OSNews

Wait, what?

Microsoft used to use BSD code in their tcp/ip stack and ftp.exe has some BSD code. What else?

Apple's kernel isn't BSD at all, AFAIK. I thought only userland stuff was? Or maybe I have that backwards.

I quote:

"Sockets are based on the UNIX sockets implementation in the Berkeley Software Distribution (BSD, release 4.3) from the University of California at Berkeley."

Source: http://search.microsoft.com/results.aspx?mkt=en-US&setlang=en-US&q=...

Apple's kernel isn't BSD at all, AFAIK.

It is originally based on Lites, which is a BSD-based server on top of the Mach microkernel. Much of the userland code is also from various BSDs.

I followed your link and do not see the text you quoted.

I'm piggy-backing on your first post so that others can read an honest, unbiased, and well-researched response to this paper.

FUD-buster: this paper has one severe flaw amongst otherwise decent logical analysis. It is not FUD from GPL fanatics, but reads more like the work of a BSD supporter that falls on the anti-relicensing side of the BSD community. The relevance of this analysis is not in any way tied to Australian law except in the one place where noted in the paper. Besides the critical flaw, the analysis seems to work in any common law jurisdiction (in my opinion).

The big flaw is that clause 2 of the BSD license, which contains the main grant of redistribution in source or binary form, modified or unmodified, indicates that the conditions of this grant are "following." Then, clause 3 requires the retention of "this list of conditions." It follows that the BSD license requires the retention of clauses 1, 3, 4, 5, and the disclaimer, but NOT clause 2.

So while redistribution of BSD code in source or binary form, modified or otherwise, requires the retention of the original copyright, the list of conditions, and the disclaimer, it does not require the retention of the main license grant, without which the license does not permit redistribution in any form (as per default copyright terms).

In short, the default application of the BSD license does seem to involve viral license propagation, but it does not prevent distributors from removing clause 2, thus preventing the propagation of redistribution rights to the recipient. It would be smart for proprietary users of BSD code to remove clause 2 from their products to bolster their defense against redistribution.

Further, as others note, the BSD license does not require the availability of corresponding source code to licensed recipients of the object code. However, it does allow the redistribution of source code corresponding to licensed object code should the recipient happen to obtain it. So, if I buy a proprietary BSD package in binary form, and some cracker steals and hosts the source code from the author's development box, then my license actually permits me to lawfully distribute the leaked source code. The cracker, it seems, would be on the wrong side of the law.

Despite its major flaw, I believe this paper achieved its intended purpose of shedding some light on the confusion surrounding the BSD license. It *is* more restrictive than you might think, because those conditions you are required to retain *do* have legal ramifications. Is a redistribution of BSD code, modified or not, required to carry the BSD license? I don't think so. But it does seem to require a subset of the BSD terms to carry with redistributed works, a subset which does not seem to grant a license to redistribute in any form.

Finally, although distributors are permitted to remove clause 2, I don't believe many of them actually do in practice. With this clause in place, any modifications of the source are redistributable, at least within the same source file, and possibly beyond.

Responsible comments on this reading are appreciated...

Edited 2007-01-17 00:42

Obviously you want to flame the lawyer with a legal opinion that points out a few things that might be different from what is commonly known.

Meanwhile all the BSD's are choke full of GPL'ed software starting with the GPL'ed GCC compiler. Yes, there are BSD licensed compliers but they arent used by these BSD flavors, so let's not turn this into a GPL vs BSD argument. Note that the paper doesnt get into this sillyness.

Edited 2007-01-16 15:34

If you didn't want a flame was you should learn to bite your tounge.

>all the BSD's are choke full of GPL'ed software starting with the GPL'ed GCC compiler.

The system itself works without any GPL code (and it's more than just a kernel), it isn't possible at all because of the GPL.
Most of the GPL software resides in ports and gcc is an addon, it's not essential for a running system. So yes, it's not about GPL vs BSD, but these are essential details!

"The system itself works without any GPL code (and it's more than just a kernel), it isn't possible at all because of the GPL. "

Yes. You can put up a kernel and base tools without any GPL'ed code. Works" != Usable for the significant portion of users though. We need more software out there and the large majority is licensed under GPL. Look at http://www.dwheeler.com/essays/gpl-compatible.html for a detailed analysis.

" So yes, it's not about GPL vs BSD, but these are essential details!"

BSD folks like Theo are so much fanatical that they wont allow GPL'ed software into the base system and not even Apache 2 because he considers it non-free but the whole system is compiled with GCC anyway. Linux distributions are however happy to put BSD code in their system.

Same for NetBSD and FreeBSD. So yes, you require GCC to even bootstrap the system. Try compiling it with any other compiler and watch it all break down.

But you don't need it for a running system - think of embedded systems, special server applications and so on.

>Linux distributions are however happy to put BSD code in their system.

Yes because GPL will eat BSD license. And *BSD people are proud in writing good code without any restrictions.

>Works" != Usable for the significant portion of users though

"The significant" <- depends on your environment, if you are a desktop-user ... but there it depends too and it's possible.

>Theo are so much fanatical

And therefore most of the best and secure software is from OpenBSD - you'll find it in Windows, Linux, *BSD, Solaris .... Because Theo is fanatic about openness and security.

>You can put up a kernel and base tools

Many people call this an operating system.

"Yes because GPL will eat BSD license. And *BSD people are proud in writing good code without any restrictions. "

GPL license is not some cannibal and your words remind me of Microsoft's FUD.Learn to stick to legal terms when talking about licenses.

"The significant" <- depends on your environment, if you are a desktop-user ... but there it depends too and it's possible. "

No. Significant means the large majority.

"
And therefore most of the best and secure software is from OpenBSD - you'll find it in Windows, Linux, *BSD, Solaris .... Because Theo is fanatic about openness and security.

It is because BSD licensed code can be easily forked off and merged into proprietary systems like Windows. Whether OpenBSD is secure or not has nothing to do with whether GPL'ed components are allowed in the base system.


"Many people call this an operating system."

Precisely. OS is useless without applications and large majority of Free software applications are written under the GPL license. So for usable applications you are quite probably relying on GPL'ed code.




