Reasons Why You Should Not Use FreeBSD

[Kano]

I dont think that you ever heard about freebsd jails, that should be enough for some types of virtualisation. Also it is partly possible to run linux drivers in userspace. Its not possible in kernel mode due to licencing but you can for example run linux usb drivers for dvb devices from userspace, so basically you can drive similar hardware up to a certain point. You can also execute linux binaries so freebsd can run to a certain degree the same apps/games. I don't get why ppl try to make one system so bad, usually there is always a usecase where a system shines.

[Libreman]

Permissive licenses have "5 freedoms":

• The freedom to run the program, for any purpose (freedom 0).
• The freedom to study how the program works, and change it so it does your computing as you wish (freedom 1). Access to the source code is NOT a precondition for this.
• The freedom to redistribute copies so you can help your neighbor (freedom 2).
• The freedom to distribute copies of your modified versions to others (freedom 3). By doing this you can give the whole community a chance to benefit from your changes. Access to the source code is NOT a precondition for this.
• The freedom to distribute copies of your modified versions to others, under different conditions (freedom 4).

Copyleft licenses are more free from software point of view. Permissive licenses are more free from user's/dev's point of view.

So you actually are claiming that an ability to take away the 4 former freedoms is a "freedom" in itself? That an ability to restrict others is in fact a "freedom"? What kind of an Orwelian corner of the universe I've wandered into - if this is not doublespeak I do not know what is ...

I truly would like to know whether people making this point are just knowingly trolling - bringing up something they know is not actually true to avoid conceding that "permissive" license just doesn't protect those freedoms the GPL does or they truly believe in that kind of logically absurd proposition. If you're struggling with the realization of just how absurd that is, imagine that someone would propose additional amendment to the US constitution which would go something like this:

Last amendment) Oh and by the way you can disregard all the aforementioned amendments if you choose to do that

And then they would argue that such a constitution is obviously more free that the original as it allows you to do anything and abolishes all the nasty restrictions that were there previously. It would be a "permissive" constitution.

Everybody would look at anybody who would propose that as a loon and rightly so ... and that's also how I see the above mentioned "5th freedom" as you called it, it nullifies all the freedoms you had previously - from the standpoint of protecting software freedom it is just as absurd as the above amendment for the constitution.

The only way out of this is to just admit that you really do not care about those freedoms being preserved as you or the people advocating for "permissive" licenses demonstrably (according to the rules of logic) don't, otherwise you would use GPL that protects them - but saying that you do not care about preserving those freedoms doesn't look good so you probably won't do that and keep pretending (or fooling yourself) that you do care about those freedoms and continue to present an ability to take away freedom as a "freedom" in itself (and I'm sure the totalitarian regimes around the world would agree to that proposition)

I am just waiting for one thing to happen. For GCC folks to rip off the "wonderful" LLVM + CLang compiler code and adapt it in GPL3+ GCC suite. That would be so ironic... Hehehe.
I can't stop giggling at thought of it.

You see this is a nice demonstration of why what I said above seems as a proper description of the situation. LightBit added another freedom as a "The freedom to distribute copies of your modified versions to others, under different conditions (freedom 4)."

Relicensing LLVM/Clang under GPLv3 is exactly that and people should be happy that it gets exercised (why would they include it otherwise), yet you refer to it as "ripping it off" and we all know that is without doubt as many in the "permissive" camp would actually feel like. But why? Because the whole so called "permissive" charade is about making things "corporate friendly" - which means eliminating the 4 freedoms so they can monopolize their contribution and take away control from the users - it has nothing to do with software freedom. If you relicense to protect those freedoms they're not going to like you very much - funny that, isn't it

Why do you think LLVM is sponsored by Apple, I mean APPLE for god sakes! Could it be more transparent? The "permissive" licensing is the corporation's attempt at "damage control", they see that they can not compete with open-source in the long term so at least they try to make it so they can rip it off (and this is true rip off because they get to close it down) and eliminate the freedoms that made it so successful.

Does anybody really need more sensible reason why to prefer GPLv3 or copyleft in general?

[blacknova]

So you actually are claiming that an ability to take away the 4 former freedoms is a "freedom" in itself? That an ability to restrict others is in fact a "freedom"? What kind of an Orwelian corner of the universe I've wandered into - if this is not doublespeak I do not know what is ...

