Screens are getting better, and a new generation doesn't care much whether it's a book or on a screen, so long as it's a decent screen. e-ink will change all of this, even for older/pickier people.
Again, for me to pirate you a movie, I have to be able to give it to you. Sure, I could mail it or something, but that's more work and expense for me. That fact means it's largely limited to me pirating things for my close friends, family or neighbors.
With what a blank tape cost, and what a movie cost, this could still be reasonably done. It's more illegal, and it's easier to track down, but nothing like this has killed the movie industry. VCRs have, in fact, strengthened it.
Even without that, the movie industry's track record with VCRs, and the music industry with DATs, suggests that either they are completely incompetent at recognizing whether a technology will kill them (or how to make money off that technology), or that they are using piracy as an excuse to gain lockin for an ulterior motive. For instance, the fact that DVDs scratch fairly easily, and that the movie industry theory is that you should only watch the same exact disc, right off the disc, suggests that they really want DRM on DVDs, not because of piracy, but to force us to buy the same thing over and over again.
For that matter, buying the same thing over and over again is an RIAA/MPAA pattern.
I agree about your AllofMP3 assessment, though. I prefer lossless or Vorbis if I'm going to buy something, but the main criteria is cheapness and lack of DRM.
You are repeating the same misconceptions in the same ways. Look at your language:
I disagree with the whole idea that it's unnecessary to protect the works of content creators.
No one is arguing "that it's unnecessary to protect the works of content creators." We're arguing that DRM, specifically, in any form, is not worth the harm it causes, and that content creators can make a profit without it. After all, they did before DRM existed.
DRM is a reality and to deny this is to be simply ignorant of current trends in media playback software/hardware stacks. All new hardware from major manufacturers will support DRM standards. If the data stream is protected, the media appliance will acknowledge and honor the DRM lock and you will be unable to do more with the content than is allowed by the DRM lock. This is reality and it is already here.
If you are astroturf, you need to listen here: If that is really and truly the reality, I will wean myself off modern media. I simply refuse to spend any money on anything that has unreasonable "protection" on it. If there are enough of us, you will lose money on DRM.
That is why I refuse to buy anything Blu-Ray until I am convinced that it's permanently cracked.
If not, I simply don't care enough. There is enough entertainment in the world that comes without strings attached.
What we want to do is make sure that things like machine-local data can be transmitted from one machine to another (deleting the original data as it moves to the next device) are preserved while things like forward-lock (which prevents copying at all) are eliminated.
Current DRM models have two problems: In order to enforce any kind of protection, they require specific software/hardware stacks, which reduces user choice -- for instance, it becomes essentially impossible to have a proper open-source media center, or even to run a closed media center on an open OS.
The second problem is, much of it is online. For instance, the music subscription services -- pay $x/mo and get as much music as you can download, but if you stop paying, they stop playing. Another example is Steam: You only pay once, but it insists on connecting to the Internet periodically to get updates and to be able to shut you down if they find two copies from the same purchase online at once. The problem with this is, I'm essentially trusting the content provider not to unfairly revoke my right to use my content -- Valve could one day decide not to let me play at all, or their servers could go down, and I'd be stuck without a game.
This puts things entirely too much out of control of the consumer, who, in a very real sense, no longer owns their stuff. Think of it this way -- the rights to a book are owned by the author, and only licensed to a publisher for a finite amount of time. If you buy a book, you own that copy, and may do whatever you want with it, other than distribute copies of significant portions of the book. Yet I never hear authors screaming about how they're being completely ripped off by those damned libraries with their damned copy machines, not to mention kids with OCR who just throw the stuff up on the Internet.
Now, look at the Music industry. No real, provable signs that Internet piracy does a thing to their sales, yet publishers own artists' song rights forever, and now they want consumers to give up any concept of owning a song, the way we have for software. Oh no, now you own a license to play this song, which they can invalidate any damn time they please.
Working against the system when you are completely outside the system is futile.
