Home Messages Index
[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index

Re: GNU and Linux Both Required to Make an "Operating System"

Hash: SHA1

____/ Rex Ballard on Wednesday 07 Apr 2010 22:28 : \____

> On Apr 7, 3:59 pm, Roy Schestowitz <newsgro...@xxxxxxxxxxxxxxx> wrote:
>> Linux is not an operating system, or is it?
> The Linux Kernel is one component of an operating system.  But GNU is
> also a critical part.  The GNU glibc library converts C framed
> procedure calls into operating system vector calls.  Because of glibc,
> the same C source code can be compiled for Linux, AIX, Solaris, HP_UX,
> and OS/X, and can be compiled without modification to run on any of
> these platforms.  Many developers write for LInux because the Linux
> code is so portable with so little effort.  Using the cygwin library,
> you can even compile source code written for Linux, to run on Windows.
> From that point on, it's layers of libraries and servers.  There are
> interface libraries for C++, PERL, Python, Ruby, and Java JVM which
> all interface to glibc.  The kernel can also call the drivers.  Then
> there is the X11 display, which is a strange combination of hardware
> drivers coupled to a dedicated server process.
> Then you have the widgets and toolkits, Athena, GTK, Qt, and Motif
> work-alikes.
> Then you have application frameworks like Apache, browsers, Java, and
> Eclipse.  These provide the fundamental properties, with the ability
> to call plug-ins which are carefully downloaded and installed by the
> user.
>> ,----[ Quote ]
>> | If you are at all familiar with my brain
>> | sneezes (I could use another word here but
>> | this is a respectable blog :) then you know
>> | I am always rabbiting on about Linux. I am
>> | always Linux this and Linux that. This gets
>> | some people upset and they claim that it
>> | should be GNU/Linux or perhaps more
>> | accurately Linux/GNU.
> Actually in terms of the layers, it's GNU/Linux.  Linux is the kernel
> and calls various device driver modules.
> GNU is "over" the Linux kernel, with it's interface library.  Then
> there are GNU applications and then GTK and then Gnome.
>> | There is a word in
>> | the English language (which I hope I am
>> | using) called context.
>> | In other words, I
>> | use the word Linux in the context of a
>> | complete Linux based distribution and
>> | believe that my readers understand that
>> | context.
> What most people call "Linux" is actually a "Linux Distribution".  A
> typical Linux distribution includes the Linux kernel, the glibc
> library, a few hundred GNU and OSS libraries, and and a few thousand
> applications and utilities.  These applications range from simple
> shell scripts and command line utilities to entire systems capable of
> emulating entire computers or computer systems.  In some cases, a
> single command translates to dozens or even hundreds of additional
> commands and capabilities.  And applications like Java applications
> can add hundreds of additional features and functions to a system.
> So a Linux distribution only "Linux" in that it's based on a Linux
> kernel.  There are other kernels such as FreeBSD, NetBSD, and
> OpenBSD.  Much like Linux, these have common libraries and structures,
> and can also run the Linux code.  Where there was a choice between BSD
> code which can be "kidnapped an enhanced" then sold or made part of a
> proprietary system like Windows or OS/X,  and GNU or OSS code which
> requires that any enhancements be shared with the rest of the
> community as patches which may or may not be incorporated into the
> master product, Linux has opted for the latter, while the BSD and Unix
> derivatives have opted for the former.
> Many, including people like Bill Joy, Marc Andreeson, and thousands of
> others got sick of their code showing up in AT&T and Microsoft systems
> and applications, without even getting so much as billing on the
> development team or a royalty check.  In many cases, these companies
> didn't even ASK for permission - because the BSD license didn't
> require them to.
> http://en.wikipedia.org/wiki/Berkeley_sockets
>> http://it.toolbox.com/blogs/locutus/linux-is-not-an-operating-system-...
>> Related:
>> Don't wanna pay for Linux? Beware of software patents
>> | Do you really want to wake up one morning and pay Microsoft or SUN
>> | royalties to have the right to use Linux on each of your boxes? Probably
>> | not. However, if companies continue playing the little game of open source 
>> | vs. software patenting, that time might come much sooner than you think.
> Actually, Open Source is one of the best defenses against software
> patents, especially for individuals and organizations that can't
> afford to spend thousands of dollars on patent searches to prove that
> they have disclosed all prior art.
> When the patent office grants a patent for a software "device", the
> lawyers and patent office only check prior patent applications and
> patents to determine whether the device has already been patented.
> However, patent law requires that both the device and the claims for
> that device be original, and that they could not be intuitively
> derived by someone knowledgeable in the technology.  These two
> requirements make it almost impossible to patent anything in Open
> Source.
> The first problem is that if the Open Source device, or the claim for
> logically similar code, has been published prior to the time that the
> patent application was received, the patent can be nullified, even if
> it had been granted.  The moment a patent holder tries to enforce the
> patent, they run the risk of some OSS software with a prior
> publication date being introduced during discovery.  In the worse case
> scenario, the patent holder has tried to sue the prior publisher, and
