is nothing wrong w. the concept of a registry in terms of
>>% stability..
>>
>>It seems like a significant single point of failure to me. It's one
>>big file that gets changed all the time and you can't boot if it's
>>corrupted.
>
> Reminds you of /etc, doesn't it?
I can still boot in SU without /etc, IIRC.
> ____
> david parsons \bi/ The only problem with /etc is that everyone and
> \/ their sister uses a different file format there.
Luckily, most are (somewhat) human readable.
--
j p d (at) d s b (dot) t u d e l f t (dot) n l .
Want to search the registry? Forget about regedit, Dump it and grep!
Richard Tobin
Kristian Rask <krask123SpamM...@mediac.dk> wrote:
>try cat /etc
> etc is a file ... screw it up and see what happens
How would one "screw it up"? It's much harder for a buggy program
to corrupt a directory than a file.
Chuck Swiger
> In article <wVed9.42$qh1....@news.ca.inter.net>,
> Patrick TJ McPhee <pt...@interlog.com> wrote:
>>% There is nothing wrong w. the concept of a registry in terms of
>>% stability..
>>
>>It seems like a significant single point of failure to me. It's one
>>big file that gets changed all the time and you can't boot if it's
>>corrupted.
>
> Reminds you of /etc, doesn't it?
Not really, no. I've had ~20 files in /etc change in 2002 out of:
769-pi# du -a /etc | wc -l
377
...total files.
Almost all of the files in /etc are human edittable, rather than being a
binary representation. I can boot single-user mode, or off a fixit CD or
floppy and have a very good chance of recovering from problems, even without
having a backup. Even with a "Windows Repair Disk", sufficient problems with
the Windows registry will require a reinstall.
Under FreeBSD, you can install a brand-new system, and simply copy
/etc/rc.conf from an old machine, which handles many of the changes and
customizations in one simple step.
-Chuck
Chuck Swiger | ch...@codefab.com | All your packets are belong to us.
-------------+-------------------+-----------------------------------
"The human race's favorite method for being in control of the facts
is to ignore them." -Celia Green
Donn Miller
> just compare Windows 9x stability with NT's. and IIRC the NT TCP/IP Stack
> is heavily based on BSD code and, well it kinda works. but if you look at
> the Windows 9x IP-Stack.....
That's like comparing an elephant turd to a dog turd. They both stink,
but one stinks up the joint a hell of a lot more. So far, I've seen two
XP installations, and none of them were stable at all. Of course, I
suspect HW problems was to blame in the second XP install I've seen.
I'd rather run the Real Thing(tm). Besides, no one can prove beyond a
doubt that MS is using any /FreeBSD/ code, other than the mention of of
a couple of FreeBSD developers' name in the XP release notes.
Donn Miller
> In article <wVed9.42$qh1....@news.ca.inter.net>,
> Patrick TJ McPhee <pt...@interlog.com> wrote:
>
>>In article <3d746aac...@news.tele.dk>,
>>Kristian Rask <krask123SpamM...@mediac.dk> wrote:
>>
>>% There is nothing wrong w. the concept of a registry in terms of
>>% stability..
>>
>>It seems like a significant single point of failure to me. It's one
>>big file that gets changed all the time and you can't boot if it's
>>corrupted.
>
>
> Reminds you of /etc, doesn't it?
But /etc doesn't have one massive file which is a single point of failure.
Donn Miller
> try cat /etc
> etc is a file ... screw it up and see what happens
But /etc is a directory entry. And so is C:\WINDOWS.
Donn Miller
> In article <3d7601ed...@news.tele.dk>,
> Kristian Rask <krask123SpamM...@mediac.dk> wrote:
>
>
>>try cat /etc
>>etc is a file ... screw it up and see what happens
>
>
> How would one "screw it up"? It's much harder for a buggy program
> to corrupt a directory than a file.
Yeah. Another thing to remember is that applications write info to the
registry when they're installed. And none of the files are written to
at all in /etc, except when the superuser edits them. The registry is
written to for registration of packages, and there's a greater chance
that an installer someplace can fuck something up. When if the system
crashes or locks up when the application installshield (or whatever it's
called) is writing info to the registry? The registry can be corrupted.
At least when packages are registering their info when FreeBSD
crashes, worst-case is that some files in /var/db/pkg get hosed, as
opposed to important configuration files residing in /etc getting all
fucked up, screwing up the system config. Even then, it would be
trivial to boot into single-user, and fix the problem, and
configurations are distributed amoung multiple files in /etc, which
eliminates the single point of failure.
I'm sick of dealing with XP-related driver problems and crashes. If XP
is the best thing MS has ever put out, they're in big trouble IMO. No
one I know likes Windows XP or ME.
jpd
[snip]
> I'm sick of dealing with XP-related driver problems and crashes. If XP
> is the best thing MS has ever put out, they're in big trouble IMO. No
> one I know likes Windows XP or ME.
They're not. Simply because there's still this large mass[0] that has the
idea they can't live without micros~1 firmly set in their concrete brains.
As long as there's a large mass willing to buy their overpriced crap, they
won't feel the slightest bit of pain. An itch, maybe, but no pain at all.
[0] Of topical idiots, if you ask me. But now lots of those will exclaim
that I'm `just bashing' their favourite bully, so to avoid hurting
their feelings, don't ask.
Bill Vermillion
Donn Miller <dmmi...@cvzoom.net> wrote:
>Richard Tobin wrote:
>> In article <3d7601ed...@news.tele.dk>,
>> Kristian Rask <krask123SpamM...@mediac.dk> wrote:
>>
>>>try cat /etc
>>>etc is a file ... screw it up and see what happens
>>
>>
>> How would one "screw it up"? It's much harder for a buggy program
>> to corrupt a directory than a file.
>Yeah. Another thing to remember is that applications write info to the
>registry when they're installed. And none of the files are written to
>at all in /etc, except when the superuser edits them.
And program specific libraries are [typically] in /usr/local/lib
so you don't go through the MS 'DLL-HELL' with one library
overwriting another and breaking the first app.
