But they're not. They're looking for a driver. They find the GPL version, which now has no mention of the BSD version, and they STOP LOOKING. This is basic human nature.
Only if that's what they wanted. If they wanted to do something prohibited by the GPL, they would continue looking for a driver under a more permissive license. If they don't want to do something prohibited by the GPL, it doesn't matter whether they get a GPL or BSD version.
Removing the BSD license from this code dcoesn't benefit the GPL version in any way. It simply damages the BSD version.
On the contrary; it makes license management easier for the linux kernel, as I've said.
Indeed, the entire point of the GPL is to make your life difficult if you don't release YOUR code under the GPL, too
No, the GPL is a permission, a positive thing; it's point is to make your life easy if you release your code under the GPL.
I don't have any control over that. The patent owner controls that. If he says "no", that's the end of it.
Everything has a price.
Rather the opposite, actually. I'm denying that YOU are entitled to MY code. My code belongs to me, just like yours belongs to you. If you choose to give yours away, and you did, then I should be able to use it.
The GPL is very much not giving it away; I can understand your disappointment with its conditions if you thought that was what it was doing, but that's a misconception on your part.
It is legal for you to attach any condition you like to the use of your code, just like the patent owner. But that does not make it right.
I have no moral obligation to give you the code (and you seem very much to agree with that, viz "I should be able to make my own decision about whether to give mine away, because it's mine."); *any* permission I give you to use it under any conditions* (be they paying me a license fee, GPLing your own code, or letting me sleep with your sister) is a unilateral offer on my part, and cannot be wrong; by making this offer I harm noone and nothing. It is of course up to you whether to accept it or not, depending on whether you think my price is too high.
There is more than enough room in this world for everyone to share only the code they want, and we'll all have plenty of code for plenty of purposes.
I'm unconvinced that there would be; witness the lack of a decent BSD-licensed optimizing C compiler. And in any case, that would mean the free code would be far behind the proprietary code in terms of features, very few people would use it, and therefore there would be very little driver and other support for free systems. Which would be a worse situation.
But as long as a license exists that pretends to be about freedom while taking those freedoms away, we're going to have a constant stream of ignorant people ending up in the prison of that license where their code isn't really free.
If it's their own code, they can always relicense it. If they needed a GPL library, then their only options were GPL or not at all, and it was their decision to write the code under those circumstances. In any case, if they suddenly switch their view on whether a license is free enough, they have only themselves to blame for any inconvenience they do themselves by doing so, and I don't see why I should be expected to feel responsible for their plight.
The GPL damages this community, and actively destroys the ideals we've worked to instill in it. It is an abomination. It needs to be stopped.
Disagree, really. I have yet to see any definition of open source or its ideals which rules out the GPL; quite the contrary in fact. I feel your community must be a different one from that of me, Linux, and at least several *BSD developers, and, I would venture to suggest, a much smaller one.
A comparison including it was done a while ago, and posted here; Konqueror was faster than anything other than Opera on the html, images and css rendering things but sucked for javascript./happy konqueror user
Not when the no moving parts method requires O(n) transistors (where n is the amount of storage) and the moving parts one only needs O(1) (or, to be truly honest, O(log n)).
That's because you're a computer nerd (one can tell by the fact that you were using the internet at all 11 years ago). Normal people can't and don't remember those numbers, so ICQ was pretty much doomed once normal people started to outnumber geeks on the internet.
Java beat Lisp because it had the huge Sun marketing machine behind it, and because Lisp isn't quite as good as its fanboys think. It's not a snazzy name that you need to become popular (though it helps a little), it's reasonable prices and decent marketing. Both of which NetApp seems to lack.
The freedom to incorporate the code with his own code however he deems appropriate.
But they have that anyway, because the BSD version of the code is still out there; if they're looking for a BSD-licensed driver for that particular card, they'll find it.
They remove a useless freedom: the freedom to take away the names of contributors altogether. This doesn't benefit the person who does it at all,
Then why does it need to be stopped? If it didn't benefit the person who did it, no-one would do it.
I don't like software patents. I think they're wrong. But I don't get to choose the way of the world. When the algorithm my product needs is patented, I have no choice but to license it. That's the law. And when that license says I can't distribute the source code, that's the law, too. When you put your code under a license that I can't legally combine with this algorithm, you don't force me to fight software patents. You simply force me to reinvent the wheel and write code that does the same thing yours does.
Or to license the patent with the ability to distribute the source; if there is enough GPL code you want to use that it would be cheaper to do that than rewrite all that code, then the rational thing for you to do is license the patent. Which is the point.
Now, if you DIDN'T make me spend two hundred hours designing, building, testing, and debugging code you've already written... what ELSE might I have chosen to do with that two hundred hours?
Do you suppose I might have been grateful, and tried to give something back to the community?
It's possible, but it seems unlikely. You're already displaying a sense of entitlement towards my code that I wrote, costing you nothing - I didn't make you spend those two hundred hours, you spent them because you wanted a program that did whatever it is. You're no worse off than if I'd never written my code.
None whatsoever, if he's a GPL bigot who doesn't understand why anyone would prefer BSD.
If, however, he's smart enough to realise that the GPL is bad for everybody and needs to die... that's a very useful freedom indeed.
Again, what useful freedom does he *actually* have? The freedom to feel all warm and fuzzy inside because one tiny part of his GPLed kernel isn't GPLed?
And look, it's controversial enough to be news on Slashdot. Don't you suppose MAYBE that means this isn't quite the trivial matter you seem to think it is?
Of course; that's the reason I'm bothering to discuss it. But so far I've yet to see one thing to convince me it's anything other than BSD zealots making a fuss about nothing.
No, but I can deny that you have that freedom in the first place, on the grounds that it's an illegal practice.
Then there's no need for the BSD license to prevent it. If the restrictions in the BSD license don't actually remove any useful freedoms, then what is the point in them?
