While it does mean you lose those particular tracks, the mentality I keep hearing from people I would expect to know better is, in a world where everyone has enough bandwidth to stream radio 24/7, nobody cares that you've lost your music collection.
You just switch to a different, competing service, and re-download everything.
The guy I had this conversation with reasoned it like this: If you're going legit, this is the cheapest way. You lose the ability to have stuff work on an iPod, but he had something else anyway. Everything he wanted to do with it, the DRM software let him do -- except play it on Linux, which he didn't want anyway (partly because it didn't work on Linux -- chicken and egg).
And the economics of it: He calculated that he'd have to subscribe to this service for 15 years straight before he'd be spending more than it would cost to buy the stuff outright on iTunes or CD. And that was just counting songs he'd already downladed -- obviously, in 15 years, he'd be downloading a lot more stuff.
Me, I'm not willing to give up my freedom like that, and stuff just has to work on Linux. Besides, I listen to a lot of Internet radio. But content as a service really isn't a problem. Software as a service, maybe, because you have your own data attached to it, but music? Who are we kidding?
First, it's actually better than it was. There used to be a separate root password, and you had to enter it every time. Now, you just enter your normal user password (via sudo), and it remembers it for some 5-10 minutes.
Second, it's not needed for everyday tasks. It's needed for messing with global system settings -- things like your screen resolution -- or for installing software. Most people don't install software every day. (Although it might get a bit annoying doing it for updates...)
Third, it can be turned off, temporarily or completely, or customized in various way. Technically, I don't know what Vista can do here -- I imagine "just turn it off" is an option.
Compare this to Vista. Some software pops up the prompt constantly, sometimes it only pops up intermittently. The vast majority of the times it pops up, there's really nothing special happening. And, to my knowledge, there's nothing like the "remember for 5 minutes" feature of Linux.
Granted, the gap isn't as big as it once was, but the reason we make fun of cancel/allow is not that they ripped off sudo, but that they ripped it off and got it so horribly wrong. Compare what Microsoft has done to what Apple has done, faced with the same problem: Microsoft has essentially kept backwards-compatibility almost all the way back to DOS, and certainly back to systems like Win95 and Win98, which had all the security of "Forget your password? Press esc at the password prompt to login as Administrator."
Apple, on the other hand, completely ditched their OS 9 architecture. Yes, OS 9 apps run on OS X, but only through an emulation layer ("Classic"), just as PPC OS X apps run on Intel OS X only through Rosetta. So, when Apple implemented sudo and a new application bundling system, they had a good shot at getting it right, because they didn't have to support all the quirks of the old programs directly.
That's what Microsoft should have done with Vista: Provide an emulation/virtualization layer, or at least something akin to a chroot, in which to run old apps in. Give it a performance hit -- and actually write a new API. When people port their apps, they do so in such a way that there are minimal "cancel/allow" windows, but for people who haven't ported, there's absolutely no cancel/allow, but they also can't harm the rest of the system.
I though Ubuntu included legal codecs, however even if they don't they could still use Linspire which does have legal codecs.
Linspire is neither free nor open source, which kind of defeats the point.
Yes, I realize codecs aren't free or open, and some drivers aren't open. But there's a difference between going with a distro which is entirely open by default, and adding just enough to make it work, and starting with a distro which is entirely closed.
And yes, Ubuntu does include legal codecs -- but only the codecs which can legally be distributed by Ubuntu as it is. Right now, a lot is supported out of the box, including wmv9, I believe. But a lot isn't, including DVD CSS.
(Frankly, there's hardly anything the community hasn't reverse-engineered, and we're not really afraid of codecs, as those are covered by patents, maybe. But most distributors are terrified of DRM, because that is covered by the DMCA, which people do tend to use and abuse to get what they want -- people with a lot deeper pockets and a lot more lawyers than your average Linux distributor.)
Then again, there's a ready-made, free/open, but maybe not legal, solution: Medibuntu. But that only works once you know what you're doing enough to know about them.
As a guy who knows what he's doing, I like taking a vanilla Kubuntu, stripping out the parts I don't like, adding the stuff I know is good (Medibuntu, nvidia-glx-new, the WineHQ repositories), and ending up with my own system. Just as with Windows, I'd not only start with a fresh XP install disk (not "restore" crap), I'd run it through nLite first. But for a newbie, it should just work out of the box, yet be as close as possible to my system under the hood, so that when they're in trouble, I can help. (I've never used Linspire, but I'm sure I could debug Ubuntu + proprietary codecs.)
With iTunes I can create a play list and burn it to cd. The only limitation is that I can only burn a specific play list 4 tymes. But I can rearrange the songs and bourn them 4 more tymes. Or I can burn the songs to cd then reimport the cd and do whatever I want. You lose a little sound quality, but then again you loose quality digital vs analogue anyway.
First, it's more than a little.
What you're saying is like saying DVD CSS and similar systems do not lock in video because you can always point a camcorder at your TV.
As for digital vs analog, most digital music, in its original form, is high enough quality that humans cannot tell the difference between it and analog. But even if this was not the case, you are dismissing a loss of quality that does not have to be there.
And what would the price point be to eliminate lost sales of hardware? $500? $1000?
Well, do the math yourself, instead of pulling numbers out of your ass.
Because when clones start crashing Apple will start to look bad, "Macs aren't any more stable than Windows PCs".
No, because they would be PCs. Thus, Apple would still have a reputation for making very good hardware -- hardware that OS X runs best on.
More and more OEMs are preinstalling Linux on PCs now, I bought one about a year ago.
Ah, true. But, as I said, they may not be doing it right -- Dell, for instance, ships a stock Ubuntu. Can you imagine if they shipped a stock Windows? People would hate them for it.
At the very least, they should be setting up video drivers. For that matter, they should be licensing some of the legal Linux versions of various codecs...
I'd be curious to know if an OEM has gotten this right, but I kind of doubt it. Most I've seen go to the other extreme -- they practically build their own distro, and everything still doesn't work out of the box.
Just use some sort of Noscript-like Firefox addon. What you're suggesting is like the old trick to disable Flash by renaming the file, and then renaming it back on the few sites you want it -- it's retarded, when there are simple extensions (add-ons) out there which let you control your plug-ins easily.
I don't agree that Virgin should be held responsible here, but at least there's an argument there. There's really no argument for bringing Creative Commons into this.