There is a lot to be said for several small pieces parts instead of
putting everything in one place. The old saying was "Don't put
all your eggs in one basket". I think at times the registry is
like "Putting all your eggs in one risk kit!".
Bill
--
Bill Vermillion - bv @ wjv . com
John S. Dyson
"Bill Vermillion" <b...@wjv.comREMOVE> wrote in message news:H1xKo...@wjv.com...
> In article <3D765767...@cvzoom.net>,
> Donn Miller <dmmi...@cvzoom.net> wrote:
> >Richard Tobin wrote:
> >> In article <3d7601ed...@news.tele.dk>,
> >> Kristian Rask <krask123SpamM...@mediac.dk> wrote:
> >>
>
> >>>try cat /etc
> >>>etc is a file ... screw it up and see what happens
> >>
> >>
> >> How would one "screw it up"? It's much harder for a buggy program
> >> to corrupt a directory than a file.
>
> >Yeah. Another thing to remember is that applications write info to the
> >registry when they're installed. And none of the files are written to
> >at all in /etc, except when the superuser edits them.
>
> And program specific libraries are [typically] in /usr/local/lib
> so you don't go through the MS 'DLL-HELL' with one library
> overwriting another and breaking the first app.
>
the circumstances) look in your local directory for the 'specially versioned' DLLs.
I use that 'feature' on 2000 when I need to create DLLs, but it has been a major
problem with upgrades on MS software. Recently (and late), I installed my
Tax software, and it felt free to install new and important system DLLs. That
is evil beyond imagination.
John
Bill Vermillion
John S. Dyson <dy...@iquest.net> wrote:
>
>"Bill Vermillion" <b...@wjv.comREMOVE> wrote in message news:H1xKo...@wjv.com...
>> In article <3D765767...@cvzoom.net>,
>> Donn Miller <dmmi...@cvzoom.net> wrote:
>> >Richard Tobin wrote:
>> >> In article <3d7601ed...@news.tele.dk>,
>> >> Kristian Rask <krask123SpamM...@mediac.dk> wrote:
>> >>
>>
>> >>>try cat /etc
>> >>>etc is a file ... screw it up and see what happens
>> >>
>> >>
>> >> How would one "screw it up"? It's much harder for a buggy program
>> >> to corrupt a directory than a file.
>>
>> >Yeah. Another thing to remember is that applications write info to the
>> >registry when they're installed. And none of the files are written to
>> >at all in /etc, except when the superuser edits them.
>>
>> And program specific libraries are [typically] in /usr/local/lib
>> so you don't go through the MS 'DLL-HELL' with one library
>> overwriting another and breaking the first app.
>One comment, and slightly off topic: MS will sometimes (I don't
>know all of the circumstances) look in your local directory for
>the 'specially versioned' DLLs.
So I guess the comment here would be "but not often enough" :-)
>I use that 'feature' on 2000 when I need to create DLLs, but it
>has been a major problem with upgrades on MS software.
YOU CREATE DLLs. Oh - I'm crushed :-( I avoid MS stuff as much
as possible.
>Recently (and late), I installed my Tax software, and it felt
>free to install new and important system DLLs. That is evil
>beyond imagination.
I'll go along with that idea.
Patrick TJ McPhee
david parsons <o...@pell.portland.or.us> wrote:
% In article <wVed9.42$qh1....@news.ca.inter.net>,
% Patrick TJ McPhee <pt...@interlog.com> wrote:
% >In article <3d746aac...@news.tele.dk>,% >% There is nothing wrong w. the concept of a registry in terms of
% >% stability..
% >
% >It seems like a significant single point of failure to me. It's one
% >big file that gets changed all the time and you can't boot if it's
% >corrupted.
%
% Reminds you of /etc, doesn't it?
No. The operative word was `gets changed all the time'. The registry gets
updated frequently. Certainly every time you install an application, but
for most applications, every time you run, and for some continuously as
you run.
Obviously, you're always going to have critical files and critical bits
of disk, but you improve stability by making them as small as possible
and reducing the chance of them becoming corrupted. To my mind, and you
know that I'm right because my spelling is quite good and I've adopted a
reasonable tone, the registry on windows goes in the opposite direction.
John S. Dyson
"Bill Vermillion" <b...@wjv.comREMOVE> wrote in message news:H1xzu...@wjv.com...
>
> >One comment, and slightly off topic: MS will sometimes (I don't
> >know all of the circumstances) look in your local directory for
> >the 'specially versioned' DLLs.
>
> So I guess the comment here would be "but not often enough" :-)
>
> >I use that 'feature' on 2000 when I need to create DLLs, but it
> >has been a major problem with upgrades on MS software.
>
> YOU CREATE DLLs. Oh - I'm crushed :-( I avoid MS stuff as much
> as possible.
>
I prefer and love (in an appropriate way) :-) FreeBSD. I tolerate the lesser
OSes like Linux or Windows, and will non-prejudicially use them when appropriate.
Frankly, I am on a mild re-education spree, where I am working to become
comfortable and near-expert on Microsoft stuff. There is alot to learn in there,
but it is irritating because of the junky quality and extremely fat APIs. For
different reasons, and with different observations, it really isn't much different
when using Linux kernel based OSes and Microsoft stuff -- I mean, alot of
things are just done incorrectly and in silly ways. Perhaps Linux is a little
better than Win2000, and Linux is certainly more Unix-like, but there are
serious eccentricities in both. As a pure end user, the eccentricities become
less important, but as an OS and kernel savvy individual, I can see some
real junky design -- but becoming obscessed with the problems with the OS
can inhibit my immediate, non-OS-developer goals.
> >Recently (and late), I installed my Tax software, and it felt
> >free to install new and important system DLLs. That is evil
> >beyond imagination.
>
> I'll go along with that idea.
>
It is INCREDIBLE that user programs can/will re-install shared objects (DLLs),
and developers in the Unix community would/should be fired for doing such
evil. There is alot of crap in the Microsoft stuff, but it is best to know
about the stuff also, because it is good to learn from other people's mistakes.
Note that the Linux crowd hadn't done enough of that, and are more oriented
towards fear and hate against their 'great satan' than willing to learn about it.