Many people call this an operating system."

>GPL license is not some cannibal and your words remind me of Microsoft's FUD.Learn to stick to legal terms when talking about licenses.

Call it whatever you want. If you take BSD code and put it under the GPL, you will have GPL-code afterwards. So yes you can tell people about the vast majority of GPL-code - tha's a nice thinking. Call it "viral" instead and if you have problems with the term, think about "free software" and "open source" - it's nonsense too without the proper context!

>Precisely. OS is useless without applications and large majority of Free software applications are written under the GPL license. So for usable applications you are quite probably relying on GPL'ed code.

But you don't have to use it, try this with the Linux kernel. Period!

Apropos licenses,

Apache - GPL?
PHP - GPL?
Mozilla - GPL?
Enlightenment - GPL?
Firefox - LGPL, GPL, MPL
Qemu - many licenses
Vim - GPL?
w3m - GPL?


and so on ...

Yes the majority is GPL software, but from the beginning? And can of course build a usuable system without GPL software - to compile it I have to use GCC. This is no war against GPL software, I'am using it too - but don't spread such a nonsense because of your very own horizon.

So yes, you require GCC to even bootstrap the system. Try compiling it with any other compiler and watch it all break down.

Actually even though the BSDs use GCC, OpenBSD is able to be completely rebuilt using Tendra, a BSD licensed compiler. It's a pain but possible.

Meanwhile all the BSD's are choke full of GPL'ed software starting with the GPL'ed GCC compiler. Yes, there are BSD licensed compliers but they arent used by these BSD flavors, so let's not turn this into a GPL vs BSD argument. Note that the paper doesnt get into this sillyness.

The BSD OS's simply include GPL software to execute. The kernel and core system do not and could not use any GPL code. All the while GPL software can consume and create derivative works using BSD code, but BSD code can't (and still remain BSD that is).

Your jab at how "the BSD's are choke full of GPL'ed software" itself is an implied GPL vs BSD argument. And the sillyness that the paper *does* get into is essentially trying to make the BSDL more GPL-like.

But even when you say that any BSD derived works must be covered under the BSD license, that still does not change anything since as deadlinkous stated, the license itself does not require source distribution.

"
The BSD OS's simply include GPL software to execute. The kernel and core system do not and could not use any GPL code. All the while GPL software can consume and create derivative works using BSD code, but BSD code can't (and still remain BSD that is). "

Well if you want only BSD code, you cant use any other licensed code. That is a no brainer to conclude. You can mix and match both of them (assuming it is the 3 clause BSD license).

"
Your jab at how "the BSD's are choke full of GPL'ed software" itself is an implied GPL vs BSD argument"

Of course not. It is just pointing out the irony of name calling in the OP's "GPL freaks"

"And the sillyness that the paper *does* get into is essentially trying to make the BSDL more GPL-like. "

Nope. It does point out a very valid legal opinion by a lawyer that you cant simply relicense any code licensed under BSD as many people do assume.

"But even when you say that any BSD derived works must be covered under the BSD license, that still does not change anything since as deadlinkous stated, the license itself does not require source distribution."

Obviously the ability to relicense is a critical difference between MIT X11 license and the BSD licenses. You might want to actually consider that for sometime before proclaiming that it doesnt change anything.

Nope. It does point out a very valid legal opinion by a lawyer that you cant simply relicense any code licensed under BSD as many people do assume.

I disagree about the weight you give to the article. They are merely educated opinions based off of Australian law , and neither US nor Australia have had any hard rulings on such matters.

But again, for all practical purposes, how does it change anything? The ONLY difference it makes to maintain the BSD license required for distribution, are that you maintain the copyright notice and that the owners and contributors of that BSDL code cannot be named for endorsement without their consent. Yet these are inconsequential for those who are already using BSD code today because they are not violating any of those terms. MS still maintains copyright notices (ok, so you'll have to look hard) and they aren't using the authors of the BSDL works in promoting their products either.

GCC compiled programs are not infected with GPL.

"Anyone else noticed a lessening of linux distro review type articles lately???"

Not really. Except that there hasnt been major releases now. Many of the popular distributions work on a bi yearly cycle. When the next stream of releases come around somewhere in April 2007 timeframe, you can start getting reviews again. Maybe you miss them so much.

>>BSD does not REQUIRE source distribution. It ALLOWS source distribution.

I agree with this argument, and it strikes me a bit that it's never said. In fact, the BSD License is generally included among Free licenses (even by the FSF), and while it might be Free in that it allows you modification and redistribution, it's certainly not an Open Source license. And it's NOT compatible with the GPL no matter what the FSF says.

Imagine Nvidia decided to distribute its binary drivers under the BSDL. Could you oblige them to give you the source code according to the license? Not at all. It would be perfectly legal to distribute the drivers in binary-only form under the BSD license. Ok, so now it would be legal to distribute a Linux kernel with the Nvidia drivers linked to it because they're under the BSDL which is compatible with the GPLv2? Certainly not, since the GPL requires you to distribute the source code along with the binaries.

The fact that all the code distributed under the BSDL is open source doesn't mean it MUST be open source. It just means that people who license under the BSDL are nice enough to provide you with the source code. But you can't legally force them to do so.

Imagine Nvidia decided to distribute its binary drivers under the BSDL. Could you oblige them to give you the source code according to the license? Not at all. It would be perfectly legal to distribute the drivers in binary-only form under the BSD license.

What would be the point? Sure they could, but the fact is if someone is openly distributing something under the BSD liscense they're giving the source away.

it's certainly not an Open Source license. And it's NOT compatible with the GPL no matter what the FSF says.

How is it not an open source liscense? Code that is distributed under the BSD liscense is just as easily available as GPL code, they differ in redistribution. Code doesn't have to be viral to be open source. The BSD liscense focuses more on giving away the code, while the GPL focuses on sharing the code while keeping honest people honest. I don't see that it makes one better than the other. Finally, the BSDL is most certainly compatible with the GPL and it's because of it's liscense; the GPL on the other hand is not compatible with the BSDL, which I beleive is a distinction that the FSF makes.