Oh, I'm sorry. Did I not make it clear that I've licensed the right to distribute compiled binaries of the algorithm? I keep forgetting that most GPL bigots know squat about how real programming works in the real world.
So they'll be dependent on you for binaries, unable to fix bugs they come across or use your code for anything else. Most GPL folks will tell you exactly why they chose the GPL - because they want no part in that situation, and don't want their code to help you place your users in it. And don't you think it's more than a little hypocritical to be using someone else's code for something where the users won't see its original, liberal license, when this is the same thing you're criticising linux for?
Seriously, if your gaming handle is based around your OS, you are kinda of... Well... A jackass. Nobody gives a shit what OS you like, just play the game.
I could say the same about any possible basis for a handle - sports teams, nationality, favourite movies. Seriously, who the hell is actually going to mind what your handle is? I can understand blocking profanity, but other than that will anyone actually care?
Re:Biggest myths of all have been around for ages.
on
Why Myths Persist
·
· Score: 2, Insightful
Not everyone who professes to be religious believes in a white robed deity sitting on a cloud chucking thunderbolts. To a logical person, the concept of an anthropomorphic divinity is laughable
Strawman. To a logical person, the concept of a divinity is laughable.
The problem is, what the hell language do you use to describe such a thing? You can call it "energy", or the "Force", but that gets you lumped in with the crystal wavers that are often more flaky than your traditional religious types. So you say God, knowing full well that 99% of the people who hear you don't have a clue what you really mean.
No, you're misunderstanding. Those of us who argue against God aren't (by and large) trying to argue against the man in the white robes, they are arguing against every concept of a deity.
Don't be so quick to dismiss those who profess to be religious. Damn near all of the greatest scientific minds of the last thousand years fall into that category.
And close to all of the greatest scientific minds of the last hundred years don't. Sure, individuals can be religious and intelligent, but on the large, statistical scale, religious people are more likely to be stupid than those who are not.
going ONLY with the same 5 senses we share with our cats and dogs, and having faith in what other humans say, is 100% GUARANTEED to keep a person mediocre.
Bollocks; it's been done by most of the outstanding scientists of the modern age.
How very strange! Here I've been going around thinking that what scientists were doing was wishful thinking. For example, "I sure hope the Scientific Method is a valid technique to verify what the universe really is, or all our work will come tumbling down around our ears" or "I sure hope modus ponens ponens never yields a contradiction," or "I sure hope the next study doesn't suggest that eating eggs is good for you." I've seen several attempts by scientists to verify the Scientific Method, but all of them use the Scientific method, which only stands for the unremarkable proposition that the Scientific Method is valid provided that the Scientific Method is valid. Now don't let me be one to deny your faith. I'm perfectly willing to allow you to use the Scientific Method to verify the Scientific method, but, in fairness, I think I should require you to allow me to use the Bible to verify the Bible.
If you like, but in that case it should be the method you live your life by. You fly by whatever the Bible says is the way to fly, and I'll fly in a scientifically constructed aeroplane, and we'll see which of us gets to Japan first. I must also ask by what Biblical method you are posting on Slashdot.
That analogy doesn't help me much as I've never hit any heroin. Would it be anything like the feeling Friedrich August Kekulé von Stradonitz got when he woke from his dream about the snake biting its tail giving him the chemical structure of benzene? Or the feeling Archimedes got upon leaving his bath?
Do those analogies help you any more, given that I doubt you're either of those people? Anyway, the answer is probably yes, but we don't think those were good things just because they made those people feel good; we think they were good because of their results. There are many things, both good an bad, which will give you a good feeling, and so your feeling alone is not a justification for anything.
What the hell is up with all the scathing remarks?! Let's remember that the ATI acquisition by AMD is new and let's be impressed, considering past support, that progress is being made in the Linux ATI drivers arena AT ALL!
But there's no *real* progress. All there is is more *talk*, and we've been hearing for *years* that the next lot of ATI drivers will make everything right. They never do, and so we've lost patience with all these *empty* announcements. When they actually release a working driver I'll start cheering. Not before.
Because it removes the recipient's knowledge of his licensing options, under the rather likely assumption that he's not going to invest much time in researching it, because the GPL crowd doesn't want anything to be released under any other license ever.
What use is the extra option the recipient gets though? If they're looking to submit changes under the BSD license, they can still do that. If they're looking to use linux under the BSD license, they still can't do that. If they're looking to implement the driver for another OS, they will spend time researching it.
Prevented? No. But submitting a change with the note that your change is only GPL and not BSD would make you look like a dick, so NOBODY WOULD DO IT.
That's not the case, because as the facts have come out it looks like that's what actually happened here - someone made substantial contributions to the driver, wanted them to be GPL-only, and so changed the license of the driver to be GPL. Also see e.g. ffmpeg's ac3 encoder; people can and do make GPLed contributions to projects under more liberal licenses.
To secure proper attribution for the original author(s). Your useful freedom is not infringed by this requirement.
You can't deny that the freedom to falsely claim authorship of a piece of code (e.g. when selling it to a client) is a useful one. Sure, it's the freedom to be a dick - but so is the freedom to redistribute under a propriety license that you get with BSD.
The BSD license recognises that when you write your changes, they are yours, and you can do whatever you want with them. This is a useful freedom. It means that if I have a patented algorithm I licensed from someone else, which I can't legally distribute or license, I can use it in my changes and still honor my existing commitment to keep it confidential. No matter what your opinion of this, that's a freedom that I have under the BSD license, and it has value.
How so? You wouldn't be able to distribute your changes, so you wouldn't need any license - you can make the changes and keep them to yourself just as well under GPL as BSD.
Which is unarguably more complicated than module X being under the GPL.
Removing the BSD license doesn't actually change anything for anyone.
So why is it a problem?