Wrong. Almost all attempts at DRM are futile. No DRM will make it completely impossible to pirate something. If it does, it will be so oppressive that consumers won't take it anymore.
Here is the system that really works for everyone: For media, make it more con
We know a working quantum computer, on a sufficient scale, can crack modern encryption in something like linear time, or at least better than the current exponential time. We know that no such computer exists now, or at least not on sufficient scale to enable the NSA to snoop all our encrypted traffic.
Or (tinfoil hat time) do we really? Will we know when this happens, or will it be classified and snapped up by the government? Would we notice that? (The way we did with the a-bomb -- contests were held for whose work could be classified the fastest.) Or would we only notice years later, when it's finally leaked...
I don't actually know how it works, but I do remember how it used to work -- the older software suspend was incredibly slow because it tried to free as much cache as possible, then swap as much out as possible, so the actual image would be as small as possible. I'd argue that it still kind of makes sense to be able to swap some stuff out before suspend -- you may not need that space on resume.
But then, the newer suspend is fast, because it writes out just about everything, and it all gets loaded in one sequential disk read -- I wish it'd trim the caches some, though.
Oh, question -- how would you write out to a file on a partition in-use? Aren't you supposed to not touch in-use filesystems, otherwise Bad Things Happen?
I'd disagree with that, also -- the trick is good unit testing (to keep bugs from getting in), don't reinvent the wheel (use existing, already-debugged middleware where possible), and other known techniques like pair programming. You can spend 2x the time getting it right the first time, instead of 3x the time fixing it later.
There was some kind of paper on this, but I can't remember where I last saw it. Basically, it was a small group that could program significantly more reliable systems than most other software shops, mainly using the above principles, given the same amount of time and resources -- which means they get to charge a premium for delivering higher quality software.
So yes, still a management issue, at least partly. The other half of the equation is, only add programmers who are good. The copy'n'paste programmer can kill any project.
Swapping out makes sense sometimes, though. For instance, there are tiny chunks of the system -- daemons and such -- that are pretty much never accessed. I'd rather reclaim that, if only to cache something worthwhile.
Also, remember that suspend2 requires swap, so figure how much of an image you'll need (and how much is cache that can be freed) and get a bit more than that. My own rule of thumb is, swap is roughly 1x to 1.5x RAM, so that I can be sure I have room for the suspend. But I have the space, and Windows doesn't use swap for this anyway, it uses hiberfil.sys
1) It's still pay-to-play (you stop paying, songs stop playing)
Same with Napster and other similar services. The point is, with a reasonable broadband connection and an even moderate interest in new music, you can easily beat the price of, say, iTunes, or your neighborhood CD store.
Stop paying, and they stop playing, but I don't think it actually deletes them, meaning you could start paying again and they'll start playing again.
I believe the DRM is even broken now...
Besides, how many services do you pay for, versus what you own? I don't own my own house yet, but if you do, do you own your own generator/windmill/solar, or do you buy energy from the power company? How about Internet -- do you run your own ISP? Backbone? What happens to your Internet/Power/Phone/Water if you stop paying?
There are other reasons I don't like this and won't buy it, a major one being that MS would be a single kill switch for all of my music, and your other two points are valid. But the pricing of this is competitive enough that it's very tempting to give up your independence.
You can have a completely different style of gameplay depending on what class you are. Skills don't often create that. Classes create limitations which can make gameplay more interesting.
The most interesting gameplay I've ever seen in an MMO was Nexus TK. Four classes: Warrior, Rogue, Poet, Mage. Warrior and Rogue are fighters, and early in the game (before level 99), they are useful for completely different things. Rogues can "ambush" around creatures, and can thus charge straight past a huge hoard of enemies and kill the boss. They can deal the most damage, but only to one target at a time. Warriors deal more net damage, but they take much more damage, and are really only effective against large hoards. Poets deal almost no damage -- they are healers. Mages can heal, but not as well as a poet, and they can paralyze the room.