How person's ability to not disclose his derivative work affect original product? The person in question cannot close or restrict distribution of original product, he can only do that to his own version or derivative work.

[susikala]

Phoronix, I honestly don't get why you always have to go with such a warmongering attitude. I get it that being a douche on the Internet=money, but is that all you care about? Setting the BSD users here against the Linux ones? The GPL supporters against the BSD ones? The KDE against the Gnome users?

Wouldn't it better to realise that both Linux and the BSDs rightfully exist (otherwise they would have already disappeared long ago) and that each one of them has its own use or purpose?

I think it would be much better to a keep to a neutral attitude, as is normally expected of journalists.

[Libreman]

How person's ability to not disclose his derivative work affect original product? The person in question cannot close or restrict distribution of original product, he can only do that to his own version or derivative work.

This has no relevance to the fact that by doing that he takes away those freedoms from the users that use it. And those who licensed that sw under "permissive" license are responsible for that - if they licensed it under GPL it wouldn't happen and users still would have all those freedoms with the derivative also, so in order to advocate for "permissive" license you need not to care about that. You need to not care that those freedoms get taken away. I do, that's why I advocate for and use GPL as many other people do for similar reasons.

People who advocate for "permissive" license don't care about those freedoms being taken away in this way ... and that would be ok if they would just be honest about it. Unfortunately most avoid conceding that point and keep pretending that's not the case because they know people care about it and if they were frank they would turn to GPL. By not being honest (either with others or often even with themselves) there is better chance some get confused into supporting it not realizing the above.

Also, it has very bad effect on innovation, if someone takes the code, adds his own improvements (and other stuff) and doesn't contribute it back (which is almost always the case if it's commercial venture) then the original gets left in the dust and the entire community loses an opportunity to improve it even further or add those improvements to existing systems - stagnation is the result, as is happening to BSD, it will die eventually. And UNIX commercial offshoots like OSX live on not contributing anything back. That is how "permissive" works, namely it doesn't - even if you overlook the obvious and very important problems about user freedom.

And another point, you said that "he can only do that to his own version or derivative work" but considering that in excess of 99% of the code was most likely written by others, is it really accurate calling it "his own version"? The buck really stops with those who've written it in the first place - they are responsible for the loss of freedom if they allow it. If someone doesn't like that the software was written with preserving user freedom in mind, they are free to write their own with different terms. So it really is about the original author and whether he/they want to preserve freedom or not ... a choice of "permissive" license says clear and loud that he/they don't.

[blacknova]

This has no relevance to the fact that by doing that he takes away those freedoms from the users that use it.

How so? Users can continue to use original product and have access to it's source code without any restriction. How come they've lost their freedoms?

[0xBADCODE]

I dont think that you ever heard about freebsd jails, that should be enough for some types of virtualisation.

Jails are just a mere "containers" rather than hypervisor, so in fact there is almost no virtualisation happens. There are 2 inherent disadvantages of this approach.

• First, using this technology, you can't boot OS with a different kernel, even if needed. It's just kind of chroot.
• Second, it relies on quite thin layer to isolate things. So it's more hackable than, say, Xen (all kernel bugs vs small hypervisor code bugs).

On advantageous side, overhead from "containers" is minimal so it's close to bare metal speed. This applies not just to jails but to all similar kinds of containers as well. But you see, for hosters, OpenVZ overtook market from *BSD a while ago and become a "de facto standard" for them (slowly moving to full hypervisors as hardware becomes more powerful and supports virtualisation in hardware). OpenVZ comes with some issues (custom patched kernel of some lower version than usually) but it works. And it offers far better manageabily and many features needed to run in hosting environments. It could even perform live migration, transparently moving certain container to another host without stopping it. If things done correctly, nobody would even notice that container has been moved. Because service is not interrupted at all, up to keeping TCP connections alive! FBSD guys are still ages behind in their technology. For more simple uses, LXC+Cgroups would still offer more features than jails, most notably, resources scheduling and limiting and so on. OpenVZ currently builds on top of LXC+Cgroups boilerplate and extends it many ways, adding features missing in mainline kernels so far. It's laughable to see this. I can remember times linux fanboys had no words to counter *bsd fanboys. But it's 2012 and to me it looks like if their bet has been put on wrong horse.

Also it is partly possible to run linux drivers in userspace.