They would have done much better, and made more technical progress more
quickly if they hadn't been so adverse about learning VM techniques (for example.) :-).
John
Bart Lateur
>Even with a "Windows Repair Disk", sufficient problems with
>the Windows registry will require a reinstall.
Hmm... I've been thinking. Are there any privilege levels set for the
Registry? Or can any user modify anything? Because *that* would be
extremely dangerous for viruses: change one file, and boom!
The Win32API docs I have, don't even mention this. But it is pre-W2K, so
things may have changed.
--
Bart.
Bart Lateur
>Besides, no one can prove beyond a
>doubt that MS is using any /FreeBSD/ code, other than the mention of of
>a couple of FreeBSD developers' name in the XP release notes.
Not any more... but with slightly older Windows versions, one could find
traces of the copyright notices in the executables.
--
Bart.
David Schultz
> Perhaps Linux is a little
> better than Win2000, and Linux is certainly more Unix-like, but there are
> serious eccentricities in both. As a pure end user, the eccentricities become
> less important, but as an OS and kernel savvy individual, I can see some
> real junky design -- but becoming obscessed with the problems with the OS
> can inhibit my immediate, non-OS-developer goals.
The scary thing is that many developers are willing to swear that the
eccentricities are in traditional Unix systems.
Kristian Rask
On Wed, 04 Sep 2002 14:42:31 -0400, Donn Miller <dmmi...@cvzoom.net>
wrote:
> Besides, no one can prove beyond a
>doubt that MS is using any /FreeBSD/ code, other than the mention of of
>a couple of FreeBSD developers' name in the XP release notes.
I thought that Win2000 went from MS'ish TCP/Stack signature to
"indistinguishable from FreeBSD 3.something" (as seen by Netcraft) a
few months before release.. and apparently not long after MS stopped
their ? attempt of converting hotmail to run on 2000
regards
Kristian
Bill Vermillion
>> >I use that 'feature' on 2000 when I need to create DLLs, but it
>> >has been a major problem with upgrades on MS software.
>> YOU CREATE DLLs. Oh - I'm crushed :-( I avoid MS stuff as much
>> as possible.
>I prefer and love (in an appropriate way) :-) FreeBSD. I
>tolerate the lesser OSes like Linux or Windows, and will
>non-prejudicially use them when appropriate. Frankly, I am
>on a mild re-education spree, where I am working to become
>comfortable and near-expert on Microsoft stuff.
You could equate that with summing up Sun Tzu's "Art of War" in about
three words. "Know your enemy". :-)
John S. Dyson
"David Schultz" <das...@HAL9000.homeunix.com> wrote in message news:al71rf$2iur$2...@agate.berkeley.edu...
a matter of note: when I become used to the Win2000 GUI environment, then
my previous Xwindows environment becomes temporarily uncomfortable. This
is probably akin to a slightly different keyboard layout syndrome.
However, knowing and understanding 'other ways' of doing things helps to
gain a deeper understanding for future development.
All in all, my original belief that an OS shouldn't just be a 'fair weather friend', but
should just do what it is supposed to persists as my view of a good OS. IMO, a good
release of FreeBSD (for example, NOT FreeBSD 2.0 that I was the primary screw-up), is
perhaps one of the very best common OSes for being performant when passively abused.
Some OSes tend to fall flat when simple heavy use cause them to choke, and
FreeBSD is one of the more robust choices.
Few OSes are robust against intentional abuse, but for simple heavy loading type
abuse, FreeBSD is really supurb, while most of my recently tried alternatives are
clunky and tend to be the proverbial 'fair weather friend.'
John
BraveNewWhirl
> You could equate that with summing up Sun Tzu's "Art of War" in about
> three words. "Know your enemy". :-)
>
> Bill
You can sum it in two words - but correctly - with "Don't fight."
John S. Dyson
"BraveNewWhirl" <braven...@postmaster.co.uk> wrote in message news:c4bc811a.02090...@posting.google.com...
destructive emotion called 'HATE.' All too often, some people are driven by 'hate'
against Microsoft, and in some ways that kind of 'hate' has created an evil
spawn (perhaps it isn't evil, but is socially/economically destructive.)
John
Lee Harr
> abuse, FreeBSD is really supurb
>
Yea. FreeBSD rocks...
I have a 1 ghz p3 which is used as a workstation and serves
2 x-terminals constantly and sometimes 3 or 4 (all running
KDE3)
Everything is smooth. Until netscape (shockwave or java
usually) starts using up all of the processor. Even then,
with something pinned at 75-80% CPU usage everything else
is usable.
It is a major task of reeducation to teach my students and
co-workers that when netscape freezes - THE COMPUTER IS
NOT "LOCKED UP"
Just hit ctrl-alt-escape, and click on the frozen window!
I would still like to be able to make sure that all processes
end when the user logs out... or get something more friendly
than top for killing runaway procs, but what is there works
great.
Patrick TJ McPhee
Bart Lateur <bart....@pandora.be> wrote:
% Chuck Swiger wrote:
%
% >Even with a "Windows Repair Disk", sufficient problems with
% >the Windows registry will require a reinstall.
%
% Hmm... I've been thinking. Are there any privilege levels set for the
% Registry? Or can any user modify anything? Because *that* would be
% extremely dangerous for viruses: change one file, and boom!
I have a lap-top with NT on it, and I have my normal user set up
as non-admin. I am able to replace parts of the operating system,
but I can't change _everything_ in the registry.
You can put security restrictions on just about anything in NT,
but by default, there is very little security.
david parsons
Bart Lateur <bart....@pandora.be> wrote:
>david parsons wrote:
>
>>>% There is nothing wrong w. the concept of a registry in terms of
>>>% stability..
>>>
>>>It seems like a significant single point of failure to me. It's one
>>>big file that gets changed all the time and you can't boot if it's
>>>corrupted.
>>
>> Reminds you of /etc, doesn't it?
>
>Except that /etc is a directory...
Which doesn't really matter -- it's still a single point of failure.
It's a BETTER point of failure, because it's on a filesystem (and
Unix filesystems tend to be much MUCH better at crash resistance
than however Windows did their registry.)