Edited 2007-01-16 19:13

>What would be the point? Sure they could, but the fact is if someone is openly distributing something under the BSD liscense they're giving the source away.

Yes, that's what I said in my last sentence. The best thing about the BSD License is not the license itself (which I don't judge, anyway), but the people who use it. Their attitude of contributing their code is great. It just has nothing to do with the license itself.

>Code doesn't have to be viral to be open source.

True, there can be licenses that are Open Source but not viral. They can say: If you distribute this code in binary form, you should provide the source code too. Nothing more than that. But the BSD License says nothing about providing the source code, hence my example about Nvidia drivers.

>I don't see that it makes one better than the other.

I never meant one is better than the other. I just wanted to point out something I believe is important: that the BSD license is not about Open Source and that it's not compatible with the GPL.

Finally, the BSDL is most certainly compatible with the GPL and it's because of it's liscense

The BSD License CAN be made compatible with he GPL if the source code is provided. But since this is not a must, then it's not automatically compatible, as is usually believed to be (see example in previous post).

"Anyone else noticed a lessening of linux distro review type articles lately???"

Well, that makes sense..since there has not been a new distro released.

"The BSD license is superior in every way and primarily for 2 reasons: "

Well that explains all the good commercial support and ISV's.

Well that explains all the good commercial support and ISV's.

I am sorry to wake you up. But code licensed under the BSDL and comparable licenses is used practically everywhere. From parts of opensource systems (X11, PostgreSQL), to closed-source systems (OS X, virtually everything that uses a TCP/IP stack that is not Linux or Solaris). BSDL-ed code is probably on of the best supported codebases (by license), though most users don't know they use BSDLed code.

You are probably mixing up the BSD license and the BSD operating system (for which commercial support is also widely available).

Edited 2007-01-16 16:01

"Well that explains all the good commercial support and ISV's."


I am sorry to wake you up. But code licensed under the BSD and comparable licenses is used practically everywhere.

Ahem. Think before you answer. Your answer is NO WAY contradicts what I said. If the license is always superior in all circumstances why is the ISV and commercial support for BSD systems lagging so much behind Linux. Why are all the vendors like IBM, HP, Novell, Red Hat, Dell etc working more on Linux? Obviously the license superiority is not a foregone conclusion by any means. The use of code licensed under BSD doesnt change the commercial support aspects of it.

Please don't spread FUD because of any real information. Red Hat is a Linux company, HP and IBM are using their own commercial Unices too, Novell is a no-brainer. Yahoo is using BSD, Google, Apple (MacOS X) is using and supporting BSD and so on. Most of the work environments are mixed ones, so one operating system has the hype, the other is working

>The use of code licensed under BSD doesnt change the commercial support aspects of it.

The lack of information doesn't change anything too!

"Please don't spread FUD because of any real information"

Nope. It is obvious that BSD commercial and ISV support lags severely behind Linux.

"Yahoo is using BSD, Google, Apple (MacOS X) is using and supporting BSD"

Oh come on. How many well known vendors provide commercial support for FreeBSD, OpenBSD and NetBSD. Thats what I am talking about. Not usage.

'The lack of information doesn't change anything too!"

Either understand the thing being asked or stay off the discussion.

Oh come on. How many well known vendors provide commercial support for FreeBSD, OpenBSD and NetBSD. Thats what I am talking about. Not usage.


And assuming you mean "call in for help"

http://openbsd.org/support.html
http://www.freebsd.org/commercial/consult_bycat.html
http://www.freebsd.org/commercial/hardware.html
http://www.freebsd.org/commercial/isp.html
http://www.freebsd.org/commercial/misc.html
http://www.freebsd.org/commercial/software_bycat.html
http://netbsd.org/gallery/consultants.html

You know how long those lists took to pull up/find? ABout 45 seconds...

And to return the question, what difference does it make how well known a vendor is? If you call IBM for support, or try to get them to implement a new feature, they'll politely tell you to go to hell (unless you actually use their particular product, and even then only if you make it worth their man/hours to help you.)

Either understand the thing being asked or stay off the discussion.
@manmist:
How about you stop insinuating people are stupid now you ill-informed twit.

"Yahoo is using BSD, Google, Apple (MacOS X) is using and supporting BSD and so on"

Really? So I can download iTunes or say Google Earth for NetBSD?

Yahoo is using an extremely hacked up and customized version of FreeBSD for their internal servers. Google, (correct me if I'm wrong but I thought) they were using also an extremely hacked up and customized version of _Linux_ for their stuff. Apple smacked together a kernel based on Mach with a FBSD userland, and whatever else from NeXT, etc. Nothing of this sounds like "supporting" to me.

Ahem. Think before you answer. Your answer is NO WAY contradicts what I said. If the license is always superior in all circumstances why is the ISV and commercial support for BSD systems lagging so much behind Linux. Why are all the vendors like IBM, HP, Novell, Red Hat, Dell etc working more on Linux?

"Think before you answer."
HAHAHAHAHAHA... Yes, asking why Red Hat works more on Linux is totally a good argument technique, way to go there...

(hint for the slow: Red Hat = Linux Distribution Company; this is a sure fire way of telling the poster does not really have the slightest idea what they are talking about. Watch for it folks, it's just that easy.)

But, no, really, lets look at this for a second:

Microsoft Windows Copyright:
This product includes software developed by the University of California, Berkeley and its contributors.
...[cut for brevity]...
Redistribution and use in source and binary forms are permitted provided that the above copyright notice and this paragraph are duplicated in all such forms and that any documentation, advertising materials, and other materials related to such distribution and use acknowledge that the software was developed by the University of California, Berkeley. The name of the University may not be used to endorse or promote products derived from this software without specific prior written permission. THIS SOFTWARE IS PROVIDED "AS IS" AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR A PARTICULAR PURPOSE.

AIX:
This software and documentation is based in part on the Fourth Berkeley Software Distribution under license from The Regents of the University of California.

HP-UX:
""

TCL/TK:
This software is copyrighted by the Regents of the University of California.

So, lets stop that for a moment and just get down to this: Microsoft, IBM, Sun, HP, Compaq (DEC), Novell, hell, just about every large company I can think of that writes commercial software uses BSD code in their software...