Face it folks, Linux has little or no serious software applications for any productive desktop use.
Openoffice. KOffice. Kontact. Evolution. Gimp. Kopete. Pidgin. Abiword. Gnumeric. Ardour. Scribus. Eclipse. KDevelop. These just off the top of my head.
Face it, Linux has plenty of serious software applications for productive desktop use. They may not have it for what you consider desktop use, which is really too bad, and if Wine can't run your app, maybe there's something worth implementing here. But to say that it has no apps is to... well... not to have tried.
Or, to say there's no "serious" apps is a bit arrogant. Many people have tried the Linux apps, and prefer the Windows/commercial counterpats -- but many people actually do prefer the Linux/free ones. Ardour is a good example here -- many people seem to think there's no competition for ProTools on Linux. And many people seem to think that there's no competition for Linux' entire sound system (JACK, etc), apps included, on any other platform.
Oh, and by the way -- there are actually some very serious apps for Linux which are not free. Maya, for instance.
Either you can continue bitching about how bad you think Windows is and spend the next decade being the distant 3rd OS in the market, or accept it and get to work to fix it.
This goes for you, too. Either you can continue bitching about how bad you think Linux is and spend the next decade putting up with Microsoft's bullshit, or you give it another shot -- after you deal with the hardware issues, by the way, seeing as Linux has to deal with both legal issues and not being preinstalled / having drivers come on CDs from the manufacturer.
I accept some people cannot work with Linux, and applications are the main reason for that. However, GP was generally not talking about applications, and for the people who can work with Linux, I don't often find people who prefer Windows after having given each a fair shot.
The only lockin Apple has is the OS running only on Apple hardware. Seeing as how Apple is a system's integrator and makes the hardware as well as software, I'm not surprised.
I have no problem with OS X being optimized for Apple hardware, only sold with Apple hardware, and only coming with drivers for apple hardware.
But Apple actually uses a TPM chip to ensure that OS X can only run on Apple hardware.
That's ignoring the whole iTunes thing.
When he was brought back he looked at the numbers and saw that by licensing the OS Apple was losing more in lost hardware sales than they made in licenses, so he ended it.
It would seem to me that the smart thing to do here would be to increase the cost of the OS on non-Apple hardware. Remember, not everyone buying a PC is a lost Mac sale, even if they run OS X on it.
However, as much business sense as it makes, I don't like the ethics of it. I bought a computer, and I want to buy an OS, which could run on it, with no technical problems, except that the OS manufacturer refuses to let me. So they don't get my money.
I bet the single biggest reason why more people don't use Linux is because there aren't many PCs that come with Linux preinstalled. And most people don't install an OS, they just buy a computer from the store, plug everything in and power up. They want it working right out of the box.
Every time I hear this, I wonder if there's any money to be made in the pre-installed Linux market. Dell seems to think so, and they haven't even bothered to do it right.
I'm sorry this post is so long... and rambling. If you'd rather wait, I can try to clean it up later.
However much you may want to make the GNU+Linux issue about naming every contributor, that's not the issue. It's not even about naming the major contributors. It's about properly describing the operating system being used.
Which, the way GNU puts it, is naming every contributor.
The way Linus would put it, it's the kernel. The way most distros would put it, it's the major software that a typical user would consider "essential", including a desktop environment.
From Wikipedia:
An operating system (OS) is the software that manages the sharing of the resources of a computer. An operating system processes raw system data and user input, and responds by allocating and managing tasks and internal system resources as a service to users and programs of the system.
Linux, the kernel, fits that description.
At the foundation of all system software, an operating system performs basic tasks such as controlling and allocating memory, prioritizing system requests, controlling input and output devices, facilitating networking and managing file systems.
Still sounds like the kernel.
Most operating systems come with an application that provides a user interface for managing the operating system, such as a command line interpreter or graphical user interface.
This sounds like bash, or X.org. It's not needed for the operating system -- in fact, it's entirely possible that there are statically linked programs used for booting a modern distro. For quite awhile, Busybox was used in the initial ramdisk, for example.
The operating system forms a platform for other system software and for application software.
Now, since we've established that the Linux kernel, by itself, is sufficient to run applications, provided that they are statically linked, that would make Linux, by itself, the OS. The platform might then be called "GNU/Linux", or it might be called something else.
But as long as it's a platform, that's a much broader definition. KDE, for example, is a platform -- the KDE libraries are massive, and Qt+KDE is theoretically cross-platform, though the latter hasn't been ported to non-Unix systems (yet).
And since it's a platform, you could say, for example, that a default install of Debian -- no GUI, not much except the apt tools, the GNU tools, and init -- is GNU/Linux.
But if we're talking about Ubuntu, that's a much larger platform, involving a lot more components than GNU alone, all of them required.
Personally, I prefer to talk about platforms, because there's less ambiguity. When I talk about operating systems, I generally mean a default install of a Linux distribution, or the kernel itself, but I accept that most of the time, when we say "Operating System", it's pretty vague. Even "kernel" can be vague -- it seems many kernel threads run as sort-of userspace processes anyway, though I'm not entirely sure how that works.
That wouldn't be reasonable because a single init script isn't a major operating system component by any reasonable definition.
Give me one, then.
The GNU project was working on their contribution to todays GNU+Linux systems for 7 years before Linus even started coding. It should be pretty obvious that that's a relevant distinction.
Indeed it is. And Dennis Richie developed C in 1972. Once again, the question remains: What is that magical quality that makes GNU special, and not simply another component?
When someone does that, someone with your "the GNU project deserves no credit" viewpoint will revert the edit because "the GNU project didn't contribute anything that significant - just some command line tools that no-one sees".
There are still some performance issues specific to some hardware
Maybe I haven't been keeping up, but wasn't there a performance issue specific to soundcards and network cards? Not just some, but all? (As in, play music and your network performance drops to 10% -- not by 10%, but to 10%.)
when I can configure a quad-core box with 4GB memory and 1TB storage under $1500 which blazes through my pro audio software with Windows Vista, I don't see any reason why I should spend twice as much (if not more) on a Mac Pro.
Maybe for the eight-core, sixteen gigabytes of RAM, or 3 terabytes of storage?
You get what you pay for...
I spend hours getting it to work with my hardware, and then pretty much don't find much use of it.