I've found that when I've had to deal with Windows that it's not so
much parts of the registry becoming corrupted, but the awful
recovery process which causes most of the trouble.
If MS had simply borrowed ffs and ffsfsck, their registry would not
be the butt of so much ill will.
____
david parsons \bi/ It's not that hard to put dirhashing on top of ffs.
\/
Richard Caley
>> You could equate that with summing up Sun Tzu's "Art of War" in about
>> three words. "Know your enemy". :-)
b> You can sum it in two words - but correctly - with "Don't fight."
IMO, more correct would be to combine the above as `Become your enemy
and lose for them'.
--
Mail me as MYFIR...@MYLASTNAME.org.uk _O_
|<
jpd
I think I'll take the entire Art of War over this kind of bickering, TYVM.
--
j p d (at) d s b (dot) t u d e l f t (dot) n l .
Three lines to explain one word is still more text than three words.
jpd
<MYFIR...@MYLASTNAME.org.uk> wrote:
> In article <c4bc811a.02090...@posting.google.com>,
> BraveNewWhirl (b) writes:
>
>>> You could equate that with summing up Sun Tzu's "Art of War" in about
>>> three words. "Know your enemy". :-)
>
> b> You can sum it in two words - but correctly - with "Don't fight."
>
> IMO, more correct would be to combine the above as `Become your enemy
> and lose for them'.
Lovely quote to take out of context.
<snarf>
`Become your enemy and lose for them'.
-- Richard Caley, cubfm
</snarf>
David Magda
[...]
> Just hit ctrl-alt-escape, and click on the frozen window!
[...]
Have you mapped this to a "kill window" function? Which window
manager?
--
David Magda <dmagda at ee.ryerson.ca>
Because the innovator has for enemies all those who have done well under
the old conditions, and lukewarm defenders in those who may do well
under the new. -- Niccolo Machiavelli, _The Prince_, Chapter VI
John S. Dyson
"jpd" <read_t...@do.not.spam.it> wrote in message news:slrnanima2.28q...@oli252.rolahola.tudelft.nl...
> On Thu, 5 Sep 2002 18:39:11 -0500, John S. Dyson <dy...@iquest.net> wrote:
> > "BraveNewWhirl" <braven...@postmaster.co.uk> wrote in message
> > news:c4bc811a.02090...@posting.google.com...
> >> >
> >> > You could equate that with summing up Sun Tzu's "Art of War" in about
> >> > three words. "Know your enemy". :-)
> >> >
> >> > Bill
> >>
> >> You can sum it in two words - but correctly - with "Don't fight."
> >>
> > Even more basic: Don't base your actions, behaviors or thoughts upon the
> > destructive emotion called 'HATE.' All too often, some people are driven
> > by 'hate'
> > against Microsoft, and in some ways that kind of 'hate' has created an evil
> > spawn (perhaps it isn't evil, but is socially/economically destructive.)
>
>
> I think I'll take the entire Art of War over this kind of bickering, TYVM.
>
1) If you make war, make sure that it is justified. Don't make war because
of hatred or ego.
2) If one hates, then it is more destructive to those with the emotional problem
than the party that is the object of the hate.
3) Allowing 'hate' to even be slightly considered in the guidance of
a project is self-defeating. 'Winning' is a rather specious goal in a large
economy, but 'excellence' and/or 'success' is indeed a useful and valid
goal.
4) When doing a overly 'me-too' development, and not studying the competition,
it doesn't improve the 'me-too' project, and only serves to artificially massage a
developer's ego.
5) Rather than trying to compete against Microsoft, Linux or *BSD, proving
that any one group is doing everything wrong, it is rather more efficient to
understand what they are doing correctly, and learn from everyone's mistakes.
In the cases for FreeBSD where I had to 'reinvent' code, it was a technically
wasted effort, but was necessary for the success of FreeBSD and the
legal constraints. There is so little 'innovation' in a 1980's OS kernel, and the
technology is so common knowledge, David Greenman and myself reimplemented
the AT&T denied sections of the OS code within about 2wks. When conceptually
new code and features were added, then that is where the bugs crept in. Luckily,
the AT&T sections of the code weren't very complex, and the reimplementation
wasn't costly. However, reimplementing those sections of the code for
developer ego reasons would have been wasteful, and IMO: insane.
For example, if Microsoft actually used the BSD TCP/IP stack as a template,
even perhaps using portions of the code, they are showing wisdom in this
case. If Microsoft had decided that BSD code is 'hippie, communist' code, useless
except for those who smoke pot all day, and it is infested with left wing ideas :-), then
they are being stupid. TCP/IP stacks don't have a political agenda, and free
software is meant for everyone to reuse for their own commercial or free
applications :-). Now, if the TCP/IP code was restrictively encumbered, perhaps
with redistribution requirements, then there might be a reason to avoid using
it. The kind of software with redistribution limitations and/or requirements isn't
really free, but is indeed appropriately called 'open source' by common parlance.
When working on the MACH VM code, I didn't throw it all out with the proclaimation
that it had too many bugs. Rather, I had reviewed the limitations, and had
determined that with it's very rough edges, it was a good start for a complete
VM system. I didn't expunge everyone elses code so as to be able to proclaim
full ownership. For some reason, I was (and still am) just interested in the
code working, and it being licensed in a way that it can be used. Frankly, the
MACH VM code has some cool ideas in it, and I learned alot from it -- it might
need to be supplanted for better SMP behavior, but it is still wise to continue
to learn from the codebase...
I am not bickering, but simply sharing a little bit of wisdom that war (of any sort)
should be done with great care and consideration. Hate is more destructive against
those who hate, rather than the object of the hatred.
I don't have ANY problems in looking at Microsoft software, or Linux software,
but by doing so, I am not studying the 'enemy.' At most, to me, Microsoft, Linux
or even FreeBSD might be the 'competition'. However, I cannot fathom where
hate makes any sense -- well, unless someone has been deceptive or has
other personal or ego agendas. There is always an opportunity to
learn from other peoples' work. For some reason, I don't feel like it is an
insult to my own ego to learn from other peoples' work, and it is incredibly
unwise to try to re-invent the world, especially without learning from other peoples'
public efforts.
