Another Stolen ElectionFollow

And sniper rifles.

And... wolves. Or something that can smell fear/treachery.

BSD is not Unix. It's one variant of unix. The guys at Berkeley decided to build a kernel and provide it for "free" to people who wanted to use it.

You could argue that BSD was "open source", but not that the "original unix kernel" was open source (which is what you claimed). I mentioned BSD as being a more open variant of Unix a couple times before you went to wiki and figured out that what I was saying was true.

Let's also be clear. There are different degrees of open source, and a number of different organizations who advocate for "open" software development, all of which use different methodologies, disagree over exactly what is required for software to be "open", and fight over who's right all the time. The idea that there's one happy group of people out there who are all in agreement about sharing ideas is a ludicrous myth. Spend a few weeks perusing the various linux dev boards and you'll quickly realize that there are more eqos at work than are helpful to any real project. Everyone knows how to do things, and none of them agree... It's comical really.


BSD specifically only made certain parts of it's operating system open, and it's different than Linux. Let me just toss two different ideas at you. BSD is of the "non-profit" variety of open software. So you are free to copy anything under that license. You're free to modify it. You're free to share your modifications. But you can't profit on them. At all. Nice, but not surprisingly, it didn't take off that much. Hobbyists played around with BSD, but you never saw anyone use it for any actually important application. For exactly the reason that there's not true "vendor" for BSD. Thus, you can't get support (or couldn't).


Linux is a different variety of "open" software. The kernel is free to obtain and modify. Similarly to BSD, there are some folks who manage the "official" kernel trees and patches, but there's nothing stopping anyone else from making their own. What's different is that the Linux kernel is not restricted to non-profit use. I can take a copy of the Linux kernel, build a distribution around it, put it in a box and sell it. I can build a company dedicated to managing and maintaining *my* version of Linux, and provide support to customers who purchase it.


Oddly, that's what was successful about Linux. The profit motive. And guess what? Redhat is just as proprietary about all the distribution code they put out there as Sun Microsystems is. The kernel and a few common utilities are open source, and Redhat does allow collaboration on their code base (if you're from a respectable company or have a name in the field), but since they are responsible for supporting their customers, they have final say as to what is "Redhat Linux". Same deal with Suse and a bunch of other variants.


The point is that you can't just toss around labels like "open source" as though they have some consistent magical meaning in all cases. The reality is vastly convoluted and incredibly inconsistent. The term means different things to different groups of people. However, it's absolutely false to say that the original "unix kernel" (whatever you think that is) was "open source". It wasn't. It's just that the original design of unix was so simplistic and revolutionary at the time, that a zillion people all copied the basic idea. They wrote their own kernels because they were operating on different hardware. Maybe your problem is that you don't really understand what a unix kernel is. It's not the actual operating system. It's just a binary file that defines devices. Today, that can be a ton of things, from hard drives, to memory locations, to bus addresses, to printers, to special file devices to run software or other specialty hardware. But back in the day, they were pretty simple. You had a computer chip. You had some memory. You had an input device and an output device. If you were lucky you had a storage device. A kernel just defines those things in a way that the rest of the OS can use.


It would take a long time for me to explain in detail the theory and design of unix. Suffice it to say that it's about simplifying things to their most basic elements and handling those elements in a consistent manner. That's what the kernel does. The OS itself runs on top of the kernel and does stuff with the elements that have been defined. Most people don't "get" unix in any sort of reasonable time frame. Certainly, I likely can't explain it to you in this forum. Most unix admins take 2-3 years and several training classes before the lights come on and they get unix. Some never do. But if you don't get it, you wont understand what I'm talking about when I say that the fact that unix kernels were widely constructed on a number of platforms and with a number of different unix OSes does *not* mean that the kernels were open source.


It's hard to make a good analogy, but it's kinda like arguing that because you wrote a paper using Word after seeing me write a paper using Word, that my paper was open source and your's is based on mine. Nope. They just both use a similar process for writing (the editor in this case). Unix was revolutionary because of the concept of defining everything as a "device" (technically as a file). This meant that the OS didn't have to understand anything about what it was sending or receiving data from/to. It just had to know to send it or what to do with it. The unix kernel concept, once invented was as simplistic and obvious an idea as inventing the wheel. You can make a lot of very different things with it.

Wow, I don't disagree with anything you posted there. Although there is a debate going on right now over the meaning of open source in the IT community, I happen to agree with your explanation.

The original Unix kernel would not fit the technical definition of "open source software," but when it is released to everyone in the community free of charge (although not "officially") then it accomplishes the same purpose because people later created the free BSD Unix OSs.