So, you have your two basic classes: fighter or caster. Then you have your main paths: Warrior, Rogue, Poet, Mage. Then, to make it more interesting, at level 99, you're able to trade experience for individual stats, so the main difference becomes what kind of items each class can use -- except, if you're obscenely rich, you can almost always find ridiculously good armor.
There's still plenty of creativity. There is real roleplaying, there are all kinds of ways to use your class, and plenty of things to exploit. For a long time, Rogues could skip fighting the Forever tree, which is required for certain levels much, much higher than 99, and requires a good few hours to kill, no matter how strong you are. There is no upper limit in strength, and the strongest character in the game (last time I played) was a Geomancer (Mage), which makes him much better at hitting individual targets -- high-level Mages can throw spells at any visible target that deal damage proportional to how much mana is available, and with several million Mana (level 99 is only a few thousand Mana), he could one-hit most people.
Sometimes I think the game seems to have settled into a WoW-like grind, only never-ending, because the most powerful people we know about only have around 2 million mana/vita, and the only cap is technological, based on the size of an int -- that's about 4.3 billion. But every time I think that, I see something creative and new to do with some aspect of the game. And a small, finite number of classes -- even if I count all 16 possible subpaths -- makes the whole thing go round.
(I could point to crafting skills as an example of a "hybrid", but while there are a few shared skills, even the crafting is something you have to choose (can't be a smith and a carpenter) and it doesn't affect fighting/hunting at all.)
There are very large, very complex programs (operating systems, databases, etc) that do not have serious problems. There are games that exist that have almost no problems.
Yes, you will always have bugs. That does not mean you can always use the "programming is hard" cop-out. When your competitors make games that run twice as fast on half the hardware -- when your game crashes twice a day and theirs never does -- something is seriously wrong.
It could be the programmers, it could be the management, but it sure as hell isn't "Waaaah! It's too complex!"
Question, though -- will it look better or worse than current games? The walls and everything looked about as good as Doom 3. About the only things that looked better were curved surfaces, but I don't think that measures up to the fact that -- just look at the models there. You're getting 20 fps on a model that seems to have less than 100 polys, with a 32 ghz cluster. I can get 60 fps on a real computer, on Quake 4, with lighting that looks almost as good, and models that look ridiculously better.
Without some means of interfacing to a computer it would be difficult to do anything with it, yet every means of accomplishing that under Unix is through an application. So what?
So it's an important question: If the GUI is part of the OS, then is running "startx" from a commandline effectively switching OSes?
Is DOS less of an OS that MacOS (pre 10) because it lacked a GUI?
That's not how I was measuring. I'd assumed DOS didn't do any kind of multitasking/multiplexing/whatever, because I've never seen it available to the user except through Windows 3.1 -- and one of the major problems there was that unless a DOS program was written specifically for Windows, it tended to assume it had the machine to itself, and never bothered to do things like release resources it wasn't using, cut down on CPU usage, or otherwise behave itself. Things tended to crash, hard, but instead of fixing the DOS apps, it was more common to modify your app to detect Windows and refuse to run under it. The exception was if you wanted to actually build a Windows GUI.
You could make a strong argument that the JRE is an OS. It's a pseudomachine after all. I don't think the same argument would be as strong for a web browser but it could be made. Just because an OS runs on top of another OS doesn't make it "not an OS".
I think OS becomes a meaningless distinction if you could argue that any scriptable/pluggable app is an OS. That would include word processors, IM programs, VLC (audio/video plugins are "programs"), any scripting language you like... If a web browser is an OS, I'd find it difficult to define what is NOT an OS.
It seems the argument is over the definition, not redefinition. You are implying there is already an accepted definition that someone else is trying to change.
Perhaps.
For a long time, I'd taken the approach towards OSes that's the classic approach towards pornography. I can't define it, but I know it when I see it. (Winston Churchill?)
It really seems like if Firefox is an OS, then we are changing an existing definition, vague as that definition may be. But that's just a gut feeling.