John
Donn Miller
> 1) If you make war, make sure that it is justified. Don't make war because
> of hatred or ego.
> 2) If one hates, then it is more destructive to those with the emotional problem
> than the party that is the object of the hate.
The perception is that FreeBSD is for people who love unix, and Linux is
for people that love Microsoft. Well, I'm not sure I hate Microsoft,
but I have enough experience with their software (about 11 years) to
know that its quality is not great enough to ensure bang for the buck.
In fact, IME, MS software delivers extremely low bang for the buck. For
me, I simply find it's not worth paying the price for MS software. I
take all the elements in the equation: the monopolistic tactics,
quality, price, licensing, and over design.
The reason I prefer to run FreeBSD is not to spite MS, but because it
delivers extremely high bang for the buck. I find it a very fun OS to
work with. Other factors are that I started out with *BSD back in the
early days, with Ultrix and whatever version of SunOS came on Sparc
IPX's. FreeBSD reminds me very much of those early versions of *nix
that I started out with that I enjoyed so much.
Also, after doing a little studying from a distance, I've found that I
really like the BSD licensing scheme, and the fact that it can encourage
competition amoung various commerical software vendors. Linux/GPL
advocates (read: immature kooks) will tell you that BSD is bad because
'ol MS can suck in all the BSD code they want, and they don't have to
contribute something back to the tree. They insinuate that this makes
MS' monopoly a lot stronger, because they have a lot of high-quality
code to work with. But my rebuttal is that this isn't necessarily so,
because, firstly, the way you use the code is as important as the
quality of the code itself. Secondly, since any company can use the BSD
code at will in a closed-source product and/or OS, this provides more
competition for any commercial monopoly.
Also, GPL advocates will point out that the BSDL is bad, because not
only can companies can use the code in closed-source proprietary
software, but they aren't required to contribute back to the codebase.
my rebuttal here is that, firstly, do we necessarily want companies
(esp. MS) contributing code the codebase, when they're famous for
writing horrible code in the first place? Secondly, just because
companies aren't required to contribute code, doesn't mean they aren't
going to. Isn't it funny how GPL/RMS worshippers will imply that just
because companies aren't required to contribute to BSDL'd code, that
automatically means they won't? They conveniently leave out that part
in their arguments.
Also, Linux advocates (read: flaming zealots) will point out that Linux
is many millions of times better than BSD, because "anyone can
contribute code to the Linux kernel". Linux zealots claim that Linux is
"for the people, by the people" for this very reason. They also claim
FreeBSD is inferior because, in their erronious view, there's only a
small core team, and no one can contribute to the FreeBSD code. Hence,
FreeBSD is an elitist, closed system, in their view. My rebuttal is,
firstly:
http://www.freebsd.org/doc/en_US.ISO8859-1/articles/contributing/index.html
Secondly, even if it were true that trillions of people are contributing
code to the Linux kernel, someone has to maintain the code that the
"many people" have contributed. If Linux is indeed a system that is
"for the people, by the people", everyone has to be familiar with
everyone else's code. It also leads to bloating.
But, fortunately, advozealots are wrong, and as the URL I gave points
out: "Contrary to what some people might have you believe, you do not
need to be a hot-shot programmer or a close personal friend of the
FreeBSD core team to have your contributions accepted." The current GPL
propaganda is that no one outside of a tiny, elitist core team can
contribute code.
Donn Miller
Donn Miller wrote:
> The perception is that FreeBSD is for people who love unix, and Linux is
> for people that love Microsoft.
^^^^
hate
John S. Dyson
"Donn Miller" <dmmi...@cvzoom.net> wrote in message news:3d798...@corp.newsgroups.com...
> As you can tell, I never proof-read my posts before hitting "send".
>
> Donn Miller wrote:
>
> > The perception is that FreeBSD is for people who love unix, and Linux is
> > for people that love Microsoft.
> ^^^^
> hate
>
who might have been reading it :-). Perhaps it was 'Freudian' in some sense.
(Maybe Linux is for those who would otherwise 'love' Microsoft, but have found
a Pied Piper to draw them from a hard life to a different kind of worse-than-optimal
condition...) Being in a world of 'just using' Linux, and comparing it's use in
applicable circumstances with 'just using' Win2000, Linux can sure seem friendlier.
However, where most of the advantages exist, FreeBSD will tend to do even
better, especially in heavily loaded systems. For developers, the free license
that is generally associated with the runtime portions of FreeBSD is an addiitional
bonus for developers, when control of a given developer's own work product could be
used for feeding heir own families -- actually strongly supporting the developer
doing developer things, rather than having to depend upon redistributor schemes
and almost guaranteeing that the control of almost any works be ceded to the
distributors... The free license terms allow the developer better control of
their add-on works by giving more freedom of licensing add-on works and
modifications... An add-on developer cannot physically rescend the freeness
of a software package that they use as a source for their add-on work, but
an anti-developer license can easily bias the control of add-on works towards
the distributor rather than the developer.
When looking at almost all of my animosity about Linux and the GPL, it was
almost 100% related to the rhetoric and not the actual license itself. The
Microsoft-hate thing bothers me (even though I do believe that SIMPLY
disliking Microsoft policies and behavior is quite valid and healthy -- it is the
intense hatred that seems destructive to those who have such feelings.) Also
the rhetoric about the GPL being a license of free software is often quite deceptive,
even though it would be valid to claim that GPLed software is often easier/freer
to re-use than most (not all) commercial software. It is just that there are
numerous "'freer' than GPL" licenses, so strongly claiming that GPL is a license
of free software is deceptive by forgetting about (ignoring) the truly more free
licenses.
I really like to avoid general Linux bashing, general GPL bashing or general
Microsoft bashing. However, specific criticism of those OSes, licenses and
overly strong advocacy is just as valid as specific criticism about FreeBSD
issues.
John
Alexander Viro
>> As you can tell, I never proof-read my posts before hitting "send".
>>
>> Donn Miller wrote:
>>
>> > The perception is that FreeBSD is for people who love unix, and Linux is
>> > for people that love Microsoft.