Wait, are you talking about Linux or Vista?
It is good to see Linux trying to catch up with Windows on the desktop and I think that it is good for people with limited desktop requirements.
Linux has been ahead of Windows for a long time now, in every way except ease of use and application support. Ease of use is subjective, and largely a result of how familiar something is -- obviously, if you've used Windows all your life, Windows is easier for you to use. Application support is a direct result of which OS is most popular -- a chicken and egg scenario.
It's only very recently that Microsoft has started to catch up in things like security -- by basically ripping off sudo. In fact, just about every improvement you see in Vista was taken from somewhere else.
And if you ever get nostalgic about the good old shell, try the new Windows PowerShell.
Why bother, when I have a perfectly functioning Bash?
But no matter. With Mono, there's a good chance I'll have PowerShell eventually, if I care.
If they want you to use vista, you will, one way or another.
Only if I want to.
Maybe I'm doomed to upgrade from XP, but it will either be a complete negation of my Windows use (I already run Linux 99% of the time), or I'll upgrade to some version after Vista. Maybe another service pack, maybe a whole other OS.
Remember ME? I never had to upgrade to that piece of shit. I just went straight to Linux, nuked my 98 partition until 2K became viable.
Go to Mac? the macintosh doesnt have many games, neither does linux.
Not as many, but still plenty. Also, to many people, WoW is gaming, and that has a native OS X version, and works well under Wine.
By the way, gaming is completely irrelevant to the question of Windows on the desktop. It's corporations that rule the desktop. What they demand, Microsoft delivers. If they want XP, then they will get it. If they want certain things fixed in Vista, they will get it.
You can barely find an OEM that will offer full time support for most non-windows systems, and actually understand, and there being a guarantee that they'll still be around in 5 years.
Ever heard of a little company called "Dell"?
Microsoft may not give a shit about their consumers, but Dell seems to.
the only positive thing is that a UI change may start requiring people to use their brains and figure shit out for once.
No, the UI change is to make them have to use less of your brain. XP didn't seem to make people smarter, did it?
Wrong. The GNU System refers to the complete OS that the GNU project set out to build (and has never formally released). X.org was not developed for the GNU System, but if they GNU System is ever released as an OS distribution it will most likely include X.org.
And when it does, will that system be called GNU/X.org?
Note also that GNU has never released a complete OS. It's easy (maybe too easy) to snipe at RMS for this -- that he's just mad because someone else finished the project that he never seemed to be planning to.
Maybe it would help to look at it this way: Red Hat Enterprise Linux is based on the GNU System in the same way that Mandriva Linux is based on Red Hat. You'll note that in the Wikipedia Article for Mandriva Linux, the fact that it's based on Red Hat is mentioned in the first paragraph. In contrast, the GNU System / GNU Project isn't mentioned in the Red Hat Enterprise Linux article at all. It's that sort of mention that this whole GNU+Linux debate is about - it would seem reasonable for the GNU Project to get credit at least where it's *obviously* appropriate.
Fine, then, edit Wikipedia pages. But stop trying to edit our language.
In any case, either you're demanding that X.org be included in the name of Unix-like operating systems with GUIs or you're not. Earlier you said you aren't - have you changed your mind?
I'm not demanding anything. I'm simply using X.org to illustrate my point: If you want me to use GNU/Linux, you must recognize X.org (among many other components) as something else that goes in that string. Thus, it would be GNU/Linux/X.org.
And if you continue that for too long, you realize that while everyone would get credit, it would simply be too long a name to put Ubuntu/GNU/Linux/X.org/KDE/Freedesktop/... all in the name.
Personally, I would rather just use "Linux", or "Ubuntu", or "KDE", or whatever I feel is relevant at the time. As an example: Often I will mention that I run a particular game under Wine. In that sentence, of course, I completely neglect to mention that Wine runs on X.org and OpenGL (but could run on the commandline instead), and is running on Linux, almost certainly linked against glibc, was compiled with gcc, and so on -- because none of it is relevant.
It's a bit like the whole hacker/cracker problem. It may make me twitch every time I hear someone talking about "hacking" into something, or how someone in a game is "hacking", but really, I know what they mean. I'd rather communicate than not communicate, and launching into a rant about why they should call it "cracking" or "cheating" (respectively), and what "hacking" really means, is a surefire way not to communicate.
Sure thing. I can't think of any major "Linux" apps that are shipped that way though.
I can think of a few -- most commercial video games, for instance, are either shipped statically linked for Linux, or as close as they can. Usually, this means they link against the system's OpenGL and their own SDL, and compile everything else in that they can. Many commercial apps take this approach, since it's the only way to guarantee that they have the libraries they need on systems that don't have good package management (Windows/OS X). And it is kind of convenient -- it means, for example, I can install a 32-bit version on a 64-bit system and expect it to work, usually without needing to install any libraries manually. (Compare this to something like Firefox.)
A representative of the GNU project has requested that Red Hat give the GNU Project credit.
Except that's not what this debate has been about. This debate has been about whether Red Hat is willing to change their name as a way of giving them credit.
X.org is offtopic because A.) they haven't asked for credit and B.) you've already stated that you aren't demandi
Because all general purpose "Linux distributions" are the GNU System with a single component swapped out.
And several other components added. Sorry, X.org is NOT part of the GNU System. Neither, for that matter, is any of KDE. From the user perspective, these are parts of the OS, yet you either exclude them from mention, or call them a part of the GNU System (as you did with X.org).
RMS "crys foul" for precisely the same reason that the NetBSD people would think I was a dick if I built a version of NetBSD with a new default shell
Several issues here:
Linux came along before HURD actually worked. So that would be more like if they'd managed to develop an OS without a shell, and you came along and wrote Bash.
That's obviously not enough, but you have to consider -- the kernel, being the monolithic beast that it is, is much more than the command interpreter. It includes and maintains every filesystem Linux supports, every driver, along with the scheduler, security model, boot process, and so on.
It's not "just a kernel" any more than GNU is "just a bunch of tools" or "just an operating system". Both are large projects, involving lots of resources.
and spread a story around about how "Chandon Seldon Developed ChandOS, a complete operating system, in 2007 because he was bored".
You keep insinuating that Linus spread the story of Linux. He didn't. Nor did he attempt to steal the credit of others.