Grandparent claims the GUI is part of an operating system. I claim I can have an operating system without a GUI -- not only can I run the GUI as a separate application, but I can still do plenty of useful things with the GUI disabled.
What you are calling "operating system" is called "kernel" by the overall community.
That much we agree on, at least until you look at microkernels. And what's DOS, then? It doesn't even do multiplexing...
You see? Even "kernel" is pretty fuzzy.
As for whether Lisp etc is an operating system: Lisp was the OS the Lisp Machines. Java has been used as the OS for the Java buttons. I do not know of any case where emacs or.NET has been used as operating systems.
Maybe I should clarify: Grandparent was claiming an OS is the environment, and they seemed to be referring to a programming environment. I'm suggesting that if any of this is true, then Java, as it is used on modern OSes, is itself an OS -- it is a self-contained programming environment. So is a web browser, by the way.
We could try to redefine OS to fit anything we like, but I think we can all agree that a web browser is not, in itself, and operating system.
Either you're sarcastic, or you're just bad at GUIs. I can understand "Linux for Dummies" or "Commandline for Dummies". But "Google for Dummies"?
All that went wrong is that there are now more dummies, period (proof: 2004 re-election), and more dummies using more computers (and games), more often.
That's fair. While I appreciate a challenge, not being able to beat a game -- especially a game with a decent plot -- will usually mean I don't want to buy that game (or genre). Example: Midnight Club II. Either I suck at driving games, or other people find it fun to play a particular race 30, 40, even 50 times to get it just right. And no saving mid-race, of course -- you have to get the entire race just right from the beginning.
I don't think they are actually manuals. Most of the ones I've seen are walkthroughs, and they include information on how to find various secrets, what the correct way through this area is, and so on. And strategy. The ones I've seen spend maybe 2 pages on the controls and basic gameplay -- the same kind of stuff you'll probably learn on your own with an in-game tutorial level.
I have seen some recent games with incredibly thick manuals -- Final Fantasy XI, for one -- because the game itself was way too complex in how it handled some simple things. Took me long enough just to get the hang of moving around. But in the same genre, there are also games like Nexus TK, which requires memorizing quite a few keystrokes to be reasonably competent. More can be done with the mouse now, but you don't want to be using the mouse in the heat of battle. I'd have appreciated a manual here, except that I was taught by my roommate -- so, Nexus doesn't need a manual, because if you can make it through the training area, you'll find plenty of people willing to help you, especially because there's an in-game reward for helping newbies.
Your statement would be better founded if you would bother to look it up first. I'll save you the trouble:
An operating system (OS) is a software program that manages the hardware and software resources of a computer. The OS performs basic tasks, such as controlling and allocating memory, prioritizing the processing of instructions, controlling input and output devices, facilitating networking, and managing files.
I don't see anything in there about UIs.
The main advantages of an operating system include:
Allows multiple programs to run concurrently.
Simplifies the programming of application software because the program does not have to manage the hardware. The operating systems manages all hardware and the interaction of software. It also gives the program a high level interface to the hardware and ways of interacting with other programs.
That is the operating system.
As well as the kernel, an operating system is often distributed with system software that manages a graphical user interface (although Windows and Macintosh have integrated these programs into the operating system), as well as utility programs for tasks such as managing files and configuring the operating system. Often times distributed with operating systems are application software that does not directly relate to the operating system's core function, but which the operating system distributor finds advantageous to supply with the operating system.
I would argue that explorer.exe is an application. It is not required -- you can replace it with another shell, and still have Windows, and still run many Windows programs unmodified.
The delineation between the operating system and application software is not precise, and is occasionally subject to controversy. From commercial or legal points of view, the delineation can depend on the contexts of the interests involved. For example, one of the key questions in the United States v. Microsoft antitrust trial was whether Microsoft's web browser was part of its operating system, or whether it was a separable piece of application software.