What changes things is when authors contribute changes under the GPL, contaminating the Linux version and preventing those changes from going into the BSD version.
But keeping the dual license would in no way have prevented this.
Premise: freedom = good.
GPL = less freedom.
Ergo, GPL = less good.
That simple.
If you hold that good and freedom are the same, then why BSD license rather than public domain? See, the Linux folks think the GPL is better than BSD (and I agree with them), so they can equally well state that it is *better* for the contributions of those who know nothing about software licensing to fall under the GPL.
> But what freedom has been removed in practice?
I am not your lawyer. If you really want to know, go get one. Honestly, I think you're just trying to make me waste time explaining a complicated subject.
Bollocks you do. You're just making excuses because you know full well no meaningful freedom has actually been removed.
You find there several example of how disruptive technologies could work. There is something like a "low-end disruption". This does not fit. Solaris does not strive to be low-end.
In what it does, it actually is; you won't get a machine with full support cheaper, and you won't get a better OS free. And to a certain extent it's new-market disruption compared to Linux, because Solaris can sell to the "big iron" customers which Linux simply can't. But I'm not thinking it will succeed because it's disruptive in that sense; rather it's just better than the alternatives.
Will Sun really invest serious money in Solaris for as long as it takes?
That's an important question, and one I'm not really sure of myself. I suspect it depends how successful it is; if no-one buys Solaris, they'll stop pushing it. But they've shown that they will continue to support users even if the technology goes out of fashion, and while they've had their share of problems I just can't see Sun going bankrupt, so even in the worst case you're unlikely to be left high and dry.
This is more or less what I said from the beginning: For the majority of Linux users is simply does not matter, whether behind their desktop a Solaris Kernel, a *BSD kernel, or a Linux kernel works. I worked on BSD machines without noticing a difference, I did not really notice a difference in different Linux kernel versions since 2.2 -> 2.4.
I agree entirely; but by the same token, there is very little preventing Linux users switching to Solaris, since they can run the same programs, so if Solaris takes off in the corporate sphere (and I think it might, though it's by no means certain, nor even more likely than not) we could well see a migration away from Linux.
For example, Solaris had schedular activations a while back. It turned out though that since very few people were interested in running new code on the platform it didn't help them. They eventually threw it out because utilizing it required the users to code to a new API. In the Linux model such a thing has an opportunity for adoption.
By *forcing* developers to recode to a new API. Under the Solaris model developers still have the choice of using the new method, so they gain everything they had under Linux, but additionally they have the choice of not changing and their code still works. Which I call a win for Solaris.
Solaris is required by market forces to maintain backwards compatability with antiquanted technologies, this means that your code will always run, but it also means solaris is slower and far less advanced then it could be.
Solaris still seems to have some pretty advanced stuff in it - dtrace, ZFS, clustering, all the big selling points. And, fundamentally, how advanced does an OS need to be? I can't think of anything linux as an OS lets applications do that solaris doesn't, wheras I still can't burn CDs in linux as well as I could in 2.4, and I've had to change what manages my/dev twice, because they decided it would be better to rip out the perfectly good API and replace it with something else.
Claiming that this makes Solaris strictly "better" and that it's because they're programmers are better, or more "work harder" as you imply,
I imply this because the excuses given in Linux's own "stable-api-nonsense.txt" seem to boil down to "we can't be bothered to maintain one".
Take your pick, but you can't call one strictly "better".
I can and do, because one of them works reliably and the other doesn't, while I have yet to see any convincing advantages in the other direction.
Personally I run Linux because it actually does what I need it to do. If I wanted a system that doesn't do much, but does it well I'd probably use Solaris. If I want something in between I'll use a BSD. Honestly Solaris isn't designed for the userland to be used much, it's mostly just a kernel for NFS, kernel based webservers etc. Their userland is still quite antiquated and far from being read for desktop use.
It's antiquated but it works; it's mostly there as a fallback these days, and if it really bothers you you can use the GNU tools instead.
And by that I mean the USERLAND, not the fsck'ing GUI stuff that all the former windows weenies turned Linux fanboys seem to care about, that's LONG away in solaris.
Huh? It's running the same X/Gnome/etc. as linux, so unless the writers have been using gratuitous nonstandard linuxisms, all the work that goes into that on Linux will also run on Solaris.
Each OS has it's merits, that's why they're all still around.
Sure, but just like with programming languages, that doesn't mean there aren't those that are better than others.
I DO care about the newest features and newest hardware support. That's why I use Linux.
I use linux on a couple of my machines for the sake of hardware support, and that's one area in which Linux has a definite advantage over Solaris. But that's a temporary, fixable thing, wheras Linux's experimental nature seems to be a deep-rooted and fundamental flaw in it.
Putting BSD software under GPL doesn't make things simpler. BSD licensing is about as simple as you get,
Yes, but "Linux is under the GPL" is simpler and easier than "Linux is under the GPL, except for module X which is BSD".
and the GPL is arguably the most complicated open source license on the planet.
and the GPL is arguably the most complicated open source license on the planet.
I can't let that one slide; just compare it to the CDDL, or the APL (to pick two fairly arbitrary examples), with their great big definition sections.
I am of the opinion that if your license strategy requires your volunteer contributors to understand the legal nuances of software licensing before they can make an informed decision on how and where to contribute, you are a sadistic asshole.
Your statement is at once entirely true and utterly irrelevant; if someone knows nothing about software licensing, how is their contribution falling under GPL any better or worse than it falling under BSD?
There are freedoms under the BSD license that don't exist under the GPL.
Removing them to make license management easier is just dickhead behavior.