Now, maybe you can define how that is not "commercial" support? Or did you mean "publicly go out of their way to say "hey, we use [insert some name for free publicity by riding a media buzzword here] in our computers and we give back to the community!" Cause, well, if that's what you mean, it's not only a bad scale to use for commercial support, it shows nothing but inane marketing attempts. Attempts that you seem to have fallen into completely...

Companies that commercially support BSD in a limited form (in no particular order, just a short list, I don't feel like reading all the names, these are just things that caught my eye): The US Government (DARPA), Yahoo, Hudson River Trading Inc, Google, SPS Studios, Vivonet Incorporated, Linux-Consulting.com, The Socialist Party of the Netherlands, Graphical Data Capture Ltd, PARSE Software Devices, Collaborative Fusion, Inc, Sun, Natural Fit Supplements, etc, etc, ad infinitum.

Hell, I could make this list 14 pages long pretty easily... You see, the difference between Linux and BSD projects is that Free/Net/OpenBSD are not buzzwords, no CTOs perk their ears up when they hear them, nobody really knows about them outside of the technical branch of the world, so there's nothing to gain by spouting "WE SUPPORT BSD! OH YEAH! YEAH BUDDY! WE ARE TOTALLY COMMUNITY PLAYERS! BUY OUR STUFF, BECAUSE WE CARE!"

See, that's the way marketing works.

The fact that hundreds of people and organizations directly support the BSD systems through monetary donations, hardware donations, and actual coding won't come up because they can't get anything out of telling people about it. That aside, the fact that nearly every software company in the world uses BSD code is, in itself, support. They take the code, they modify it, and the world gets to use products that otherwise either wouldn't exist, would be poorer quality, or would just take years longer to develop.

That aside, a great many of the companies who have borrowed BSD code, while not wanting to put their own changes out to the public, have gone on to make monetary donations to the projects.

And I don't know about the rest of you, but I see "This product includes software developed by the University of California, Berkeley and its contributors." as "commercial support" every damn time I come across it.

Edited 2007-01-17 07:18

But code licensed under the BSDL and comparable licenses is used practically everywhere. From parts of opensource systems (X11, PostgreSQL), to closed-source systems (OS X, virtually everything that uses a TCP/IP stack that is not Linux or Solaris).

I've heard that used as an argument by various people for the BSDL, and the usage of various BSD software as a primary reason, but it means absolutely nothing. Unless those companies and projects contribute anything back to the BSDL code they are using, it doesn't help anyone in any way, and it doesn't help the health of the BSD project in question.

An example among others -> Apple. They give something back in terms of code, money, build/mirror servers and so on.

>Unless those companies and projects contribute anything back to the BSDL code they are using,

They are stealing GPL code too and you will not know most of them, because it's impossible, maybe only by accident. So in the end you have GPL, which makes the life of developers very difficult if you mix up with other licenses.

An example among others -> Apple. They give something back in terms of code, money, build/mirror servers and so on.

Which is why something like Darwin is practically dead, and no one will be able to do anything truly useful with it. The only thing that truly matters to an open source project is code.

They are stealing GPL code too and you will not know most of them

Wow. So the GPL doesn't really work properly, and people just steal the code anyway? Do you have any evidence for that?

So in the end you have GPL, which makes the life of developers very difficult if you mix up with other licenses.

The benefits quite clearly outweigh any disadvantages on that score. As explained though, everyone knows exactly what the GPL entails, but some just want to pretend that it's difficult for their own ends ;-).

They are stealing GPL code too and you will not know most of them

Wow. So the GPL doesn't really work properly, and people just steal the code anyway? Do you have any evidence for that?


Haha, sorry folks, I know it's probably still a touchy subject, but some folks seem to have the memories of goldfish.

Segedunum, if you want evidence that people just steal GPL software for the hell of it expecting not to get caught... do a google search for a little thing called "CherryOS"

Here, I'll save you the trouble: http://en.wikipedia.org/wiki/CherryOS

You see, software licenses are like the lock on your door. If someone gets a sudden urge, they'll jiggle the knob and the walk away after finding it locked... but that doesn't mean door locks work, it means most people will give up and leave... but a locksmith can still get you into your house... so, it stands to reason that someone else could just as easily.

As with anything else, it's just there to keep "honest" people "honest".

The cherry OS people just happened to be the stupid punk teenager who breaks a window with a rock rather than learning how to pick the deadbolt, but you can bet there are a crap-load of people who spend the time to learn how the lock works before trying to rob you... and those are the ones nobody catches.

Segedunum, if you want evidence that people just steal GPL software for the hell of it expecting not to get caught... do a google search for a little thing called "CherryOS"

Here, I'll save you the trouble: http://en.wikipedia.org/wiki/CherryOS

Is that all you can come up with? Really? From your opening sentence I thought you might actually have something to say, and then you pull that out.

The fact that that is about the only thing you can come up with just proves me right - the GPL simply works - and violations are dealt with as they should be.

The BSD license is superior in every way and primarily for 2 reasons:
1) Simplicity
2) Far less need to legislate to enforce

But alas, not one of the BSDs has reached the usage or critical mass that Linux has, primarily due to the license used. The GPL compels companies and people to put their code back into the kernel, which then snowballs the whole project along. There is no inequality.

An awful lot of people want to pretend that all that is religious fanaticism, but the fact is that Linus' choice for a license that would dictate that is simple, straightforward and everyone knows where they stand.

The people and companies who pretend that the GPL is complicated are those who simply want to use other peoples' code or create their own closed world for their own benefits - which incidentally, doesn't help any of the BSD projects in any way.

Edited 2007-01-16 16:50

>But alas, not one of the BSDs has reached the usage or critical mass that Linux has,


Because of,

>Why isn't BSD better known?
For a number of reasons, BSD is relatively unknown:

1. The BSD developers are often more interested in polishing their code than marketing it.
2. Much of Linux's popularity is due to factors external to the Linux projects, such as the press, and to companies formed to provide Linux services. Until recently, the open source BSDs had no such proponents.
3. BSD developers tend to be more experienced than Linux developers, and have less interest in making the system easy to use. Newcomers tend to feel more comfortable with Linux.
4. In 1992, AT&T sued BSDI, the vendor of BSD/386, alleging that the product contained AT&T-copyrighted code. The case was settled out of court in 1994, but the spectre of the litigation continues to haunt people. As recently as March 2000 an article published on the web claimed that the court case had been ``recently settled''.