This just seems ridiculous to me. Well-designed operating systems do not require any applications at all.
Look at it this way: I can install or uninstall Linux UIs, or switch between them, often on the fly -- without ever rebooting. By that I mean desktop environments (gnome/kde), window managers (sawfish/metacity/fluxbox/windowmaker), window servers (xfree/x.org/xgl/ywindows), standalone GUI framebuffer apps (links2), commandlines (bash/csh/ksh/ruby/python/mysql)... The only thing that requires a reboot is a kernel update or a crash. Everything else only requires variosu services to be started/restarted. I can run most of these simultaneously.
So, I've come to think of "operating system" in a fairly loose way. It's useful to refer to Windows as a whole to be an "operating system", so that one can contrast it with Linux or Mac OS X, without constantly having to say "Windows and Windows applications vs Linux and various open source applications" to make my point.
But it also helps to remember that most of userspace should be regarded as an application and/or an API, so as not to be confused about things like... well, like what you just said. Take Debian/Ubuntu. Ubuntu tracks Debian very closely -- I believe you can even "upgrade" a Debian installation to Ubuntu by changing your sources.
Each of these is unique enough to where without the special attention of package maintainers and translators, software would be hell to run across them.
Indeed? You could always take the Windows approach -- id seems to be able to distribute a single binary package for Quake 4 that works, out-of-the-box, on any x86 or amd64 Linux. It can't be taki
As for the promise of Vista vapor ware --well, lets just wait.
We can still theorize about how it might work, or how Apple might implement something similar. Or you can download the Vista beta.
Firefox like other apps, and its plugins ONLY have write access to any place the user has such access to. This means that if there is a bug or hole in Firefox, malware can damage any file the user can write to.
Yes, I understand it.
If a program saves code in user space, that code will still cause a dialog to come up the first time it is run as an independent program.
...wha?
If the extension wants to use a port or address that is not yet on the always allow list of Little Snitch, then there would be a warning.
Presumably you have to allow Firefox access to the entire Internet, right? Or at least to port 80 to everywhere. It would be quite annoying getting a popup every time you went to a new domain, or an existing one on a new round-robin address.
Any possible malware would have to *know* that it was running under my real account. If it did nothing in the test account it would do nothing on my regular account since that is not admin either.
There are other differences besides admin.
Access control lists don't have rules as such, but you can set up some pretty fine grained permissions for users and their directories.
What about programs?
An example of the simplest of problems is: Quake 4, running on my Linux, only needs read/write access to ~/.quake4, and read access to/opt/quake4 and one or two things in/lib. It may need to read one or two other things -- OpenAL configs, OpenGL libs, and sufficient access to the X server and/dev/sound -- but these are all well-defined things.
I should be able to easily set up this or any other game to be easily restricted to these areas. I'd be willing to trust the game not to screw up its own config and savegames, and I can easily backup/restore those anyway.
I know only vaguely of ways to make this possible, and I do not know of any way to make it easy/efficient enough to be practical.
Slashdot Mirror
Screens are getting better, and a new generation doesn't care much whether it's a book or on a screen, so long as it's a decent screen. e-ink will change all of this, even for older/pickier people.
With what a blank tape cost, and what a movie cost, this could still be reasonably done. It's more illegal, and it's easier to track down, but nothing like this has killed the movie industry. VCRs have, in fact, strengthened it.
Even without that, the movie industry's track record with VCRs, and the music industry with DATs, suggests that either they are completely incompetent at recognizing whether a technology will kill them (or how to make money off that technology), or that they are using piracy as an excuse to gain lockin for an ulterior motive. For instance, the fact that DVDs scratch fairly easily, and that the movie industry theory is that you should only watch the same exact disc, right off the disc, suggests that they really want DRM on DVDs, not because of piracy, but to force us to buy the same thing over and over again.
For that matter, buying the same thing over and over again is an RIAA/MPAA pattern.