But what freedom has been removed in practice? The freedom to put your own contributions under the BSD license? Nope, still there. The freedom to contribute your own changes to the BSD branch as well? Nope, still there, you are the copyright holder after all. The freedom to contribute another developer's changes to the BSD branch against their will? That was never there anyway, they could make their contributions GPL only and then you wouldn't be able to port them to BSD. The freedom to port the driver to some other OS with a GPL-incompatible license, say Solaris? Nope, still there, you just start from the BSD copy. The freedom to use the entire Linux kernel under BSD? Never existed. So go on, what freedom has *actually* been lost here?
This alone makes it hard for Solaris to challenge Linux. With Windoze I heard very often the argument that it is so successful in companies because so many users use it at home and got used to it. This is probably true, but works in this case for Linux. There is no reason for an end user to change. And let's be truthful, if there is no real pressing need, even professionals are a bit reluctant to change and learn something new.
It could still change, it'll just take time - there *are* pressing reasons for businesses to change to Solaris. And if/when it becomes big in business, then it'll be picked up by schools; it's already fairly popular in some universities, and this will trickle down, and schools are always strapped for cash but in need of support, so the all-in-one deals you get from Sun (hardware, software, support) have got to look pretty attractive to them. And when you've got the children using it at school, and perhaps the parents at work, and you can get a free version off the 'net, and dual boot for a while if you want/need to, home users are going to switch - it's not as if it's any *worse* than the alternatives, and there's value in using the same software you use elsewhere. It'd take a while, but so did linux; it's still possible for Solaris to conquer the world at this stage.
Ok, this might be true. But I did not solely mean official written and released documentation. Really for every problem I ever had with Linux, I found without effort a solution in Google. Solaris has a much smaller user base, would I be also so lucky to find solutions even for arcane problems with Solaris?
Well, the great advantage of Solaris is that you mostly don't need to. But, in my (admittedly fairly limited - it's only happened once or twice) experience, when you do need to go to the internet it's helpful - the community may be small, but it's quite active, e.g. the newsgroup is pretty responsive. Also the community is probably larger than for any given linux distribution.
What if it is not Sun software? Suppose I have problems with Firefox? When I had problems with Linux it wasn't necessarily Linux, but some 3rd party program I needed.
Sun ships their own, supported version (IIRC), and this is the case for many popular programs. Of course if you're going to load up with random programs from across the net you're no better off than you were with Linux, but solaris as shipped comes with a pretty complete set of applications for most needs; for corporate desktops you wouldn't want people installing any other software.
Is this something I want or care for?
How much you care is of course down to personal preference, but I think everyone sees it as good; it makes it a lot easier to use programs you only use occasionally if they behave like the ones you use more often.
Even under Solaris I'd probably want to code for KDE or using pure Qt. I'd want to run Firefox, which is not from Sun. I'd want to use a couple of other programs, which are also not from Sun.
Fair enough (though I have to ask why you want those particular programs), but that's pretty unusual. Sun is aiming at the people who just want a web browser, office suite, development environment etc. and don't much care what it is, which I think is the majority of computer users. If you've got your heart set on KDE/firefox then you're not asking "what is the best OS" but rather "what is the best OS to run these particular programs on". And that may well be Linux, though KDE/Solaris is quite mature, but it's not officially supported by Sun so that negates many of the advantages Solaris has.
So please could anyone tell me, what are the USP's of Solaris?
Frankly, there aren't many for end users; it's advantages are mostly for enterprise. That said:
I disagree with your claim that Linux has the best documentation; in my experience linux documentation is very variable, and some packages just aren't documented at all. Solaris has good, consistent documentation, all in the same format, written in the same style, using the same conventions, which is very useful.
Vendor support; sure e.g. Red Hat will sell you this, but they can't fix every problem because it's often not their software. Sun can and does get the programmer who wrote the software you're using to look at your bug, if they need to - and they'll support your software and hardware (also made by their own engineers) for less than RHEL charges for the software.
Stability and maturity; these are mostly subjective, but my solaris machine never one crashed until its hardware failed, while even the my debian stable box crashed once in the same period (to be fair I should mention that my OpenBSD machine also never crashed in the same period); certainly the greater maturity of Solaris is undeniable.
Consistency; the Solaris software is all from the same source, so there's a lot less of the e.g. differing styles for commandline arguments that you see on Linux, and the GUI admin tools are better integrated in the system, wheras those shipped by many Linux distros can have a "bolted-on" feel, IME (e.g. YaST frequently nukes any changes to the config files you made by hand).
Software support; sun's own software obviously has Solaris as its primary platform, but also many of the programs we think of as "linux" utilities e.g. GNU tar have solaris as their original platform; also since it has Sun behind it the big Oracle-type application vendors are more likely to support it.
If I recall correctly, sun has a pretty good clustering product for Solaris, which Linux rather lacks.
There are various specific things that solaris has, e.g. DTrace or ZFS, but these won't interest you unless you're working in that particular area.
From the CDDL: "that Source Code form
must be distributed only under the terms of this License." So, basically, no; CDDL is bringing in just as heavy restrictions as the GPL, two licenses with such restrictions cannot coexist in the same work, and the GPL was there first.
as will.NET once people have gotten used to it, forgotten the hype that surrounds it and started to find the same flaws in it as Java has (after all, C# practically is Java).
But C#, coming as it does after Java, corrects many of Java's stupidities; it's a lot easier to interface with native code, it doesn't have the ludicracies of Java's type system, it's got generics properly integrated in since they were there from the start (Java has them now but a lot of the huge standard library isn't converted to them yet), it does the GUI right by leaving that to bindings to native libraries, and MS has got the framework right by shipping.net with multiple language capabilities, rather than the Java thing of ruining what's actually quite a nice virtual machine environment with the most horrible language I have had the misfortune to program in. Sure, in many ways it's a clone of Java, but it's a better clone of Java.
What can Sun Micro Systems bring to the table that rest of the Linux could not? Its name, some kind of relationships with corporations and provide "not a bunch of amateurs in their spare time, this OS is backed by professionals" kind of sales talk. But that niche is already occupied by IBM.