His original name for Linux was "Freeax". It was originally intended to be a simple terminal emulator. Then he wanted it to be able to download, so he had to add support for a filesystem, and enough else that he decided to just make it a POSIX-compliant kernel -- but it was still intended to be merely a stopgap measure until HURD was released, which was supposed to be in a year.
The only reason it is still called "Linux" today was, when someone else was giving Linus an FTP directory (some free hosting), for whatever reason, the other guy decided to call it "Linux", after Linus' somewhat embarrassing pet name.
It also wasn't merely because he was bored. The terminal emulator part was created out of necessity -- basically, being a college student, he'd rather write some software than go buy a piece of hardware.
If you don't combine it with the GNU system, you have either nothing or a reasonably bare-bones embedded OS that can't even run dynamically linked binaries, much less any "Linux Applications".
You do know there are statically linked applications, right?
And you do know that glibc isn't the only libc for Linux, right? And that gcc isn't the only compiler? (Even if it was, a compiler is hardly an excuse for demanding that naming -- BSD wasn't called GNU/BSD before some variants started switching to pcc.)
Of course, let's say we only include GNU/Linux -- well, then, you have Emacs, but no Vi. You might have a mail client. (Neither of which, I might add, are installed on my "GNU/Linux" system.) And you have absolutely no GUI -- I'm not just talking about X.org, I'm talking about most of the applications run on top of X.org -- I don't run GNOME. The closest you might have is Links, which, it could be argued, is an application, not an OS component.
Not only that, but even GNOME isn't technically GNU -- it's now maintained by something called the "GNOME Foundation."
The refusal of RMS (and you) to accept the name "Linux" is truly petty. What's more, it's a refusal to accept that you are now part of a larger whole. Yes, you are an essential part of Ubuntu, and of most Linux distributions (not all), but you are one of many essential parts, and you don't make up the majority anymore, neither in disk space nor in lines of code.
Yes, you deserve to be recognized!
But so do so many others that there is no name that could possibly be fair to all of them.
Yes, you are part of the operating system!
But the term is so loosely defined in the first place that you cannot possibly claim to be the complete operating system (short of the kernel).
And you refuse to see lines where they do exist, or try to place them in nonsensical places to prove your point.
Can't dispute that, because you've actually said nothing here. What constitutes "nonsensical"?
You can feel free to talk about NetBSD/Xorg all you want - I'll accept the convention and keep calling it NetBSD.
Yet you don't accept the convention regarding Linux -- you insist on GNU/Linux.
My point about Xorg is not that I wish it to be called NetBSD/Xorg. My point is that it isn't called NetBSD/Xorg, it's called NetBSD, and the Xorg people don't cry foul. Why, then, does GNU cry foul about Linux?
He used something along the lines of "amateur sleuth". And when he used "nerd", it was meant as a bad word -- do you call a mechanic a "motorhead"? No, you call him a mechanic, and if he's designing cars as well as fixing them, you call him an engineer. In other words, you call him a fucking expert.
Java gaming? maybe, sure it is cross platform, but your app is horribly VM limited and performance will sucky no matter how you tweak.
Java itself is open source, so you can tweak it if you have a problem.
Besides, this has been discussed to death, and I seem to remember the conclusion is that Java can be fast, and that most of the reasons people don't like it for game development are purely academic. For example, complaining about a "stop-the-world" garbage collector -- I believe newer VMs use a generational gc, and even if they didn't, it's still not a significant impact -- it's stopping the world for less than a frame.
frozen-bubble keeps getting revived but in the end is not compileable with newer versions of SDL_perl.
So what's wrong with using an old version?
the better the toolkit, the longer lived the project - look at the old quake engines...
Yeah, let's go back to Quake, where they had to invent their own fucking C! Or maybe something like Unreal Tournament, which, to its latest version, has developed its own embedded scripting language.
The thing is, you're talking about reinventing the wheel for no apparent reason, and you seem to have arbitrarily selected which toolkits a project should use, and which it shouldn't. Yes, scons will be the death of the project! (Except it's just a build system, and Make isn't that good.)
Yeah, let's throw out Python and develop our own scripting language instead! Or better, use straight C!
I just love tracking down memory leaks. Let's do our own memory management, for no reason!
But, conversely, let's use GTK+ instead of rolling our own UI entirely in GL!
Look, I realize there are toolkits out there that suck. But that's no reason to throw the idea of a toolkit out, and just go with as bare-bones as you can get. And frankly, I'd be much more encouraged to develop a game that, say, embedded Python -- a game where I could actually have an eval loop when I hit tilde, and control much of the game from there.
I don't think the controls are what makes it hard. It's an interesting control method, sure, but once you get used to it, no harder than an FPS...
Consider an FPS, actually. You could have a game with really "easy" controls, like an auto-aim. We call that an aimbot, aka, CHEATING.
Or, you learn the controls you're given, and you find that it's not just the controls that make the game hard. It's the game itself -- in Neverball, it's the physics of momentum in a world where it's easy to fall off the edge.
By the way -- this is hardly new. A LOT of old games have much more frustrating gameplay. In general, it's arcade-style games anyway -- the idea, I guess, being that you'd lose, but you'd be SO close that you'd have to try again, so you'd put in more quarters. But it is actually fun -- the flipside of it being frustrating is, it's that much more rewarding to finally beat it.
I don't like the trend these days towards easier games, by the way -- towards games which have an "I can win" setting (even if I like that setting), which have like 5-10 hours of gameplay to make sure everyone can finish it (even if they don't really have the time to be playing games anyway), to the point where it's just not worth it to buy a game anymore, if you can rent a console game and finish it before you have to take it back.
If there's the option to do more stuff, you have to be very careful to design a UI which will lead ordinary users away from choosing that option. Worse are the "power users", who will try to explore every single option, whether or not they understand it.
It's a deliberate break from the "GNOME or KDE" question you get asked at install time. If you don't know and don't care, you get GNOME. If you know, there's always a way to get more choice -- you could download Kubuntu if you want KDE, for example.
And if you really know what you're doing, you could download the Alternate or Server install CD, install from that (a more powerful installer anyway), and add packages as you need them. Or you can do a normal install, and remove packages, add other ones in later.
In fact, I believe it's possible to "upgrade" a system between Ubuntu and Kubuntu and back.