And why someone on the Earth would need this piece of headache? You see, Linux comes with source. It costs nothing (if you don't want to pay for support, etc). So you can just use it and have fun and profit. Rather than fight with those (unnatural) technical difficulties. It's not my fault someone incapable of writing drivers under "proper" license in a timely manner. In fact I partially attribute this failure to BSD license. You see, there are/were some vendors who took *BSD code pieces but "forgot" to return anything in exchange. That's one of reasons why BSD has been outrun by Linux where you should show your kernel code. As the result of this approach, Linux currently supports ton of hardware platforms, devices, etc. Including those you can actually buy. In fact many vendors supply it in their devices by default. Many SDKs for evalboards assume Linux as well, etc. And all this comes with source. When license offers option not to share source, it's tempting to do so. However this backstabs project as whole, slowing down it's development.

Its not possible in kernel mode due to licencing

That's where BSD politics comes in game. They care about padding interests of proprietary corporations and their abilities to close source in proprietary projects far more than about what would happen to their users. As for me it looks like if their licensing is their worst curse ever. If you allow in your laws to do some bad things, bad things would happen. Ironically, bad things happened to BSD guys themselves. Proprietary vendors took sources here and there and most of time they "forgot" to help their upstream in development. Just because it has been allowed by license. Some of those even cowardly switched to Linux ... when they got idea nobody would buy their "super-puper-proprietary-stuff" when there is opensource Linux, available free of charge.

but you can for example run linux usb drivers for dvb devices from userspace, so basically you can drive similar hardware up to a certain point.

Except that on Linux you just install OS and it works but in FBSD you should start heroic fight with dozen of technical difficulties instead. Half of them also comes from politics. At the end of day I need things running and I don't need extra headaches.

You can also execute linux binaries so freebsd can run to a certain degree the same apps/games.

But again, in Linux it would work out of the box. And you see, these guys dont even emulate all clone() flags in their stupid linuxlator. Because this would require to implement half of LXC in FBSD. This implies that Chromium (and other apps using it) would be unable to isolate themselves into LXC containers, etc. It's 2.6.24 kernel featiure, ages ago. Now it's 2012 and 3.4 kernel just launched. Why should I restrict myself to pre-2.6.24 artefacts then?

I don't get why ppl try to make one system so bad, usually there is always a usecase where a system shines.

Maybe because they had hard time to find that use case? In fact I only know 2 "usecases" where it "shines". Apple ripped off some code to (proprietary) MacOS X and Juniper did same for (proprietary) JunOS. So those two got their usecases and could be happy. For some obscure reason this does not looks like huge feature to me.

[0xBADCODE]

How so? Users can continue to use original product and have access to it's source code without any restriction. How come they've lost their freedoms?

Simple! If device runs Linux, I can ask for source for device I'm using, learn it and change it if I need this. Will not work with Juniper and BSD "freedom" though, etc. So it's just a couple of corporations who can actually enjoy by BSD freedoms. And everyone else loses. As you see, I'm not as powerful as Juniper so I can't do a complete porting and development cycle on my own and almose everyone else seems to be quite the same. This also led to situation where FBSD formally supports MIPS but there is almost no real-world MIPS hardware where you can actually RUN it. So it looks like if BSD freedom is completely theoretical one

[Libreman]

How so? Users can continue to use original product and have access to it's source code without any restriction. How come they've lost their freedoms?

Do I really need to repeat myself? I've said it already - I do not think it's hard to get unless you're trying very hard not to get it. Either you care about those freedoms in which case you want to protect them (GPL) and prevent others to use your code and take them away from the users of the modified software (they can do whatever they want with their own code) or you do not care about those freedoms in which case you do not mind somebody using your code and taking those freedoms away from its users.

How can you care about preserving those freedoms, consider them important and use a license that allows to disregard them? It's inconsistent and doesn't make sense. And indeed when you just talk with people from both camps you'll find that's exactly the case, people from the "permissive" camp generally do no have much problem with Apple's walled gardens, using MS or other exploits specific to proprietary software - if they do have any complaint it's just technical. It's the opposite in the Linux camp, many people care mostly about the freedom and are willing to even let slide minor technical glitches or inadequacies to preserve it in the hope it will get solved long term (which it slowly is happening, step by step).

It's not really hard to see the difference in mindset - of course there are exceptions in each camp and they tend to overlap but overall I find it quite accurate.

[LightBit]

If somebody distributes copies of his modified versions to others under different conditions, only his modifications will be under different conditions, beacause original version is still available under the same conditions.