Sun does a fully integrated system - their hardware, running their OS, with their applications on it, and their own support - and does it a helluva lot cheaper than IBM.
As for SUNW's vaunted professionalism, they fumbled the lead they had in unix and are struggling to keep up.
The technology that leads isn't always the one that's technically best. Solaris is actually a better system than Linux - it's better designed, more mature, more stable (which, yes, it does by losing out on some of the leading edge innovation - but that's a good tradeoff to make in the corporate sphere), and a lot better integrated as an OS than any linux distribution - unsurprising given that most of it comes from a single vendor and was written together (and as for the GNU tools, most of them were originally written on solaris, and generally work at least as well there as on Linux).
No it doesn't. I run Linux/PPC and I *never* see hardware manufacturers releasing drivers for their hardware on it.
Erm, yeah, because the fact that no drivers are being released on a system which doesn't have a stable ABI means that a stable ABI doesn't help. Oh, wait. If you'd made a comparison with an OS which had a stable ABI you'd have a point.
The only way to get a decent driver for Linux (Not Linux/x86-32, but Linux) is for it to be in the main kernel tree.
Yes, and this is a problem, especially since it extends outside the kernel - whoops, we just broke the CD burning/wireless/video interfaces because we felt like it. No-one maintaining the program you use for them any more? Then you're SOL. Solaris can use drivers (and programs) that were written ten years ago, without any modifications to the driver, because it's got a well designed kernel and its developers are willing to put the effort in to maintain a stable ABI.
Slashdot Mirror
Ah, a wikipedia article tagged for original research, truly an authoritative source of information.
Only if that's what they wanted. If they wanted to do something prohibited by the GPL, they would continue looking for a driver under a more permissive license. If they don't want to do something prohibited by the GPL, it doesn't matter whether they get a GPL or BSD version.
Removing the BSD license from this code dcoesn't benefit the GPL version in any way. It simply damages the BSD version.
On the contrary; it makes license management easier for the linux kernel, as I've said.
Indeed, the entire point of the GPL is to make your life difficult if you don't release YOUR code under the GPL, too
No, the GPL is a permission, a positive thing; it's point is to make your life easy if you release your code under the GPL.
I don't have any control over that. The patent owner controls that. If he says "no", that's the end of it.
Everything has a price.
Rather the opposite, actually. I'm denying that YOU are entitled to MY code. My code belongs to me, just like yours belongs to you. If you choose to give yours away, and you did, then I should be able to use it.
The GPL is very much not giving it away; I can understand your disappointment with its conditions if you thought that was what it was doing, but that's a misconception on your part.
It is legal for you to attach any condition you like to the use of your code, just like the patent owner. But that does not make it right.
I have no moral obligation to give you the code (and you seem very much to agree with that, viz "I should be able to make my own decision about whether to give mine away, because it's mine."); *any* permission I give you to use it under any conditions* (be they paying me a license fee, GPLing your own code, or letting me sleep with your sister) is a unilateral offer on my part, and cannot be wrong; by making this offer I harm noone and nothing. It is of course up to you whether to accept it or not, depending on whether you think my price is too high.
There is more than enough room in this world for everyone to share only the code they want, and we'll all have plenty of code for plenty of purposes.
I'm unconvinced that there would be; witness the lack of a decent BSD-licensed optimizing C compiler. And in any case, that would mean the free code would be far behind the proprietary code in terms of features, very few people would use it, and therefore there would be very little driver and other support for free systems. Which would be a worse situation.
But as long as a license exists that pretends to be about freedom while taking those freedoms away, we're going to have a constant stream of ignorant people ending up in the prison of that license where their code isn't really free.
If it's their own code, they can always relicense it. If they needed a GPL library, then their only options were GPL or not at all, and it was their decision to write the code under those circumstances. In any case, if they suddenly switch their view on whether a license is free enough, they have only themselves to blame for any inconvenience they do themselves by doing so, and I don't see why I should be expected to feel responsible for their plight.
The GPL damages this community, and actively destroys the ideals we've worked to instill in it. It is an abomination. It needs to be stopped.
Disagree, really. I have yet to see any definition of open source or its ideals which rules out the GPL; quite the contrary in fact. I feel your community must be a different one from that of me, Linux, and at least several *BSD developers, and, I would venture to suggest, a much smaller one.
A comparison including it was done a while ago, and posted here; Konqueror was faster than anything other than Opera on the html, images and css rendering things but sucked for javascript. /happy konqueror user
Not when the no moving parts method requires O(n) transistors (where n is the amount of storage) and the moving parts one only needs O(1) (or, to be truly honest, O(log n)).
Yes, because economics of scale mean that doing the same thing in a larger country is less efficient. Oh, wait.
That's because you're a computer nerd (one can tell by the fact that you were using the internet at all 11 years ago). Normal people can't and don't remember those numbers, so ICQ was pretty much doomed once normal people started to outnumber geeks on the internet.
Java beat Lisp because it had the huge Sun marketing machine behind it, and because Lisp isn't quite as good as its fanboys think. It's not a snazzy name that you need to become popular (though it helps a little), it's reasonable prices and decent marketing. Both of which NetApp seems to lack.
But they have that anyway, because the BSD version of the code is still out there; if they're looking for a BSD-licensed driver for that particular card, they'll find it.
They remove a useless freedom: the freedom to take away the names of contributors altogether. This doesn't benefit the person who does it at all,
Then why does it need to be stopped? If it didn't benefit the person who did it, no-one would do it.
I don't like software patents. I think they're wrong. But I don't get to choose the way of the world. When the algorithm my product needs is patented, I have no choice but to license it. That's the law. And when that license says I can't distribute the source code, that's the law, too. When you put your code under a license that I can't legally combine with this algorithm, you don't force me to fight software patents. You simply force me to reinvent the wheel and write code that does the same thing yours does.