Now, granted, maybe it would be a good thing to have all of this power as a convenient GUI option. But the choice is there, if you know where to look.
The name of the distribution is "Nexenta OS". The OS family is "GNU/Solaris". The name of the OS I'm running is "Ubuntu". The family is "GNU/Linux".
Splitting hairs. Especially considering only the GNU people are convinced that GNU deserves special mention as part of the operating system.
Given that what their distributing is a GNU-class Unix-like operating system, that seems rather impolite.
See, that's where we disagree.
You look at Ubuntu, and you see it as a GNU operating system. I see it as something which I call "Linux", but which, if I deconstruct it, GNU is really only a very small part of it.
Your best response so far seems to be drawing lines where none exist. Saying "This bit over here is an essential part of the operating system, and this isn't." Or saying "It's really the GNU operating system, but it uses X.org."
You still haven't given me a good reason, other than convention (as adopted by other systems), to explain why X.org should not be included in the name. And where that convention is followed, GNU usually isn't mentioned either.
Also, look at this page -- the longer description of Ubuntu doesn't include a mention of Linux, GNU, or any other project.
One final point: When you say "GNU operating system" and don't mention Linux, people don't know what you mean. At best, they might know about HURD, and assume you're running that. But when you say "Linux" or a "Linux distro", no one is confused. The worst that happens is you offend RMS.
While I love KDE, GTK is 'free'er and I have to respect that. If I had to pick a graphics toolkit for commercial use, it wouldn't be Qt.
If it was a one-man killer app kind of project, yeah, I'd go for LGPL.
If it was open, I'd probably use Qt. Conversely, if it was a large, commercial project, I'd probably use Qt, just to have the commercial support from Trolltech, and also because I figure the more people actually pay for Qt, the more they can afford to improve it, proprietary and open versions.
Perhaps you've forgotten, but this comment was referring to this project, whose whole point is exactly what I was talking about. They tend to be a bit more programmable, I'd imagine, so they don't have to have as many hardware designs (and so they can re-use existing hardware).
But you can take your objections up with them -- I was only pointing out that saying we have open drivers is not answering the original question here, which was whether the OpenCores stuff would be used.
Allow one player to host, and allow it to fail over to another player. Use a central server only to track the state of who's in the game, who's the server, etc.
Problem solved, and it helps a bit with the cheating.
If it took me all of thirty seconds to come up with that, why didn't Blizzard just do it that way, instead of making it hell for anyone behind a NAT? And this will only become more of a problem in the future; I have seen ISPs throw all their clients behind one massive NAT gateway, rather than, say, going to IPv6.
Slashdot Mirror
While it does mean you lose those particular tracks, the mentality I keep hearing from people I would expect to know better is, in a world where everyone has enough bandwidth to stream radio 24/7, nobody cares that you've lost your music collection.
You just switch to a different, competing service, and re-download everything.
The guy I had this conversation with reasoned it like this: If you're going legit, this is the cheapest way. You lose the ability to have stuff work on an iPod, but he had something else anyway. Everything he wanted to do with it, the DRM software let him do -- except play it on Linux, which he didn't want anyway (partly because it didn't work on Linux -- chicken and egg).
And the economics of it: He calculated that he'd have to subscribe to this service for 15 years straight before he'd be spending more than it would cost to buy the stuff outright on iTunes or CD. And that was just counting songs he'd already downladed -- obviously, in 15 years, he'd be downloading a lot more stuff.
Me, I'm not willing to give up my freedom like that, and stuff just has to work on Linux. Besides, I listen to a lot of Internet radio. But content as a service really isn't a problem. Software as a service, maybe, because you have your own data attached to it, but music? Who are we kidding?
First, it's actually better than it was. There used to be a separate root password, and you had to enter it every time. Now, you just enter your normal user password (via sudo), and it remembers it for some 5-10 minutes.
Second, it's not needed for everyday tasks. It's needed for messing with global system settings -- things like your screen resolution -- or for installing software. Most people don't install software every day. (Although it might get a bit annoying doing it for updates...)
Third, it can be turned off, temporarily or completely, or customized in various way. Technically, I don't know what Vista can do here -- I imagine "just turn it off" is an option.
Compare this to Vista. Some software pops up the prompt constantly, sometimes it only pops up intermittently. The vast majority of the times it pops up, there's really nothing special happening. And, to my knowledge, there's nothing like the "remember for 5 minutes" feature of Linux.
Granted, the gap isn't as big as it once was, but the reason we make fun of cancel/allow is not that they ripped off sudo, but that they ripped it off and got it so horribly wrong. Compare what Microsoft has done to what Apple has done, faced with the same problem: Microsoft has essentially kept backwards-compatibility almost all the way back to DOS, and certainly back to systems like Win95 and Win98, which had all the security of "Forget your password? Press esc at the password prompt to login as Administrator."
Apple, on the other hand, completely ditched their OS 9 architecture. Yes, OS 9 apps run on OS X, but only through an emulation layer ("Classic"), just as PPC OS X apps run on Intel OS X only through Rosetta. So, when Apple implemented sudo and a new application bundling system, they had a good shot at getting it right, because they didn't have to support all the quirks of the old programs directly.
That's what Microsoft should have done with Vista: Provide an emulation/virtualization layer, or at least something akin to a chroot, in which to run old apps in. Give it a performance hit -- and actually write a new API. When people port their apps, they do so in such a way that there are minimal "cancel/allow" windows, but for people who haven't ported, there's absolutely no cancel/allow, but they also can't harm the rest of the system.
Maybe they should license a Wine variant?
Linspire is neither free nor open source, which kind of defeats the point.
Yes, I realize codecs aren't free or open, and some drivers aren't open. But there's a difference between going with a distro which is entirely open by default, and adding just enough to make it work, and starting with a distro which is entirely closed.
And yes, Ubuntu does include legal codecs -- but only the codecs which can legally be distributed by Ubuntu as it is. Right now, a lot is supported out of the box, including wmv9, I believe. But a lot isn't, including DVD CSS.
(Frankly, there's hardly anything the community hasn't reverse-engineered, and we're not really afraid of codecs, as those are covered by patents, maybe. But most distributors are terrified of DRM, because that is covered by the DMCA, which people do tend to use and abuse to get what they want -- people with a lot deeper pockets and a lot more lawyers than your average Linux distributor.)