>> ^^^^
>> hate
>>
>Yes, on first reading, your post appeared to be flame-bait for Linux advocates
>who might have been reading it :-). Perhaps it was 'Freudian' in some sense.
<shrug> one soundbite is worth another - both reek with advocacy. Let me
put it that way - the difference between Linux, FreeBSD, Windows, etc.
advocates is that between syphilitic whores prefering this, this and that
positions. Dealing with any of them is... unsanitary, regardless of your
own preferences.
--
"You're one of those condescending Unix computer users!"
"Here's a nickel, kid. Get yourself a better computer" - Dilbert.
Bill Vermillion
John S. Dyson <dy...@iquest.net> wrote:
>"jpd" <read_t...@do.not.spam.it> wrote in message news:slrnanima2.28q...@oli252.rolahola.tudelft.nl...
>> On Thu, 5 Sep 2002 18:39:11 -0500, John S. Dyson <dy...@iquest.net> wrote:
>> > "BraveNewWhirl" <braven...@postmaster.co.uk> wrote in message
>> > news:c4bc811a.02090...@posting.google.com...
>> >> >
>> >> > You could equate that with summing up Sun Tzu's "Art of War" in about
>> >> > three words. "Know your enemy". :-)
>> >> You can sum it in two words - but correctly - with "Don't fight."
>> > Even more basic: Don't base your actions, behaviors or
>> > thoughts upon the destructive emotion called 'HATE.' All too
>> > often, some people are driven by 'hate' against Microsoft,
>> > and in some ways that kind of 'hate' has created an evil
>> > spawn (perhaps it isn't evil, but is socially/economically
>> > destructive.)
>> I think I'll take the entire Art of War over this kind of
>> bickering, TYVM.
>I saw no bickering. However, the following is important to consider:
>1) If you make war, make sure that it is justified. Don't make
>war because of hatred or ego.
>2) If one hates, then it is more destructive to those with the
>emotional problem than the party that is the object of the hate.
>3) Allowing 'hate' to even be slightly considered in the
>guidance of a project is self-defeating. 'Winning' is a rather
>specious goal in a large economy, but 'excellence' and/or
>'success' is indeed a useful and valid goal.
Wow - just because I mentioned Sun Tzu I didn't expect a thread and
flame-war to develop.
Though "Art of War" was written centuries ago that book became
popular in the '80s in the corporate world, and was a "must read"
for many young wanna-be executives. When applied to this
century's civiliation Sun Tzu's concepts can be applied to any
competitive venture. Too many seem to take the word 'war'
literaly and think only in military terms.
Bill Vermillion
>> 1) If you make war, make sure that it is justified. Don't make war because
>> of hatred or ego.
>> 2) If one hates, then it is more destructive to those with the emotional problem
>> than the party that is the object of the hate.
>The perception is that FreeBSD is for people who love unix, and
>Linux is for people that love Microsoft. Well, I'm not sure I
>hate Microsoft, but I have enough experience with their software
>(about 11 years) to know that its quality is not great enough
>to ensure bang for the buck. In fact, IME, MS software delivers
>extremely low bang for the buck. For me, I simply find it's not
>worth paying the price for MS software. I take all the elements
>in the equation: the monopolistic tactics, quality, price,
>licensing, and over design.
>The reason I prefer to run FreeBSD is not to spite MS, but because it
>delivers extremely high bang for the buck. I find it a very fun OS to
>work with. Other factors are that I started out with *BSD back in the
>early days, with Ultrix and whatever version of SunOS came on Sparc
>IPX's. FreeBSD reminds me very much of those early versions of *nix
>that I started out with that I enjoyed so much.
My first Unix type system was on Xenix back in 1983. That was
before AT&T started the code bloat with System III and continued
that vein with SysV. For meBSD was like going home again.
Donn Miller
> I really like to avoid general Linux bashing, general GPL bashing or general
> Microsoft bashing. However, specific criticism of those OSes, licenses and
> overly strong advocacy is just as valid as specific criticism about FreeBSD
> issues.
Do you know anything about the types of schedulers used in FreeBSD
compared to Linux? I've heard that the scheduler in FBSD is what's
known as a "priority feedback" scheduler. I was also reading some stuff
on kernetrap that claims that Linux's scheduler has some flaws that the
core team is trying to correct. Supposedly, the current Linux kernel
doesn't pre-empt processes very efficiently (hence Robert Love's work),
and it's also an O(n) scheduler, whatever that means. One guy claims
that sometimes it's even O(n^2). There was some hub-bub on the lkm
about a push for an O(1) behavior, which is supposed to improve Linux'
multi-tasking.
Also, there were some blurbs about Linux caching as much memory as
possible, and then swapping under pressure. But, I saw something about
dynamically-shrinking caches on kerneltrap, which is supposed to cache
smaller amounts of RAM, and change the cache based on avail. physical
memory. I'm guessing this is what FreeBSD has been doing overall.
It looks like the Linux kernel has a few BSD-isms, though, such as sysctl.
John S. Dyson
"Donn Miller" <dmmi...@cvzoom.net> wrote in message news:3d7a1...@corp.newsgroups.com...
> John S. Dyson wrote:
>
> > I really like to avoid general Linux bashing, general GPL bashing or general
> > Microsoft bashing. However, specific criticism of those OSes, licenses and
> > overly strong advocacy is just as valid as specific criticism about FreeBSD
> > issues.
>
> Do you know anything about the types of schedulers used in FreeBSD
> compared to Linux? I've heard that the scheduler in FBSD is what's
> known as a "priority feedback" scheduler. I was also reading some stuff
> on kernetrap that claims that Linux's scheduler has some flaws that the
> core team is trying to correct. Supposedly, the current Linux kernel
> doesn't pre-empt processes very efficiently (hence Robert Love's work),
> and it's also an O(n) scheduler, whatever that means. One guy claims
> that sometimes it's even O(n^2). There was some hub-bub on the lkm
> about a push for an O(1) behavior, which is supposed to improve Linux'
> multi-tasking.