Or to license the patent with the ability to distribute the source; if there is enough GPL code you want to use that it would be cheaper to do that than rewrite all that code, then the rational thing for you to do is license the patent. Which is the point.
Now, if you DIDN'T make me spend two hundred hours designing, building, testing, and debugging code you've already written... what ELSE might I have chosen to do with that two hundred hours?
Do you suppose I might have been grateful, and tried to give something back to the community?
It's possible, but it seems unlikely. You're already displaying a sense of entitlement towards my code that I wrote, costing you nothing - I didn't make you spend those two hundred hours, you spent them because you wanted a program that did whatever it is. You're no worse off than if I'd never written my code.
If, however, he's smart enough to realise that the GPL is bad for everybody and needs to die... that's a very useful freedom indeed.
Again, what useful freedom does he *actually* have? The freedom to feel all warm and fuzzy inside because one tiny part of his GPLed kernel isn't GPLed?
And look, it's controversial enough to be news on Slashdot. Don't you suppose MAYBE that means this isn't quite the trivial matter you seem to think it is?
Of course; that's the reason I'm bothering to discuss it. But so far I've yet to see one thing to convince me it's anything other than BSD zealots making a fuss about nothing.
No, but I can deny that you have that freedom in the first place, on the grounds that it's an illegal practice.
Then there's no need for the BSD license to prevent it. If the restrictions in the BSD license don't actually remove any useful freedoms, then what is the point in them?
Oh, I'm sorry. Did I not make it clear that I've licensed the right to distribute compiled binaries of the algorithm? I keep forgetting that most GPL bigots know squat about how real programming works in the real world.
So they'll be dependent on you for binaries, unable to fix bugs they come across or use your code for anything else. Most GPL folks will tell you exactly why they chose the GPL - because they want no part in that situation, and don't want their code to help you place your users in it. And don't you think it's more than a little hypocritical to be using someone else's code for something where the users won't see its original, liberal license, when this is the same thing you're criticising linux for?
I could say the same about any possible basis for a handle - sports teams, nationality, favourite movies. Seriously, who the hell is actually going to mind what your handle is? I can understand blocking profanity, but other than that will anyone actually care?
Strawman. To a logical person, the concept of a divinity is laughable.
The problem is, what the hell language do you use to describe such a thing? You can call it "energy", or the "Force", but that gets you lumped in with the crystal wavers that are often more flaky than your traditional religious types. So you say God, knowing full well that 99% of the people who hear you don't have a clue what you really mean.
No, you're misunderstanding. Those of us who argue against God aren't (by and large) trying to argue against the man in the white robes, they are arguing against every concept of a deity.
Don't be so quick to dismiss those who profess to be religious. Damn near all of the greatest scientific minds of the last thousand years fall into that category.
And close to all of the greatest scientific minds of the last hundred years don't. Sure, individuals can be religious and intelligent, but on the large, statistical scale, religious people are more likely to be stupid than those who are not.
Bollocks; it's been done by most of the outstanding scientists of the modern age.
If you like, but in that case it should be the method you live your life by. You fly by whatever the Bible says is the way to fly, and I'll fly in a scientifically constructed aeroplane, and we'll see which of us gets to Japan first. I must also ask by what Biblical method you are posting on Slashdot.
That analogy doesn't help me much as I've never hit any heroin. Would it be anything like the feeling Friedrich August Kekulé von Stradonitz got when he woke from his dream about the snake biting its tail giving him the chemical structure of benzene? Or the feeling Archimedes got upon leaving his bath?
Do those analogies help you any more, given that I doubt you're either of those people? Anyway, the answer is probably yes, but we don't think those were good things just because they made those people feel good; we think they were good because of their results. There are many things, both good an bad, which will give you a good feeling, and so your feeling alone is not a justification for anything.
But there's no *real* progress. All there is is more *talk*, and we've been hearing for *years* that the next lot of ATI drivers will make everything right. They never do, and so we've lost patience with all these *empty* announcements. When they actually release a working driver I'll start cheering. Not before.
What use is the extra option the recipient gets though? If they're looking to submit changes under the BSD license, they can still do that. If they're looking to use linux under the BSD license, they still can't do that. If they're looking to implement the driver for another OS, they will spend time researching it.
Prevented? No. But submitting a change with the note that your change is only GPL and not BSD would make you look like a dick, so NOBODY WOULD DO IT.
That's not the case, because as the facts have come out it looks like that's what actually happened here - someone made substantial contributions to the driver, wanted them to be GPL-only, and so changed the license of the driver to be GPL. Also see e.g. ffmpeg's ac3 encoder; people can and do make GPLed contributions to projects under more liberal licenses.
To secure proper attribution for the original author(s). Your useful freedom is not infringed by this requirement.
You can't deny that the freedom to falsely claim authorship of a piece of code (e.g. when selling it to a client) is a useful one. Sure, it's the freedom to be a dick - but so is the freedom to redistribute under a propriety license that you get with BSD.
The BSD license recognises that when you write your changes, they are yours, and you can do whatever you want with them. This is a useful freedom. It means that if I have a patented algorithm I licensed from someone else, which I can't legally distribute or license, I can use it in my changes and still honor my existing commitment to keep it confidential. No matter what your opinion of this, that's a freedom that I have under the BSD license, and it has value.
How so? You wouldn't be able to distribute your changes, so you wouldn't need any license - you can make the changes and keep them to yourself just as well under GPL as BSD.
Which is unarguably more complicated than module X being under the GPL.
Removing the BSD license doesn't actually change anything for anyone.
So why is it a problem?
What changes things is when authors contribute changes under the GPL, contaminating the Linux version and preventing those changes from going into the BSD version.
But keeping the dual license would in no way have prevented this.
Premise: freedom = good.
GPL = less freedom.