One detail that the lawsuit did clarify is the naming: in the 1980s, BSD was known as ``BSD UNIX''. With the elimination of the last vestige of AT&T code from BSD, it also lost the right to the name UNIX. Thus you will see references in book titles to ``the 4.3BSD UNIX operating system'' and ``the 4.4BSD operating system''
5. There is a perception that the BSD projects are fragmented and belligerent. The Wall Street Journal recently spoke of ``balkanization'' of the BSD projects. Like the law suit, this perception bases mainly on ancient history.

http://www.lemis.com/bsdpaper.html

Because of,

Because of that fact that more code has gone into Linux on a wide variety of things than has gone into any BSD kernel.

The BSD developers are often more interested in polishing their code than marketing it.

Polishing their code for what purpose, exactly? And Linux developers don't polish their code as much? (Which is utterly meaningless by the way)

Much of Linux's popularity is due to factors external to the Linux projects, such as the press, and to companies formed to provide Linux services.

And why did Linux get more press with more companies forming around it? You simply haven't explained why that became the case.

Because when people contributed code to Linux, there was no ifs or buts about creating closed or differently licensed extensions (and gave people little incentive to do so since it wasn't possible), that code gradually improved the quality and functionality of Linux which in turn attracted companies and more developers and press attention increased as a result.

It's been a major factor in the code within the Linux kernel remaining open sourced and remaining together as a coherent whole, as well as various userspace projects as well. Without that you get fragmentation and a lack of coherance.

Until recently, the open source BSDs had no such proponents.

They still don't.

BSD developers tend to be more experienced than Linux developers

There's simply no evidence for that whatsoever, and again, it doesn't explain why more code has gone into Linux than any BSD for all kinds of hardware and purposes.

and have less interest in making the system easy to use.

I don't see any Linux system being necessarily more easy to use than any BSD system. The only reason why Linux is perhaps easier to use, and works, in some ways is because that's where most of the code goes. Guess why........

In 1992, AT&T sued BSDI, the vendor of BSD/386, alleging that the product contained AT&T-copyrighted code. The case was settled out of court in 1994

It still doesn't explain why more code has flowed into Linux.

in the 1980s, BSD was known as ``BSD UNIX''. With the elimination of the last vestige of AT&T code from BSD, it also lost the right to the name UNIX.

I don't follow. Linux isn't UNIX either. How does that mean that less is contributed to any BSD OS?

There is a perception that the BSD projects are fragmented and belligerent. The Wall Street Journal recently spoke of ``balkanization'' of the BSD projects.

And it has that perception even when it has a primary license that allows the closed source drivers and proprietary development some people believe are so important? Go figure.

The Linux community has that perception as well, and again, it still doesn't explain why more code continues to flow into Linux and projects that are written for Linux first.

There's simply no evidence for that whatsoever, and again, it doesn't explain why more code has gone into Linux than any BSD for all kinds of hardware and purposes.

Quantity is not quality. "More code" is not the same as "better system."

It still doesn't explain why more code has flowed into Linux.

It has nothing to do with the respective licenses and everything to do with the nature of the developer communities.

Linux Torvalds is a likeable guy who deliberately put together a system of participation that encouraged less experienced developers to become part of the process and thereby grow into better developers. LKML, Linux janitors, and a bunch of other initiatives have made developing for Linux attractive.

The BSD developers are less egalitarian and more interested in having a tighter focus on what they do.

Neither approach is better or worse than the other, but they lead to very different dynamics and that has nothing to do with the licensing approach.

Had the BSDI suit been settled earlier, Linus would, he says, have forked a BSD distro and gone from there. He'd have lived with the BSD license and BSD development, at least in his variant, would be just as Linux development is now.

you forgot 6:

Linux has a cute cuddly penguin mascot, and who couldn't love penguins??? BSD has mascots ranging from the devil, a bug, a flag, and a spiky fish who might be into bondage...

I don't really lean one way or another on this topic, just had to chime in though:

"2. Much of Linux's popularity is due to factors external to the Linux projects, such as the press, and to companies formed to provide Linux services. Until recently, the open source BSDs had no such proponents."

That is a horrendous argument. You are contradicting yourself if anything. Linux didn't start out as some sort of astroturfing concept. You're more or less saying Linux is successful because people payed attention to it. Must be a reason for that. (again, not leaning towards either side, but that paragraph shot down half your argument.)

But alas, not one of the BSDs has reached the usage or critical mass that Linux has, primarily due to the license used.

OS/X, which is BSD derived, has roughly the same market share as all of the Linux distros combined.

OS/X, which is BSD derived, has roughly the same market share as all of the Linux distros combined.

Which means what to anyone? It's simply not BSD, and Apple just doesn't care that they use BSD code nor do they give a damn about it. Can you run OS X applications on any BSD derived OS? Errrrr, no. No point whatsoever.

It means that arguments about licenses based on market share are silly, because the market share of Linux is tiny.


Some women just don't want to be told that they are being used. Just the way the world is I suppose.

and some enjoy the use. tastes differ.

It means that arguments about licenses based on market share are silly, because the market share of Linux is tiny.

OS X is not BSD in any way shape or form - Apple just lifted a proportion of code and modified large parts of it for themselves. Is there some part of that that wasn't clear?

Trying to somehow claim that Linux's share is small when compared with BSD because BSD code has been lifted and shoehorned into various other operating systems that will never contribute anything back is just plain laughable, but that's what many BSD proponents try to claim sadly.

rying to somehow claim that Linux's share is small when compared with BSD because BSD code has been lifted and shoehorned into various other operating systems that will never contribute anything back is just plain laughable, but that's what many BSD proponents try to claim sadly.