>
FreeBSD scheme, and couldn't significantly improve on it. The guys at Berkeley
and AT&T apparently really knew what they were doing. I worked on some
schemes that would allow for quicker rescheduling and/or avoid some of the fairness
issues when mixing I/O with compute bound processes, but with little success.
>
> Also, there were some blurbs about Linux caching as much memory as
> possible, and then swapping under pressure. But, I saw something about
> dynamically-shrinking caches on kerneltrap, which is supposed to cache
> smaller amounts of RAM, and change the cache based on avail. physical
> memory. I'm guessing this is what FreeBSD has been doing overall.
>
The feedback mechanisms in FreeBSD VM were based more upon 'stability'
rather than 'policy.' Increasing the amount of filesystem cache as an impulse
function, when in a state of heavy memory load it will tend to destablize the
paging system, and perhaps cause thrashing. Because of the natural delays
in the paging system, allowing for a rapid increase in file cache memory
pressure will actually hurt performance, especially if the purpose of the growing
memory is to increase the amount of deferred writes beyond a common
sense amount. FreeBSD really does have a scheme that allows the use of
relatively unused memory as filesystem cache, and avoids excess pressure on
process memory.
One reason why FreeBSD will tend to converge to a working set, actually quieting
down during heavy paging, is that it is actually fairly (not perfectly) control-system
stable WRT the paging behavior.
John
jpd
[snip!]
> Though "Art of War" was written centuries ago that book became
> popular in the '80s in the corporate world, and was a "must read"
> for many young wanna-be executives. When applied to this
> century's civiliation Sun Tzu's concepts can be applied to any
> competitive venture. Too many seem to take the word 'war'
> literaly and think only in military terms.
<reflection>
Seems to me the 80's were full of ``gurus of the hour'', everyone had
to have read their works and such. And then we had a new Big Example
to follow. Seemed some people made a decent sum out of the hype, too.
</reflection>
IMO, blindly following One Work in religious faith, will not magically
Make All Wrongs Right. Doesn't have to stop you learn from text tho.
I think I like the _Art of War_. I also liked _The Ruler_[0], until I
read a comment that it is a very dark joke. Then I liked it even more.
[0] _Il Principe_ by Machiavelli, title also translated as _The Prince_.
John S. Dyson
"Alexander Viro" <vi...@weyl.math.psu.edu> wrote in message news:alcc12$l...@weyl.math.psu.edu...
> In article <dyge9.1475$I4.2...@news.iquest.net>,
> John S. Dyson <dy...@iquest.net> wrote:
> >
> >"Donn Miller" <dmmi...@cvzoom.net> wrote in message news:3d798...@corp.newsgroups.com...
> >> As you can tell, I never proof-read my posts before hitting "send".
> >>
> >> Donn Miller wrote:
> >>
> >> > The perception is that FreeBSD is for people who love unix, and Linux is
> >> > for people that love Microsoft.
> >> ^^^^
> >> hate
> >>
> >Yes, on first reading, your post appeared to be flame-bait for Linux advocates
> >who might have been reading it :-). Perhaps it was 'Freudian' in some sense.
>
> <shrug> one soundbite is worth another - both reek with advocacy. Let me
> put it that way - the difference between Linux, FreeBSD, Windows, etc.
> advocates is that between syphilitic whores prefering this, this and that
> positions. Dealing with any of them is... unsanitary, regardless of your
> own preferences.
>
or not I like FreeBSD, or tolerate Linux/Win2000, extreme advocacy based upon emotion
from either side is tiresome. If an individual developer had done innovative work on Linux
or FreeBSD, then some 'bias' would be expected. Even being an advocate based upon
a concrete feature set is also reasonable. However, blind advocacy, and cult-like behavior
(either in a defensive sense from the predominant advocacy position, or in an aggressive
jihad (in the bad sense) trying to topple the predominant player in the market) aren't really
helpful for the advocates' rhetorical position.
For example, I am typing on a Win2000 system that I use for casual computing and frankly,
it works very well for it's purpose. However, the system is sitting behind a free OS firewall,
and is peers on a LAN with various other machines running free, commercial and open source
OS code. Even this machine that I am typing on is a dual boot, where the unix disk allocation
is actually significantly larger than the Win2000 config.
In some ways, especially where the additional memory could be used for improved performance,
the Win2000 runs much more slowly than the machine being
booted as a FreeBSD system. However, I also don't have the source code or resources to
run a specific very performant circuit analysis program as a FreeBSD application, so am stuck with
running it on Win2000. Do I feel a grudge because I have to run Win2000? No, of course not.
In fact, the need to run Win2000 fairly often has inspired me to learn more about the details
of the OS, and I am not a weaker or inferior person because I have increased my know-how.
Of course, I'd rather be running a more performant OS, whether it is of the unix-ilk, or of the
WinNT ilk. However, luckily, my system is large enough that the applications that I like
to run on Win2000 don't 'load' it heavily, but it is somewhat disappointing that the unused
memory isn't used more effectively for caching. Some of the privacy and security issues
with Win2000 (and more importantly, WinXP) are of great concern to me, and should be
of great concern to anyone whose pocketbook could be picked by intruders, or personal
information by Microsoft as a predator. The natural answer to this shouldn't be a rallying
cry, but should be a 'voting with one's pocketbook', along with criticism of the privacy
intrusion.
John
David Schultz
> Do you know anything about the types of schedulers used in FreeBSD
> compared to Linux? I've heard that the scheduler in FBSD is what's
> known as a "priority feedback" scheduler. I was also reading some stuff
> on kernetrap that claims that Linux's scheduler has some flaws that the
> core team is trying to correct. Supposedly, the current Linux kernel
> doesn't pre-empt processes very efficiently (hence Robert Love's work),
> and it's also an O(n) scheduler, whatever that means. One guy claims
> that sometimes it's even O(n^2). There was some hub-bub on the lkm
> about a push for an O(1) behavior, which is supposed to improve Linux'
> multi-tasking.