Then again, there's a ready-made, free/open, but maybe not legal, solution: Medibuntu. But that only works once you know what you're doing enough to know about them.
As a guy who knows what he's doing, I like taking a vanilla Kubuntu, stripping out the parts I don't like, adding the stuff I know is good (Medibuntu, nvidia-glx-new, the WineHQ repositories), and ending up with my own system. Just as with Windows, I'd not only start with a fresh XP install disk (not "restore" crap), I'd run it through nLite first. But for a newbie, it should just work out of the box, yet be as close as possible to my system under the hood, so that when they're in trouble, I can help. (I've never used Linspire, but I'm sure I could debug Ubuntu + proprietary codecs.)
First, it's more than a little.
What you're saying is like saying DVD CSS and similar systems do not lock in video because you can always point a camcorder at your TV.
As for digital vs analog, most digital music, in its original form, is high enough quality that humans cannot tell the difference between it and analog. But even if this was not the case, you are dismissing a loss of quality that does not have to be there.
Well, do the math yourself, instead of pulling numbers out of your ass.
No, because they would be PCs. Thus, Apple would still have a reputation for making very good hardware -- hardware that OS X runs best on.
Ah, true. But, as I said, they may not be doing it right -- Dell, for instance, ships a stock Ubuntu. Can you imagine if they shipped a stock Windows? People would hate them for it.
At the very least, they should be setting up video drivers. For that matter, they should be licensing some of the legal Linux versions of various codecs...
I'd be curious to know if an OEM has gotten this right, but I kind of doubt it. Most I've seen go to the other extreme -- they practically build their own distro, and everything still doesn't work out of the box.
Just use some sort of Noscript-like Firefox addon. What you're suggesting is like the old trick to disable Flash by renaming the file, and then renaming it back on the few sites you want it -- it's retarded, when there are simple extensions (add-ons) out there which let you control your plug-ins easily.
And this involves Creative Commons itself, how?
I don't agree that Virgin should be held responsible here, but at least there's an argument there. There's really no argument for bringing Creative Commons into this.
Openoffice. KOffice. Kontact. Evolution. Gimp. Kopete. Pidgin. Abiword. Gnumeric. Ardour. Scribus. Eclipse. KDevelop. These just off the top of my head.
Face it, Linux has plenty of serious software applications for productive desktop use. They may not have it for what you consider desktop use, which is really too bad, and if Wine can't run your app, maybe there's something worth implementing here. But to say that it has no apps is to... well... not to have tried.
Or, to say there's no "serious" apps is a bit arrogant. Many people have tried the Linux apps, and prefer the Windows/commercial counterpats -- but many people actually do prefer the Linux/free ones. Ardour is a good example here -- many people seem to think there's no competition for ProTools on Linux. And many people seem to think that there's no competition for Linux' entire sound system (JACK, etc), apps included, on any other platform.
Oh, and by the way -- there are actually some very serious apps for Linux which are not free. Maya, for instance.
This goes for you, too. Either you can continue bitching about how bad you think Linux is and spend the next decade putting up with Microsoft's bullshit, or you give it another shot -- after you deal with the hardware issues, by the way, seeing as Linux has to deal with both legal issues and not being preinstalled / having drivers come on CDs from the manufacturer.
I accept some people cannot work with Linux, and applications are the main reason for that. However, GP was generally not talking about applications, and for the people who can work with Linux, I don't often find people who prefer Windows after having given each a fair shot.
I have no problem with OS X being optimized for Apple hardware, only sold with Apple hardware, and only coming with drivers for apple hardware.
But Apple actually uses a TPM chip to ensure that OS X can only run on Apple hardware.
That's ignoring the whole iTunes thing.
It would seem to me that the smart thing to do here would be to increase the cost of the OS on non-Apple hardware. Remember, not everyone buying a PC is a lost Mac sale, even if they run OS X on it.
However, as much business sense as it makes, I don't like the ethics of it. I bought a computer, and I want to buy an OS, which could run on it, with no technical problems, except that the OS manufacturer refuses to let me. So they don't get my money.
Every time I hear this, I wonder if there's any money to be made in the pre-installed Linux market. Dell seems to think so, and they haven't even bothered to do it right.
I'm sorry this post is so long... and rambling. If you'd rather wait, I can try to clean it up later.
Which, the way GNU puts it, is naming every contributor.
The way Linus would put it, it's the kernel. The way most distros would put it, it's the major software that a typical user would consider "essential", including a desktop environment.
From Wikipedia:
Linux, the kernel, fits that description.
Still sounds like the kernel.
This sounds like bash, or X.org. It's not needed for the operating system -- in fact, it's entirely possible that there are statically linked programs used for booting a modern distro. For quite awhile, Busybox was used in the initial ramdisk, for example.
Now, since we've established that the Linux kernel, by itself, is sufficient to run applications, provided that they are statically linked, that would make Linux, by itself, the OS. The platform might then be called "GNU/Linux", or it might be called something else.
But as long as it's a platform, that's a much broader definition. KDE, for example, is a platform -- the KDE libraries are massive, and Qt+KDE is theoretically cross-platform, though the latter hasn't been ported to non-Unix systems (yet).
And since it's a platform, you could say, for example, that a default install of Debian -- no GUI, not much except the apt tools, the GNU tools, and init -- is GNU/Linux.
But if we're talking about Ubuntu, that's a much larger platform, involving a lot more components than GNU alone, all of them required.
Personally, I prefer to talk about platforms, because there's less ambiguity. When I talk about operating systems, I generally mean a default install of a Linux distribution, or the kernel itself, but I accept that most of the time, when we say "Operating System", it's pretty vague. Even "kernel" can be vague -- it seems many kernel threads run as sort-of userspace processes anyway, though I'm not entirely sure how that works.
Give me one, then.
Indeed it is. And Dennis Richie developed C in 1972. Once again, the question remains: What is that magical quality that makes GNU special, and not simply another component?
Maybe I haven't been keeping up, but wasn't there a performance issue specific to soundcards and network cards? Not just some, but all? (As in, play music and your network performance drops to 10% -- not by 10%, but to 10%.)
Maybe for the eight-core, sixteen gigabytes of RAM, or 3 terabytes of storage?
You get what you pay for...
Wait, are you talking about Linux or Vista?