It's only sad from a 'linux' point of view. The whole myth about the GPL is that it encourages a huge amount of source code feedback, which keeps things improving. It's true to a certain extent, but really the vast majority of the code is produced by a core team (just like the BSDs).

The BSD point of view is that they have *better* code than anyone else, and it *should* be shoehorned into other projects, to make those projects better. Contributions back from these other OS's isn't really needed (unless it's something minor, like a fix for a device driver).

Really, it's just a difference of ideologies.

It's only sad from a 'linux' point of view.

No it's just plain sad, but if BSD people are OK with it then fine.

The whole myth about the GPL is that it encourages a huge amount of source code feedback

It's simply not a myth. The fact is that the integrity of the Linux kernel, or any GPL licensed or open sourced project, has to be maintained if it is to survive or get anywhere. That code feedback is built into the license, and ensures people and companies contribute fairly. One developer's contributions will be matched by others.

It's true to a certain extent, but really the vast majority of the code is produced by a core team

A core team that has to abide by the rules, along with countless contributions from other interested parties using Linux who share their contributions and get contributions in return from others. Those other interested parties would probably have just created closed extensions or drivers in a BSD OS, making the many people less willing to contribute as a result.

It works.

The BSD point of view is that they have *better* code than anyone else

Well, there's no evidence for that.

and it *should* be shoehorned into other projects, to make those projects better.

I've read that a thousand times on BSD mailing lists, and honestly, I just find it to be an excuse to justify the historical licensing decisions of the past.

OS X is not BSD in any way shape or form - Apple just lifted a proportion of code and modified large parts of it for themselves. Is there some part of that that wasn't clear?

That part where it matters to the discussion.

Trying to somehow claim that Linux's share is small when compared with BSD because BSD code has been lifted and shoehorned into various other operating systems that will never contribute anything back is just plain laughable, but that's what many BSD proponents try to claim sadly.

The Linux kernel is a small part of a "Linux" system. Linux systems tend to be licensed under a wide range of licenses, and only a portion of the code is licensed via GPL.

The point is that it is silly to base arguments about the effectiveness of licenses on 'market share.'

You don't seem to realize that behind your backlash against BSD, you're supporting that point.

That part where it matters to the discussion.

There is no part of what you mentioned that is relevant to any discussion. OS X simply isn't BSD, nor is it even remotely related, and trying to claim it as some market share win for BSD via OS X versus Linux as you're trying to do is just daft.

The Linux kernel is a small part of a "Linux" system. Linux systems tend to be licensed under a wide range of licenses

A large proportion of it is licensed under the GPL, and the kernel being the core of it is pretty important. Getting your hardware to run is pretty fundamental, and the GPL has been a huge part in keeping it together.

No one has the motivation to work much on a BSD kernel because they've not sure if their contributions will be matched by others, or whether others will simply attach their own binary only extensions or simply take the code and run. That underlying suspicion has held back any BSD.

The point is that it is silly to base arguments about the effectiveness of licenses on 'market share.'

You don't seem to realize that behind your backlash against BSD, you're supporting that point.

You were the one talking about Linux market share being small, so I'm not entirely sure what knots you're tying yourself into. You were also trying to take that further by implying that OS X using BSD code has some bearing on any of the BSDs market share - which it simply doesn't.

Besides, it's not a backlash against BSD. It's simply the truth. If BSD people are OK with that then there's no reason to get upset. They are simply not using a license that makes enough people confident about contributing on a level playing field in an open source project. That word source, and what gets done with it, is pretty important.

OS X is not BSD in any way shape or form - Apple just lifted a proportion of code and modified large parts of it for themselves. Is there some part of that that wasn't clear?

XNU itself is made up of a Mach-based microkernel combined with a kernel-space BSD support layer, one which bears more than a passing resemblance to LITES, as well as FreeBSD.

You can't say ‘they lifted a proportion of code’ and in the same breath say ‘is not BSD in any way shape or form’: the two are completely mutually exclusive. If it wasn't BSD in any way, it wouldn't contain any BSD code: it does contain BSD code—rather a lot of it, in fact. To say that it's categorically ‘not BSD’ is a little like saying that Solaris isn't System V, and I'm not sure anybody's about to claim that.

Edited 2007-01-16 23:45

"Just so you know BSD's are bigger then Mac OS X in market share , its market income that they have a problem with. We know how small all BSD's are vs GNU/Linux."

Your point being? You claiming that the licens is better because more people are using a super hyped up OS? Take a look at Windos then... Their EULA must be the way to go then.

you need to realize that Microsoft and Apple are using the BSD license to remove market income and market share directly from BSD

What evidence do you have of this?

Do you really think that web server admins are moving their systems from FreeBSD to OS X because they share some driver code and common utilities?

Or that network admins are trading in their OpenBSD firewalls for Microsoft ISA server, due to the fact that MS borrowed code for their Services for UNIX product?

I don't think so.

"Depends on what your goal is. If lying and destroying BSD is your goal , yes Microsoft Eula and attacking GNU/Linux is the way to go. Its sure going to bring BSD downfall. "

What i don't understand is how attacking linux will bring *bsd down.

"If you want BSD to thrive , you need to realize that Microsoft and Apple are using the BSD license to remove market income and market share directly from BSD without any compensation or contribution to it. "

I really would like to see you back that up with something.

"If as you suggest Microsoft is the way to go , then why are you not following your own advice ?"

No i was just following your 'logic'. There are more windows users then linux users and if we apply your 'logic' thats the way to go then.

Dude, it seems to me that every time I see a comment by you, your calling somebody a lier. Why is that? Is there some specific reason or is it just your limited vocabulary?

Just because someone has a different view or opinion than you does not make them a lier. Your inability to be civil about such things hint at a considerable chip on your shoulder.

And for the record, I am not a BSD user. That does not mean that I don't have a vast amount of respect for the BSDs.

Lets see :

http://www.osnews.com/usercomments.php?uid=266

I suppose you did not knew that there was an history button on comments ... you press on the name of the previous commenter and it lead you to there profile , press on comments and you can read what else they have discussed you might have missed , if you really care about what they say that is.

Yes , I believe I am limited in my insult and proper use of the English language in reply to certain deserving individual , but thats because the moderator oppose the use of many insults and have told me so.