Ergo, GPL = less good.
That simple.
If you hold that good and freedom are the same, then why BSD license rather than public domain? See, the Linux folks think the GPL is better than BSD (and I agree with them), so they can equally well state that it is *better* for the contributions of those who know nothing about software licensing to fall under the GPL.
> But what freedom has been removed in practice?
I am not your lawyer. If you really want to know, go get one. Honestly, I think you're just trying to make me waste time explaining a complicated subject.
Bollocks you do. You're just making excuses because you know full well no meaningful freedom has actually been removed.
In what it does, it actually is; you won't get a machine with full support cheaper, and you won't get a better OS free. And to a certain extent it's new-market disruption compared to Linux, because Solaris can sell to the "big iron" customers which Linux simply can't. But I'm not thinking it will succeed because it's disruptive in that sense; rather it's just better than the alternatives.
Will Sun really invest serious money in Solaris for as long as it takes?
That's an important question, and one I'm not really sure of myself. I suspect it depends how successful it is; if no-one buys Solaris, they'll stop pushing it. But they've shown that they will continue to support users even if the technology goes out of fashion, and while they've had their share of problems I just can't see Sun going bankrupt, so even in the worst case you're unlikely to be left high and dry.
This is more or less what I said from the beginning: For the majority of Linux users is simply does not matter, whether behind their desktop a Solaris Kernel, a *BSD kernel, or a Linux kernel works. I worked on BSD machines without noticing a difference, I did not really notice a difference in different Linux kernel versions since 2.2 -> 2.4.
I agree entirely; but by the same token, there is very little preventing Linux users switching to Solaris, since they can run the same programs, so if Solaris takes off in the corporate sphere (and I think it might, though it's by no means certain, nor even more likely than not) we could well see a migration away from Linux.
By *forcing* developers to recode to a new API. Under the Solaris model developers still have the choice of using the new method, so they gain everything they had under Linux, but additionally they have the choice of not changing and their code still works. Which I call a win for Solaris.
Solaris is required by market forces to maintain backwards compatability with antiquanted technologies, this means that your code will always run, but it also means solaris is slower and far less advanced then it could be.
Solaris still seems to have some pretty advanced stuff in it - dtrace, ZFS, clustering, all the big selling points. And, fundamentally, how advanced does an OS need to be? I can't think of anything linux as an OS lets applications do that solaris doesn't, wheras I still can't burn CDs in linux as well as I could in 2.4, and I've had to change what manages my /dev twice, because they decided it would be better to rip out the perfectly good API and replace it with something else.
Claiming that this makes Solaris strictly "better" and that it's because they're programmers are better, or more "work harder" as you imply,
I imply this because the excuses given in Linux's own "stable-api-nonsense.txt" seem to boil down to "we can't be bothered to maintain one".
Take your pick, but you can't call one strictly "better".
I can and do, because one of them works reliably and the other doesn't, while I have yet to see any convincing advantages in the other direction.
Personally I run Linux because it actually does what I need it to do. If I wanted a system that doesn't do much, but does it well I'd probably use Solaris. If I want something in between I'll use a BSD. Honestly Solaris isn't designed for the userland to be used much, it's mostly just a kernel for NFS, kernel based webservers etc. Their userland is still quite antiquated and far from being read for desktop use.
It's antiquated but it works; it's mostly there as a fallback these days, and if it really bothers you you can use the GNU tools instead.
And by that I mean the USERLAND, not the fsck'ing GUI stuff that all the former windows weenies turned Linux fanboys seem to care about, that's LONG away in solaris.
Huh? It's running the same X/Gnome/etc. as linux, so unless the writers have been using gratuitous nonstandard linuxisms, all the work that goes into that on Linux will also run on Solaris.
Each OS has it's merits, that's why they're all still around.
Sure, but just like with programming languages, that doesn't mean there aren't those that are better than others.
I DO care about the newest features and newest hardware support. That's why I use Linux.
I use linux on a couple of my machines for the sake of hardware support, and that's one area in which Linux has a definite advantage over Solaris. But that's a temporary, fixable thing, wheras Linux's experimental nature seems to be a deep-rooted and fundamental flaw in it.
Yes, but "Linux is under the GPL" is simpler and easier than "Linux is under the GPL, except for module X which is BSD".
and the GPL is arguably the most complicated open source license on the planet.
and the GPL is arguably the most complicated open source license on the planet.
I can't let that one slide; just compare it to the CDDL, or the APL (to pick two fairly arbitrary examples), with their great big definition sections.
I am of the opinion that if your license strategy requires your volunteer contributors to understand the legal nuances of software licensing before they can make an informed decision on how and where to contribute, you are a sadistic asshole.
Your statement is at once entirely true and utterly irrelevant; if someone knows nothing about software licensing, how is their contribution falling under GPL any better or worse than it falling under BSD?
There are freedoms under the BSD license that don't exist under the GPL.
Removing them to make license management easier is just dickhead behavior.
But what freedom has been removed in practice? The freedom to put your own contributions under the BSD license? Nope, still there. The freedom to contribute your own changes to the BSD branch as well? Nope, still there, you are the copyright holder after all. The freedom to contribute another developer's changes to the BSD branch against their will? That was never there anyway, they could make their contributions GPL only and then you wouldn't be able to port them to BSD. The freedom to port the driver to some other OS with a GPL-incompatible license, say Solaris? Nope, still there, you just start from the BSD copy. The freedom to use the entire Linux kernel under BSD? Never existed. So go on, what freedom has *actually* been lost here?