I agree about your AllofMP3 assessment, though. I prefer lossless or Vorbis if I'm going to buy something, but the main criteria is cheapness and lack of DRM.
Where do you get your statistics from?
Oh, and how about we compare every song bought legally, period, versios how many illegally downloaded? Your statistic, even if true, is misleading.
Interesting idea. There are major downsides, but at least this has worked in the past -- it's called a concert.
You are repeating the same misconceptions in the same ways. Look at your language:
No one is arguing "that it's unnecessary to protect the works of content creators." We're arguing that DRM, specifically, in any form, is not worth the harm it causes, and that content creators can make a profit without it. After all, they did before DRM existed.
If you are astroturf, you need to listen here: If that is really and truly the reality, I will wean myself off modern media. I simply refuse to spend any money on anything that has unreasonable "protection" on it. If there are enough of us, you will lose money on DRM.
That is why I refuse to buy anything Blu-Ray until I am convinced that it's permanently cracked.
If not, I simply don't care enough. There is enough entertainment in the world that comes without strings attached.
Current DRM models have two problems: In order to enforce any kind of protection, they require specific software/hardware stacks, which reduces user choice -- for instance, it becomes essentially impossible to have a proper open-source media center, or even to run a closed media center on an open OS.
The second problem is, much of it is online. For instance, the music subscription services -- pay $x/mo and get as much music as you can download, but if you stop paying, they stop playing. Another example is Steam: You only pay once, but it insists on connecting to the Internet periodically to get updates and to be able to shut you down if they find two copies from the same purchase online at once. The problem with this is, I'm essentially trusting the content provider not to unfairly revoke my right to use my content -- Valve could one day decide not to let me play at all, or their servers could go down, and I'd be stuck without a game.
This puts things entirely too much out of control of the consumer, who, in a very real sense, no longer owns their stuff. Think of it this way -- the rights to a book are owned by the author, and only licensed to a publisher for a finite amount of time. If you buy a book, you own that copy, and may do whatever you want with it, other than distribute copies of significant portions of the book. Yet I never hear authors screaming about how they're being completely ripped off by those damned libraries with their damned copy machines, not to mention kids with OCR who just throw the stuff up on the Internet.
Now, look at the Music industry. No real, provable signs that Internet piracy does a thing to their sales, yet publishers own artists' song rights forever, and now they want consumers to give up any concept of owning a song, the way we have for software. Oh no, now you own a license to play this song, which they can invalidate any damn time they please.
Wrong. Almost all attempts at DRM are futile. No DRM will make it completely impossible to pirate something. If it does, it will be so oppressive that consumers won't take it anymore.
Here is the system that really works for everyone: For media, make it more con
We know a working quantum computer, on a sufficient scale, can crack modern encryption in something like linear time, or at least better than the current exponential time. We know that no such computer exists now, or at least not on sufficient scale to enable the NSA to snoop all our encrypted traffic.
Or (tinfoil hat time) do we really? Will we know when this happens, or will it be classified and snapped up by the government? Would we notice that? (The way we did with the a-bomb -- contests were held for whose work could be classified the fastest.) Or would we only notice years later, when it's finally leaked...
I don't actually know how it works, but I do remember how it used to work -- the older software suspend was incredibly slow because it tried to free as much cache as possible, then swap as much out as possible, so the actual image would be as small as possible. I'd argue that it still kind of makes sense to be able to swap some stuff out before suspend -- you may not need that space on resume.
But then, the newer suspend is fast, because it writes out just about everything, and it all gets loaded in one sequential disk read -- I wish it'd trim the caches some, though.
Oh, question -- how would you write out to a file on a partition in-use? Aren't you supposed to not touch in-use filesystems, otherwise Bad Things Happen?
I'd disagree with that, also -- the trick is good unit testing (to keep bugs from getting in), don't reinvent the wheel (use existing, already-debugged middleware where possible), and other known techniques like pair programming. You can spend 2x the time getting it right the first time, instead of 3x the time fixing it later.