> the prior publisher could end up being granted the patent, giving him
> the power to get an injunction against the previous holder - because
> he filed a fraudulent patent application.
> If the Open Source code come out after the initial patent is granted,
> the patent holder must prove that the open source code was a direct
> result of knowledge of the patented device AND the specific
> implementation of that device.  So if joe college comes up with code
> that lets you send stock quotes over a unix pipe, and those quotes are
> encapsulated in XML, you would have to prove that his implementation
> was deliberately copied from yours, AND that his attempt to use the
> tool in the manner made in your patent claim.
> An even bigger problem is that many of these Open Source application
> were created as classroom assignments, often to freshman and sophomore
> programming students.  If it turns out that one student actually DID
> go to the patent office archive, dug up the patented code, and even
> cited the patent in his completed homework, there might be 10-20 other
> fully functional implementations available, at similar costs.
>> | Take Microsoft for instance, they have a great code repository called
>> | CodePlex, one of their show-rooms for open source development advocacy.
> The problem with CodePlex is that you have to sign a consent agreement
> which states that you will only use the code for development and
> support of Windows applications.  Essentially, if you agree to the
> license, you forfeit your ability to create any kind of similar code
> for Linux.  Furthermore, if you take the code from CodePlex and try to
> contribute it to Linux, you will have to provide a certificate of
> authenticity in which you state that the code is your original code
> and not a derivative of any other copyrighted or patented work.  This
> means that if you lie, you are not only guilty of patent violation,
> but you are also guilty of fraud.  You will go to jail.
> When someone asserts a patent claim against Linux, or OSS, the
> response is very simple.  If you tell them which code you think
> violates your patent, they will replace your "device" with a different
> implementation.  This was a source of great frustration for SCO,
> because the Linux community kept telling them - show us the code you
> think we stole, and we'll stop using it.  The problem was that SCO
> didn't want Linux to stop using the code, they wanted IBM to pay SCO
> $3 billion for using the code.  The same is true with Microsoft.
> The bigger problem for both SCO and Microsoft, is that when they
> finally DID disclose what they assumed were patent infringements, the
> Linux community not only provided the certificate of authenticity, but
> also the revision history showing that their implementation was
> actually prior art.
> The really ticklish problem for a company like Microsoft, is that if
> they assert a patent claim against GNU code, and the GNU code turns
> out to be the prior art, Richard Stallman could demand that the ENTIRE
> Microsoft application be placed in Open Source - because the Microsoft
> code violated the GNU copyrights.
>> | Except that in the last months, developers have lost their ability to
>> | license their code under the GPLv3. As we've shown earlier, Microsoft donât
>> | like the GPLv3 because it provides protection against software patenting.
> The irony is that GPLv3 doesn't even prevent you from enforcing
> patents against proprietary versions you have licensed to other
> companies.  It does however, require that you agree NOT to donate the
> code, have it enhanced by hundreds of contributors who donate their
> time for testing, enhancements, bug fixes, promotion, distribution,
> and incorporation into other projects - then attempt - 5-10 years
> later, to demand $3 billion dollars in patent royalties from one or
> more companies with deep pockets - because you changed your mind about
> enforcing your patents.
>> | The latest victim to date is MySQL who has always had a very strong stance
>> | against software patent.  But like Icarus flying too closely from the sun,
>> | MySQL burned its wings: their anti-patent manifesto has abruptly
>> | disappeared from their site. (The complete story here).
> Sun bought MySQL support organization.
> Oracle is in the process of buying Sun.
> It's possible that Oracle will be offering new tools for MySQL, such
> as CASE Method, or incorporating it into some of their products, like
> Fusion, for things like "Community" or "Express" editions of the
> product.  MySQL wouldn't be able to continue to tirade against
> software patents if it's now becoming a key component connected to
> other components that are patented.  Especially if there are big bucks
> available to the MySQL team as part of the deal.
> IIRC, there have been numerous attempts to buy off Richard Stallman,
> but he has been loyal to the thousands of people who have designed,
> coded, tested, fixed, enhanced, and promoted GNU software.
> FSF has filed a few lawsuits to enforce GPL, and most companies get a
> rude awakening from the Judge during the preliminary ruling and then
> are relieved to find that the GNU terms are VERY reasonable.
>> http://blog.milkingthegnu.org/2008/03/when-the-ip-law.html

I once read that Torvalds was also offered huge sums of money. That's why OSDL was made.

- -- 
		~~ Best of wishes

Previously-unsurpassed exposure makes carnation-faced men
http://Schestowitz.com  |    RHAT Linux     |     PGP-Key: 0x74572E8E
 02:20:01 up 142 days, 22:51,  2 users,  load average: 0.16, 0.22, 0.27
      http://iuron.com - Open Source knowledge engine project
Version: GnuPG v1.4.9 (GNU/Linux)


[Date Prev][Date Next][Thread Prev][Thread Next]
Author IndexDate IndexThread Index