Do I have a chip on my shoulder , yes , every time I see BSD , as to do with my past , as a BSD advocate , user , contributor , financial backer and translator.

Reality showed me I was wrong to believe BSD's lies , as to do with some of my code that I was blocked from using even do liar in the BSD camp will tell you that original cant be closed.

That being said , if your painting of me and your lack of research is any indicator , you where probably more interested in scoring some insult on me then actually discuss something intelligent.

I could pull the discussion Cloudy , from is comment history , where market share of GNU/Linux was discussed
and where it was shown to him that its not tiny and this one more then one occasion , that is why Cloudy , for me , is a liar , he discard truth and reality and repeat the same nonsense and lies that he discussed in the past.

Sorry , If my comment disturbed you.

"Yes , I believe I am limited in my insult and proper use of the English language in reply to certain deserving individual , but thats because the moderator oppose the use of many insults and have told me so. "

I think you might be just limited in your use of English in general. If you're a native speaker, please proofread your posts, they're a debacle to read.

And its such a travesty you can't insult on what is supposed to be an intelligent discussion board. You apparently can't have an intelligent discussion at all if the thread with arielb is any indication...

And yes, your comment do offend me. Most of them do. Your replies like,
""Honestly"

I know your not honest. "
really contribute to the community.

Oooooh! I love this history thing. Thanks for pointing it out.

Wow, a history function. Who would of thought!

Your French Canadian, right? So I'm not going to point out how hard it is to read your comments cause at least you try, but the personal attacks are not gaining you any support.

As for intelligent discussion, you have got to be kidding. You lash out at other peoples comment because they don't fit in to your world view. How can you consider that to be intelligent discussion?

I have neither the need nor the inclination to research your past posts and have only pointed out your seeming inability to be civil in your comments. If that's a big deal for you, or your strafing at the expletive limitation of this forum is getting you down, I suggest you p*ss off and go somewhere else!

>Your French Canadian, right? So I'm not going to point out how hard it is to read your comments cause at least you try

Attacks about the language are irrelevant in a site where at least 50% of the readers learned English as a foreign language.

Edited 2007-01-17 16:48

I am interested in what you discuss as market share , I know your wrong and are mixing market income with market share. But I want to confirm what your discussing first. If its your usual lyes and nonsense I am not interested.

I would wager he is looking at something like this: http://marketshare.hitslink.com/report.aspx?qprid=2

Windows XP 85.3%
Mac+MacIntel 5.67% (4.15 & 1.52, respectively)
Windows 2000 5%
Windows 98 1.77%
Windows ME 0.89&
Windows NT 0.68%
Linux 0.37%

Admittedly, that only shows internet usage, likely doesn't include web servers, etc, but their results do show a drop in "Linux" in the past 4 months and a great increase in MacIntel, and I'd wager most of the 0.07% that was lost was some sort of migration in that direction.

But, let's assume for a moment that web servers should count, and we'll make a stupid rash assumption and say that every single web server running apache in the world is running linux(which is, obviously, false)... well, web servers only actually make up about .5% of the worlds computers, so... at 60% (apache share) of .5% + .37% that would bring Linux up to 0.67%... which is still way below Mac...

Now, I can't say that's the way things sit right now, I'm just saying that from a short look at the numbers that's the way it appears

"their results do show a drop in "Linux" in the past 4 months and a great increase in MacIntel, and I'd wager most of the 0.07% that was lost was some sort of migration in that direction."

Why?

Having spent a long time playing with this site...because its great. Why would *if* the figures are accurate. Show that Mac is taking Linux's market share.

When if could be.
After WGA people tried linux as an alternative.
People interested in OS trying out Vista
Sites used in the figures become popular with Linux users.
People finding Linux is not for them.
etc etc.

What is clear from the *these* figures is Linux is becoming increasingly popular.

*What* is interesting, that you have left out of your post is Linux, according to this site has *massive* fluctuations in Linux figures. Particularly with massive peaks in April and August, which is where you get your *last four months from* with no real explanation, as no other OS is shown to have this massive variation.

I have looked for an explanations, including Ubuntu launches or WGA, but no single event explains such dramatic changes.

The only thing we could *ascertain* from this data is 25% increase in Linux usage every year, of users of those sites.

You are trying to deliberately mislead people and that is wrong.

Edited 2007-01-17 11:18

"OS/X, which is BSD derived, has roughly the same market share as all of the Linux distros combined."

You're confusing desktop presence with overall presence. Yes, there's more mac desktops out there than linux desktops (I'd guess). Now how about servers? Picture would appear to change there...

Anyhow, it's a moot point. As someone else said, OSX != a BSD distro.

BSDs have a large presence on servers too you know.

of course it does. however I'd think it safe to say it doesn't come close to the enterprise presence of linux at this point. if it did, you'd see more folks trying to make money off it, which largely, you don't.

> create their own closed world for their own benefits
> - which incidentally, doesn't help any of the BSD
> projects in any way.

I'm a GPL/LGPL supporter. It follows a quid quo pro idea (I scratch your back if you scratch mine) that ensures that you won't be "ripped off". It's the license set I'd use if I were to write software for open source.

But you're forgetting one key advantage of the BSD license -- sometimes you have code that you want to have as broadly used as possible so *want* to be ripped off (as long as you get the credit). For instance, think of the Unix standard and TCP/IP. Both succeeded precisely because they were BSD-licensed. TCP/IP in particular isn't the best protocol out there, but since it could be "ripped off" it could be used directly in everything from routers to MS-Windows to Mainframes. If the TCP/IP stack had been LGPLed, it would have been a less popular (in particular, MS-Windows would likely push it's own protocols the way they've pushed active directory).

Before anyone mentions active directory as a example of why the BSD license fails, consider this. I the Kerberos code and protocol wasn't available in BSD form, Microsoft would have no reason to stick to an existing protocol and would have developed something a lot more complex that would require a lot more reverse engineering to implement.

It helped adoption of a single protocol used by the majority of the computers around the world, sooner rather than later. It DID help and you're blind to think otherwise.

It helped adoption of a single protocol used by the majority of the computers around the world, sooner rather than later.