Linux has been ahead of Windows for a long time now, in every way except ease of use and application support. Ease of use is subjective, and largely a result of how familiar something is -- obviously, if you've used Windows all your life, Windows is easier for you to use. Application support is a direct result of which OS is most popular -- a chicken and egg scenario.
It's only very recently that Microsoft has started to catch up in things like security -- by basically ripping off sudo. In fact, just about every improvement you see in Vista was taken from somewhere else.
Why bother, when I have a perfectly functioning Bash?
But no matter. With Mono, there's a good chance I'll have PowerShell eventually, if I care.
I'm curious why people choose to leave Microsoft for a company which does even more DRM and lock-in...
Maybe because their stuff works?
But it does raise the question -- why not Linux? It'd be much cheaper.
Only if I want to.
Maybe I'm doomed to upgrade from XP, but it will either be a complete negation of my Windows use (I already run Linux 99% of the time), or I'll upgrade to some version after Vista. Maybe another service pack, maybe a whole other OS.
Remember ME? I never had to upgrade to that piece of shit. I just went straight to Linux, nuked my 98 partition until 2K became viable.
Not as many, but still plenty. Also, to many people, WoW is gaming, and that has a native OS X version, and works well under Wine.
By the way, gaming is completely irrelevant to the question of Windows on the desktop. It's corporations that rule the desktop. What they demand, Microsoft delivers. If they want XP, then they will get it. If they want certain things fixed in Vista, they will get it.
Ever heard of a little company called "Dell"?
Microsoft may not give a shit about their consumers, but Dell seems to.
No, the UI change is to make them have to use less of your brain. XP didn't seem to make people smarter, did it?
And when it does, will that system be called GNU/X.org?
Note also that GNU has never released a complete OS. It's easy (maybe too easy) to snipe at RMS for this -- that he's just mad because someone else finished the project that he never seemed to be planning to.
Fine, then, edit Wikipedia pages. But stop trying to edit our language.
I'm not demanding anything. I'm simply using X.org to illustrate my point: If you want me to use GNU/Linux, you must recognize X.org (among many other components) as something else that goes in that string. Thus, it would be GNU/Linux/X.org.
And if you continue that for too long, you realize that while everyone would get credit, it would simply be too long a name to put Ubuntu/GNU/Linux/X.org/KDE/Freedesktop/... all in the name.
Personally, I would rather just use "Linux", or "Ubuntu", or "KDE", or whatever I feel is relevant at the time. As an example: Often I will mention that I run a particular game under Wine. In that sentence, of course, I completely neglect to mention that Wine runs on X.org and OpenGL (but could run on the commandline instead), and is running on Linux, almost certainly linked against glibc, was compiled with gcc, and so on -- because none of it is relevant.
It's a bit like the whole hacker/cracker problem. It may make me twitch every time I hear someone talking about "hacking" into something, or how someone in a game is "hacking", but really, I know what they mean. I'd rather communicate than not communicate, and launching into a rant about why they should call it "cracking" or "cheating" (respectively), and what "hacking" really means, is a surefire way not to communicate.
I can think of a few -- most commercial video games, for instance, are either shipped statically linked for Linux, or as close as they can. Usually, this means they link against the system's OpenGL and their own SDL, and compile everything else in that they can. Many commercial apps take this approach, since it's the only way to guarantee that they have the libraries they need on systems that don't have good package management (Windows/OS X). And it is kind of convenient -- it means, for example, I can install a 32-bit version on a 64-bit system and expect it to work, usually without needing to install any libraries manually. (Compare this to something like Firefox.)
Except that's not what this debate has been about. This debate has been about whether Red Hat is willing to change their name as a way of giving them credit.
Let me deconstruct this, one by one:
And several other components added. Sorry, X.org is NOT part of the GNU System. Neither, for that matter, is any of KDE. From the user perspective, these are parts of the OS, yet you either exclude them from mention, or call them a part of the GNU System (as you did with X.org).
Several issues here:
Linux came along before HURD actually worked. So that would be more like if they'd managed to develop an OS without a shell, and you came along and wrote Bash.
That's obviously not enough, but you have to consider -- the kernel, being the monolithic beast that it is, is much more than the command interpreter. It includes and maintains every filesystem Linux supports, every driver, along with the scheduler, security model, boot process, and so on.
It's not "just a kernel" any more than GNU is "just a bunch of tools" or "just an operating system". Both are large projects, involving lots of resources.
You keep insinuating that Linus spread the story of Linux. He didn't. Nor did he attempt to steal the credit of others.
His original name for Linux was "Freeax". It was originally intended to be a simple terminal emulator. Then he wanted it to be able to download, so he had to add support for a filesystem, and enough else that he decided to just make it a POSIX-compliant kernel -- but it was still intended to be merely a stopgap measure until HURD was released, which was supposed to be in a year.
The only reason it is still called "Linux" today was, when someone else was giving Linus an FTP directory (some free hosting), for whatever reason, the other guy decided to call it "Linux", after Linus' somewhat embarrassing pet name.
It also wasn't merely because he was bored. The terminal emulator part was created out of necessity -- basically, being a college student, he'd rather write some software than go buy a piece of hardware.
You do know there are statically linked applications, right?
And you do know that glibc isn't the only libc for Linux, right? And that gcc isn't the only compiler? (Even if it was, a compiler is hardly an excuse for demanding that naming -- BSD wasn't called GNU/BSD before some variants started switching to pcc.)
Of course, let's say we only include GNU/Linux -- well, then, you have Emacs, but no Vi. You might have a mail client. (Neither of which, I might add, are installed on my "GNU/Linux" system.) And you have absolutely no GUI -- I'm not just talking about X.org, I'm talking about most of the applications run on top of X.org -- I don't run GNOME. The closest you might have is Links, which, it could be argued, is an application, not an OS component.
Not only that, but even GNOME isn't technically GNU -- it's now maintained by something called the "GNOME Foundation."
The refusal of RMS (and you) to accept the name "Linux" is truly petty. What's more, it's a refusal to accept that you are now part of a larger whole. Yes, you are an essential part of Ubuntu, and of most Linux distributions (not all), but you are one of many essential parts, and you don't make up the majority anymore, neither in disk space nor in lines of code.
Yes, you deserve to be recognized!