That is my point, that everyone had access to the original Unix kernel free of charge and could use it so it spawned later generations of free, open source, software packages including BSD Unix which was later incorporated in Linux, although Linus said that his main inspiration was actually the proprietary Sun version of Unix.

So, yeah, good history lesson regarding the technical definition of open source, and we agree on paper trails for electronic machines too. I'm amazed.

Wrong. They basically copied the original Unix code and then re-wrote it because they wanted to get it away from the copyrights held by AT&T. One reason the courts favored them is because AT&T had released the Unix kernel code to the public domain by giving it to the faculty and staff of universities and publishing it in textbooks. This was discussed in Unix System Laboratories vs. Berkley in 1992.

http://en.wikipedia.org/wiki/USL_v._BSDi







You just contradicted yourself. You said it didn't did that at all, and then mentioned BSD as the exception.



Again, you just contradicted yourself. The open source components are what allow people to make their "own" distributions, different from the open source publicly available distributions.




I agree completely. I've never presented any conspiracy theory simply empirical evidence. I do think suspicious past elections should be investigated, however, as long as it is not cost prohibitive.

The fact that they had to do this should be your first clue that the original code wasn't "open source".

Look. It's a silly argument. The fact that there is a 25 year history of lawsuits over who owns which code and who doesn't pretty much proves my point. The code was not released as "open source". There's a specific range of meanings for that phrase, none of which matches the historical path that various unix branches took developmentally. BSD did some things with its kernel which today we might call open source, but the code they used to develop their kernel in the first place most definitely was not.


Part of the problem is that there were questions as to whether computer code was protected under patent or copyright or something else entirely. It took the law most of that 25 year time period to iron out the issue. So yeah. It's muddled. But coming in after the fact and declaring the original unix kernel "open source" is just plain absurd.

Sure. Except my point was that the statement that the "original unix kernel was open source" is simply incorrect. Now, we can go through an entire history of variants of unix if we wish, discussing how different kernels were developed for different hardware platforms, with each generation changing how they worked and how they licensed their code, eventually leading to some branches which became what we now call "open source", but the argument I was responding to was that unix systems were more secure because they were "open source", which is not only incorrect, but also irrelevant *and* absurd.


The statement about the "original unix kernel" is just factually incorrect. Period. But the argument he was attempting to support is also incorrect. The most secure computing systems are the opposite of open source. Most of the best mechanisms for securing data and systems were developed under licensing (or are wholly proprietary or just plain secret), not as open source projects. The original assertion is just plain wrong on every single level...

Edited, Nov 3rd 2008 4:02pm by gbaji

Not me confusing it though. The phrase "open source" was used to refer to Linux kernels and "the original unix kernel" within the same context. Now, perhaps he was context switching and didn't bother to tell anyone, but that's *exactly* the issue I was bringing up. No early kernels can be said to have been "open source". For one, the term and concept didn't exist, and for another, they were *not* granted the same sort of open licensing rights as the term means today.


The most correct term to describe the original AT&T kernel (really the source for compiling unix kernels) would be (as you mentioned) "licensed source". Licensed for super cheap is still licensed. That the code got out "into the wild" of academia and they eventually lost their control over it doesn't mean it's correct to label their source code "open" by any stretch of the imagination. That label implies that the code was released in an open format for anyone to use as they wish, which was absolutely not the case.

You also have to remember that the laws didn't catch up with regards to computer code until the late 80s to early 90s. While we can look back from today and say that they lost their legal rights because they didn't follow the correct steps to protect them, it was well after the "you've lost it" time period before the law was clear as to what would have been needed. Again though, it's pretty silly to go back in time and label that "open source" because they didn't protect their code using legal processes that weren't formalized for another 20-30 years.



Um... It's pretty irrelevant anyway. The entire example is just "odd". The kernel source code really doesn't have anything to do with how secure a computer system is. That falls into the arena of port daemons and inet processes, which were (again) almost exclusively developed by private companies using largely proprietary code. The point is that the use of open source code for secure applications is only a very very recent thing. And most of that wasn't driven by a need to create secure systems, but to bring security to open systems (which were open for cost reasons, not security reasons).

A great example of this is to follow the history of the development of ssh. It started as a simple freeware product, but had some serious flaws. The adoption of the software by companies making proprietary variants added significant security and closed the serious loopholes. When open systems wanted to use ssh, they went back to the less secure free version to start, taking several years to catch up to where the proprietary versions were.


My point is that "open source" doesn't really mean "more secure" in any real way. While it can be secure, it's not secure because it's open. And when we're talking about application security (like a voting system), open source is almost always the wrong way to go.