A protocol Microsoft didn't really want to adopt, and they needed something - fast ;-).

It DID help and you're blind to think otherwise.

From the point of view of a BSD licensed project, helping Microsoft was a rather pointless exercise.

Uh. You didn't even address what I said. It helped, whether BSD cared or Microsoft wanted it to. It did. The end.

Not just MS, but everyone had access to BSD code to use in their BSD stacks, that's why TCPIP became the standard for the Internet. it was used by everybody

Firefox is the most popular open source app by far, probably one of the most essential for other OS's because of its website compatibility and is not GPL. It's tri-licensed with gpl but they won't accept any gpl code because of the MPL.

Apache is also extremely important to linux and linux wouldn't be the "server" OS without it. Linux wouldn't go anywhere without X, either.

Interesting enough arielb, I can take X or Apache and make it much better, and re-release it under the GPL and there is nothing these projects can do about it. You can licence BSD code under GPL but not GPL code under BSD. GPL is a popular licence for many reasons, but it is also a much "stronger" open source licence then BSD.

I perfer the GPL because I think it respects my copyright as a developer more then the BSD does. I wouldn't like it if a company like Microsoft or Apple was to my code in their closed applications without compensation. The GPL allows me to licence my code and make money in this way, while the BSD does not.

Edited 2007-01-16 20:05

I perfer the GPL because I think it respects my copyright as a developer more then the BSD does. I wouldn't like it if a company like Microsoft or Apple was to my code in their closed applications without compensation. The GPL allows me to licence my code and make money in this way, while the BSD does not.

Sure, that's your preference.
However, I would be thrilled if code that *I gave away* was included in a commercial product that was to be used by possibly millions of people. Talk about bragging rights.

Per tell openwookie, what things have you released under the BSD licence?

Per tell openwookie, what things have you released under the BSD licence?

Nothing yet. I develop web apps, so its rare that I get to write anything that would be beneficial to release.

I have a couple of small ruby libraries that I could release, I just have no time to polish them up (documentation, etc). When I do release them, they will be released under the ruby licence (which is similar to BSD).

Linus doesn't get any compensation from Novell or Redhat from using his code. Yes, the code is out there but only the companies that didn't really write the code will actually make money from it. That's because you can't make money from just selling open source software. You can only make money from support and that only benefits big companies!

Also with bsd you can still make an app or OS that's open source and still have the option of making another version with proprietary features such as codecs and fonts.

>Linus doesn't get any compensation from Novell or >Redhat from using his code.

Well Linus is employed by OSDL. Both Novell and Red Hat are members of OSDL.

"Red Hat and VA Linux, both leading developers of Linux-based software, presented Torvalds with stock options in gratitude for his creation. In 1999, both companies went public and Torvalds' net worth shot up to roughly $20 million [1]."
http://en.wikipedia.org/wiki/Linus_Torvalds

So I think Red Hat gave something back? How much money did *BSD developers get from Yahoo,Google,Apple or Microsoft?

Edited 2007-01-16 22:22

Linus doesn't get any compensation from Novell or Redhat from using his code.


From wikipedia (and "Just for fun" by Linus himself):

"Red Hat and VA Linux, both leading developers of Linux-based software, presented Torvalds with stock options in gratitude for his creation. In 1999, both companies went public and Torvalds' net worth shot up to roughly $20 million"

Gratitude isn't the same as legally required compensation. Fortunately for Linus, there was a lot of dotcom hype and he sold before the stocks busted. Most gpl developers can't count on being compensated at all.

Is Linus getting anything from these deals:
http://open.itworld.com/4917/061220mslinux/page_1.html

No, but Microsoft can actually make money from linux. It's really hard to beat microsoft when it comes to support for big companies. They could try to knock off redhat and then the only way anyone could make money from linux is by working with microsoft.

"The GPL compels companies and people to put their code back into the kernel"

As long as you don't distribute your changes you are not compelled to do anything.

"Ask a lawyer and you can't trust yourself anymore "

Ask two lawyers and you will get three different opinions. :-)

As far as I read the article, the arguments were absolutely valid.

Throwing shit back without being responsive to the points is a very bad style of argumentation, FYI.

While following the IBM/SCO/Linux case, I had always looked at Groklaw as some sort of legal authority on the case.

The problem is that many other people did, as well. PJ and her crew are free software activists, with a clear bias that influences their "reporting". They've never tried to hide this, but many people seem to blissfully ignore that and treat Groklaw as some sort of authority.

The SCO case was something the entire community was united in, so that bias was easy to overlook because we all pretty much shared the same opinion anyways. To her credit, PJ did a remarkable job of combing reams of legal documentation to bring the core issues to light, much to SCO's chagrin. But there's no denying the reporting was decidedly biased against SCO, as SCO pointed out repeatedly and even attempted to initiate a pro-SCO information site. However, none of us care(d) because, well, it's SCO.

SCO has quieted down, so now the Groklaw community is stirring whatever other pots they can find. That is certainly their right, and all the more power to them. But at the end of the day, Groklaw is simply an uber-blog espousing personal opinions with facts and documentation selectively chosen to back them. There's is not even the slightest attempt at balanced reporting or seeking counter-opinions.

The breaking point for me was the scam about Novell forking OpenOffice. That was over the top, even for them. And they were called on it by many others, which hopefully validated the fact that Groklaw is an opinion site, a very well written one at that, but not an authoritative resource.

I think the "BSD" and "GPL" camps need to see that they have one common enemy, Microsoft, instead of fighting each other how about attack Microsoft instead?

Not everyone considers MS to be 'the enemy'. That tends to be a 'linux' point of view.

"Linux people do what they do because they hate Microsoft. We do what we do because we love Unix"
-- De Raadt

Excellent quote.

I generally applaud anyone striving to become better than him- or herself rather than someone striving to become better than someone else.

The aim for any FLOSS project should not to beat Microsoft, Apple, Sun etc. It should be to beat itself.

Catchy quote, but that doesn't make it true. It sometimes seems the camp I've heard spew the most vitriol about another "competing" OS and its users is the BSD camp against Linux.

Jealousy perhaps?