It could still change, it'll just take time - there *are* pressing reasons for businesses to change to Solaris. And if/when it becomes big in business, then it'll be picked up by schools; it's already fairly popular in some universities, and this will trickle down, and schools are always strapped for cash but in need of support, so the all-in-one deals you get from Sun (hardware, software, support) have got to look pretty attractive to them. And when you've got the children using it at school, and perhaps the parents at work, and you can get a free version off the 'net, and dual boot for a while if you want/need to, home users are going to switch - it's not as if it's any *worse* than the alternatives, and there's value in using the same software you use elsewhere. It'd take a while, but so did linux; it's still possible for Solaris to conquer the world at this stage.
Ok, this might be true. But I did not solely mean official written and released documentation. Really for every problem I ever had with Linux, I found without effort a solution in Google. Solaris has a much smaller user base, would I be also so lucky to find solutions even for arcane problems with Solaris?
Well, the great advantage of Solaris is that you mostly don't need to. But, in my (admittedly fairly limited - it's only happened once or twice) experience, when you do need to go to the internet it's helpful - the community may be small, but it's quite active, e.g. the newsgroup is pretty responsive. Also the community is probably larger than for any given linux distribution.
What if it is not Sun software? Suppose I have problems with Firefox? When I had problems with Linux it wasn't necessarily Linux, but some 3rd party program I needed.
Sun ships their own, supported version (IIRC), and this is the case for many popular programs. Of course if you're going to load up with random programs from across the net you're no better off than you were with Linux, but solaris as shipped comes with a pretty complete set of applications for most needs; for corporate desktops you wouldn't want people installing any other software.
Is this something I want or care for?
How much you care is of course down to personal preference, but I think everyone sees it as good; it makes it a lot easier to use programs you only use occasionally if they behave like the ones you use more often.
Even under Solaris I'd probably want to code for KDE or using pure Qt. I'd want to run Firefox, which is not from Sun. I'd want to use a couple of other programs, which are also not from Sun.
Fair enough (though I have to ask why you want those particular programs), but that's pretty unusual. Sun is aiming at the people who just want a web browser, office suite, development environment etc. and don't much care what it is, which I think is the majority of computer users. If you've got your heart set on KDE/firefox then you're not asking "what is the best OS" but rather "what is the best OS to run these particular programs on". And that may well be Linux, though KDE/Solaris is quite mature, but it's not officially supported by Sun so that negates many of the advantages Solaris has.
Frankly, there aren't many for end users; it's advantages are mostly for enterprise. That said:
I disagree with your claim that Linux has the best documentation; in my experience linux documentation is very variable, and some packages just aren't documented at all. Solaris has good, consistent documentation, all in the same format, written in the same style, using the same conventions, which is very useful.
Vendor support; sure e.g. Red Hat will sell you this, but they can't fix every problem because it's often not their software. Sun can and does get the programmer who wrote the software you're using to look at your bug, if they need to - and they'll support your software and hardware (also made by their own engineers) for less than RHEL charges for the software.
Stability and maturity; these are mostly subjective, but my solaris machine never one crashed until its hardware failed, while even the my debian stable box crashed once in the same period (to be fair I should mention that my OpenBSD machine also never crashed in the same period); certainly the greater maturity of Solaris is undeniable.
Consistency; the Solaris software is all from the same source, so there's a lot less of the e.g. differing styles for commandline arguments that you see on Linux, and the GUI admin tools are better integrated in the system, wheras those shipped by many Linux distros can have a "bolted-on" feel, IME (e.g. YaST frequently nukes any changes to the config files you made by hand).
Software support; sun's own software obviously has Solaris as its primary platform, but also many of the programs we think of as "linux" utilities e.g. GNU tar have solaris as their original platform; also since it has Sun behind it the big Oracle-type application vendors are more likely to support it.
If I recall correctly, sun has a pretty good clustering product for Solaris, which Linux rather lacks.
There are various specific things that solaris has, e.g. DTrace or ZFS, but these won't interest you unless you're working in that particular area.
From the CDDL: "that Source Code form must be distributed only under the terms of this License." So, basically, no; CDDL is bringing in just as heavy restrictions as the GPL, two licenses with such restrictions cannot coexist in the same work, and the GPL was there first.
But C#, coming as it does after Java, corrects many of Java's stupidities; it's a lot easier to interface with native code, it doesn't have the ludicracies of Java's type system, it's got generics properly integrated in since they were there from the start (Java has them now but a lot of the huge standard library isn't converted to them yet), it does the GUI right by leaving that to bindings to native libraries, and MS has got the framework right by shipping .net with multiple language capabilities, rather than the Java thing of ruining what's actually quite a nice virtual machine environment with the most horrible language I have had the misfortune to program in. Sure, in many ways it's a clone of Java, but it's a better clone of Java.
Sun does a fully integrated system - their hardware, running their OS, with their applications on it, and their own support - and does it a helluva lot cheaper than IBM.
As for SUNW's vaunted professionalism, they fumbled the lead they had in unix and are struggling to keep up.
The technology that leads isn't always the one that's technically best. Solaris is actually a better system than Linux - it's better designed, more mature, more stable (which, yes, it does by losing out on some of the leading edge innovation - but that's a good tradeoff to make in the corporate sphere), and a lot better integrated as an OS than any linux distribution - unsurprising given that most of it comes from a single vendor and was written together (and as for the GNU tools, most of them were originally written on solaris, and generally work at least as well there as on Linux).
Erm, yeah, because the fact that no drivers are being released on a system which doesn't have a stable ABI means that a stable ABI doesn't help. Oh, wait. If you'd made a comparison with an OS which had a stable ABI you'd have a point.
The only way to get a decent driver for Linux (Not Linux/x86-32, but Linux) is for it to be in the main kernel tree.
Yes, and this is a problem, especially since it extends outside the kernel - whoops, we just broke the CD burning/wireless/video interfaces because we felt like it. No-one maintaining the program you use for them any more? Then you're SOL. Solaris can use drivers (and programs) that were written ten years ago, without any modifications to the driver, because it's got a well designed kernel and its developers are willing to put the effort in to maintain a stable ABI.