The FreeBSD scheduler has always been O(1), and I'll bet that if the
Linux folks have any sense, their default scheduler (also priority
feedback) is O(1) as well. But that says nothing about how good the
scheduler is at choosing which process should run. Moreover,
different schedulers are appropriate for different workloads. Do you
care about long-running tasks or interactive tasks? Which is more
important, fairness or real-time guarantees? In some situations, an
/inherently/ O(n) scheduling algorithm, such as lottery scheduling,
may be appropriate. There are patches for at least three schedulers
for FreeBSD, and probably at least as many for Linux, though most
people are happy with the standard ones, as John pointed out.
The issue of preemption is completely different, and much more
controversial. With preemption, interactive processes seem a bit more
snappy because they can run in response to an event even when another
process is in the middle of a long-running system call. But you pay a
significant performance penalty for that---more context switches,
additional locking in the kernel, and more overall complexity. So
it's an issue of response time versus throughput. In a real-time
operating system, the improved response time can be crucial, but
FreeBSD and Linux are decidedly not real-time systems. The FreeBSD
folks have not yet decided whether to support preemption, but if they
do, I wouldn't expect to see it until 6.0.
> Also, there were some blurbs about Linux caching as much memory as
> possible, and then swapping under pressure. But, I saw something about
> dynamically-shrinking caches on kerneltrap, which is supposed to cache
> smaller amounts of RAM, and change the cache based on avail. physical
> memory. I'm guessing this is what FreeBSD has been doing overall.
The Linux VMM has some problems with self-tuning, so I'm told. I
looked at some of it a few months ago, and it seems to have quite a
few special-purpose hacks to address specific cases that are probably
handled automatically in FreeBSD. But Rik (and probably others) have
been looking carefully at what FreeBSD has done, so I imagine that
will show as the Linux virtual memory system matures.
H. Dziardziel
<dy...@iquest.net> wrote:
>
>
snip
>
>In some ways, especially where the additional memory could be used for improved performance,
> the Win2000 runs much more slowly than the machine being
>booted as a FreeBSD system. However, I also don't have the source code or resources to
>run a specific very performant circuit analysis program as a FreeBSD application, so am stuck with
>running it on Win2000. Do I feel a grudge because I have to run Win2000? No, of course not.
>In fact, the need to run Win2000 fairly often has inspired me to learn more about the details
>of the OS, and I am not a weaker or inferior person because I have increased my know-how.
>
>Of course, I'd rather be running a more performant OS, whether it is of the unix-ilk, or of the
Are you the one who has coined (coded?) "performant'? Why not
"better performing"? There must be nuances I am overlooking
Thanks.
>
snip
Bill Vermillion
jpd <data...@elsewhere.net> wrote:
>On Sat, 07 Sep 2002 12:57:14 GMT, Bill Vermillion <b...@wjv.comREMOVE> wrote:
>[snip!]
>> Though "Art of War" was written centuries ago that book became
>> popular in the '80s in the corporate world, and was a "must read"
>> for many young wanna-be executives. When applied to this
>> century's civiliation Sun Tzu's concepts can be applied to any
>> competitive venture. Too many seem to take the word 'war'
>> literaly and think only in military terms.
><reflection>
>Seems to me the 80's were full of ``gurus of the hour'', everyone had
>to have read their works and such. And then we had a new Big Example
>to follow. Seemed some people made a decent sum out of the hype, too.
></reflection>
The sheep always had to have a new leader. A lot was just to be
different I think.
>IMO, blindly following One Work in religious faith, will not magically
>Make All Wrongs Right. Doesn't have to stop you learn from text tho.
As evinced so well by Joseph Campbell in "The Powers of Myth".
Seem to be more similarities than differences.
>I think I like the _Art of War_. I also liked _The Ruler_[0], until I
>read a comment that it is a very dark joke. Then I liked it even more.
>[0] _Il Principe_ by Machiavelli, title also translated as _The
>Prince_.
Books that have stood the test of time.
Arnoud Galactus Engelfriet
Donn Miller <dmmi...@cvzoom.net> wrote:
>I can see where the GPL can foster software patents. Normally, when
>software companies take the BSD code, and incorporate it into their own
>products with changes, they don't have to make the changes available to
>everyone, as the GPL stipulates. So, there's no need for a patent. But
>since the GPL states that any modifications to the GPL'd code be made
>public, well, I can see the need for additional protection by patenting
>the code. Too bad the GPL doesn't state anything about patents. Here,
>the complexity of the GPL works against itself, and the simplicity of
>the BSDL is the "winner".
If you distribute GPL-licensed code, you can't use your own
patent to restrict usage or redistribution of that code by
the recipient(s). GPL section 6:
6. Each time you redistribute the Program (or any work based on the
Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject
to these terms and conditions. You may not impose any further
restrictions on the recipients' exercise of the rights granted
herein. You are not responsible for enforcing compliance by third
parties to this License.
If you enforce a patent against the recipient, you're imposing
a further restriction on his exercise of the rights he got
under the GPL. That means you lose your rights under the GPL
and your product suddenly is infringing on someone else's
copyright.
Arnoud
--
Arnoud Engelfriet, (almost) Dutch patent attorney - Speaking only for myself
Patents, copyright and IPR explained for techies: http://www.iusmentis.com/
Richard Caley
age> If you distribute GPL-licensed code, you can't use your own
age> patent to restrict usage or redistribution of that code by
age> the recipient(s). GPL section 6: [etc]
This tickles something I have wondered/worried about. Let's say I have
permission from a patent holder to use some patented technique in an
extenstion I am constructing to some GPLed code. Can a patent holder
give enough permission that the code can be distributed under the GPL
without effectively sinking the patent?
This bites if, as some have proposed, some community funded
organisation should be set up to patent everything under the sun and
then give everyone a licence to use it all, as a mechanism to block
commercial nusance patents.
(of course, hitting the US patent office and the EU equivalent with a
REALLY BIG clue stick is more what is needed, but the odds against
that... )
BraveNewWhirl
> Though "Art of War" was written centuries ago that book became
> popular in the '80s in the corporate world, and was a "must read"
> for many young wanna-be executives. When applied to this
> century's civiliation Sun Tzu's concepts can be applied to any
> competitive venture. Too many seem to take the word 'war'
> literaly and think only in military terms.
>
> Bill
Not only that. Quite a few folks think it is concerned with competition.