But so do so many others that there is no name that could possibly be fair to all of them.
Yes, you are part of the operating system!
But the term is so loosely defined in the first place that you cannot possibly claim to be the complete operating system (short of the kernel).
I was switching between them, but nice to know they'll all actually coexist.
Demonstrate it, then.
Can't dispute that, because you've actually said nothing here. What constitutes "nonsensical"?
Yet you don't accept the convention regarding Linux -- you insist on GNU/Linux.
My point about Xorg is not that I wish it to be called NetBSD/Xorg. My point is that it isn't called NetBSD/Xorg, it's called NetBSD, and the Xorg people don't cry foul. Why, then, does GNU cry foul about Linux?
He used something along the lines of "amateur sleuth". And when he used "nerd", it was meant as a bad word -- do you call a mechanic a "motorhead"? No, you call him a mechanic, and if he's designing cars as well as fixing them, you call him an engineer. In other words, you call him a fucking expert.
Java itself is open source, so you can tweak it if you have a problem.
Besides, this has been discussed to death, and I seem to remember the conclusion is that Java can be fast, and that most of the reasons people don't like it for game development are purely academic. For example, complaining about a "stop-the-world" garbage collector -- I believe newer VMs use a generational gc, and even if they didn't, it's still not a significant impact -- it's stopping the world for less than a frame.
So what's wrong with using an old version?
Yeah, let's go back to Quake, where they had to invent their own fucking C! Or maybe something like Unreal Tournament, which, to its latest version, has developed its own embedded scripting language.
The thing is, you're talking about reinventing the wheel for no apparent reason, and you seem to have arbitrarily selected which toolkits a project should use, and which it shouldn't. Yes, scons will be the death of the project! (Except it's just a build system, and Make isn't that good.)
Yeah, let's throw out Python and develop our own scripting language instead! Or better, use straight C!
I just love tracking down memory leaks. Let's do our own memory management, for no reason!
But, conversely, let's use GTK+ instead of rolling our own UI entirely in GL!
Look, I realize there are toolkits out there that suck. But that's no reason to throw the idea of a toolkit out, and just go with as bare-bones as you can get. And frankly, I'd be much more encouraged to develop a game that, say, embedded Python -- a game where I could actually have an eval loop when I hit tilde, and control much of the game from there.
I don't think the controls are what makes it hard. It's an interesting control method, sure, but once you get used to it, no harder than an FPS...
Consider an FPS, actually. You could have a game with really "easy" controls, like an auto-aim. We call that an aimbot, aka, CHEATING.
Or, you learn the controls you're given, and you find that it's not just the controls that make the game hard. It's the game itself -- in Neverball, it's the physics of momentum in a world where it's easy to fall off the edge.
By the way -- this is hardly new. A LOT of old games have much more frustrating gameplay. In general, it's arcade-style games anyway -- the idea, I guess, being that you'd lose, but you'd be SO close that you'd have to try again, so you'd put in more quarters. But it is actually fun -- the flipside of it being frustrating is, it's that much more rewarding to finally beat it.
I don't like the trend these days towards easier games, by the way -- towards games which have an "I can win" setting (even if I like that setting), which have like 5-10 hours of gameplay to make sure everyone can finish it (even if they don't really have the time to be playing games anyway), to the point where it's just not worth it to buy a game anymore, if you can rent a console game and finish it before you have to take it back.
If there's the option to do more stuff, you have to be very careful to design a UI which will lead ordinary users away from choosing that option. Worse are the "power users", who will try to explore every single option, whether or not they understand it.
It's a deliberate break from the "GNOME or KDE" question you get asked at install time. If you don't know and don't care, you get GNOME. If you know, there's always a way to get more choice -- you could download Kubuntu if you want KDE, for example.
And if you really know what you're doing, you could download the Alternate or Server install CD, install from that (a more powerful installer anyway), and add packages as you need them. Or you can do a normal install, and remove packages, add other ones in later.
In fact, I believe it's possible to "upgrade" a system between Ubuntu and Kubuntu and back.
Now, granted, maybe it would be a good thing to have all of this power as a convenient GUI option. But the choice is there, if you know where to look.
Splitting hairs. Especially considering only the GNU people are convinced that GNU deserves special mention as part of the operating system.
See, that's where we disagree.
You look at Ubuntu, and you see it as a GNU operating system. I see it as something which I call "Linux", but which, if I deconstruct it, GNU is really only a very small part of it.
Your best response so far seems to be drawing lines where none exist. Saying "This bit over here is an essential part of the operating system, and this isn't." Or saying "It's really the GNU operating system, but it uses X.org."
You still haven't given me a good reason, other than convention (as adopted by other systems), to explain why X.org should not be included in the name. And where that convention is followed, GNU usually isn't mentioned either.
Also, look at this page -- the longer description of Ubuntu doesn't include a mention of Linux, GNU, or any other project.
One final point: When you say "GNU operating system" and don't mention Linux, people don't know what you mean. At best, they might know about HURD, and assume you're running that. But when you say "Linux" or a "Linux distro", no one is confused. The worst that happens is you offend RMS.
If it was a one-man killer app kind of project, yeah, I'd go for LGPL.
If it was open, I'd probably use Qt. Conversely, if it was a large, commercial project, I'd probably use Qt, just to have the commercial support from Trolltech, and also because I figure the more people actually pay for Qt, the more they can afford to improve it, proprietary and open versions.
Perhaps you've forgotten, but this comment was referring to this project, whose whole point is exactly what I was talking about. They tend to be a bit more programmable, I'd imagine, so they don't have to have as many hardware designs (and so they can re-use existing hardware).
But you can take your objections up with them -- I was only pointing out that saying we have open drivers is not answering the original question here, which was whether the OpenCores stuff would be used.
Allow one player to host, and allow it to fail over to another player. Use a central server only to track the state of who's in the game, who's the server, etc.
Problem solved, and it helps a bit with the cheating.
If it took me all of thirty seconds to come up with that, why didn't Blizzard just do it that way, instead of making it hell for anyone behind a NAT? And this will only become more of a problem in the future; I have seen ISPs throw all their clients behind one massive NAT gateway, rather than, say, going to IPv6.
...which would be why they called it "Nexenta OS", and not "GNU/Solaris"?
I'm aware of the domain name, but I somehow doubt that would be enough for RMS.