There was some kind of paper on this, but I can't remember where I last saw it. Basically, it was a small group that could program significantly more reliable systems than most other software shops, mainly using the above principles, given the same amount of time and resources -- which means they get to charge a premium for delivering higher quality software.
So yes, still a management issue, at least partly. The other half of the equation is, only add programmers who are good. The copy'n'paste programmer can kill any project.
Swapping out makes sense sometimes, though. For instance, there are tiny chunks of the system -- daemons and such -- that are pretty much never accessed. I'd rather reclaim that, if only to cache something worthwhile.
Also, remember that suspend2 requires swap, so figure how much of an image you'll need (and how much is cache that can be freed) and get a bit more than that. My own rule of thumb is, swap is roughly 1x to 1.5x RAM, so that I can be sure I have room for the suspend. But I have the space, and Windows doesn't use swap for this anyway, it uses hiberfil.sys
How do those features and benefits scale, though?
Same with Napster and other similar services. The point is, with a reasonable broadband connection and an even moderate interest in new music, you can easily beat the price of, say, iTunes, or your neighborhood CD store.
Stop paying, and they stop playing, but I don't think it actually deletes them, meaning you could start paying again and they'll start playing again.
I believe the DRM is even broken now...
Besides, how many services do you pay for, versus what you own? I don't own my own house yet, but if you do, do you own your own generator/windmill/solar, or do you buy energy from the power company? How about Internet -- do you run your own ISP? Backbone? What happens to your Internet/Power/Phone/Water if you stop paying?
There are other reasons I don't like this and won't buy it, a major one being that MS would be a single kill switch for all of my music, and your other two points are valid. But the pricing of this is competitive enough that it's very tempting to give up your independence.
You can have a completely different style of gameplay depending on what class you are. Skills don't often create that. Classes create limitations which can make gameplay more interesting.
The most interesting gameplay I've ever seen in an MMO was Nexus TK. Four classes: Warrior, Rogue, Poet, Mage. Warrior and Rogue are fighters, and early in the game (before level 99), they are useful for completely different things. Rogues can "ambush" around creatures, and can thus charge straight past a huge hoard of enemies and kill the boss. They can deal the most damage, but only to one target at a time. Warriors deal more net damage, but they take much more damage, and are really only effective against large hoards. Poets deal almost no damage -- they are healers. Mages can heal, but not as well as a poet, and they can paralyze the room.
So, you have your two basic classes: fighter or caster. Then you have your main paths: Warrior, Rogue, Poet, Mage. Then, to make it more interesting, at level 99, you're able to trade experience for individual stats, so the main difference becomes what kind of items each class can use -- except, if you're obscenely rich, you can almost always find ridiculously good armor.
There's still plenty of creativity. There is real roleplaying, there are all kinds of ways to use your class, and plenty of things to exploit. For a long time, Rogues could skip fighting the Forever tree, which is required for certain levels much, much higher than 99, and requires a good few hours to kill, no matter how strong you are. There is no upper limit in strength, and the strongest character in the game (last time I played) was a Geomancer (Mage), which makes him much better at hitting individual targets -- high-level Mages can throw spells at any visible target that deal damage proportional to how much mana is available, and with several million Mana (level 99 is only a few thousand Mana), he could one-hit most people.
Sometimes I think the game seems to have settled into a WoW-like grind, only never-ending, because the most powerful people we know about only have around 2 million mana/vita, and the only cap is technological, based on the size of an int -- that's about 4.3 billion. But every time I think that, I see something creative and new to do with some aspect of the game. And a small, finite number of classes -- even if I count all 16 possible subpaths -- makes the whole thing go round.
(I could point to crafting skills as an example of a "hybrid", but while there are a few shared skills, even the crafting is something you have to choose (can't be a smith and a carpenter) and it doesn't affect fighting/hunting at all.)
There are very large, very complex programs (operating systems, databases, etc) that do not have serious problems. There are games that exist that have almost no problems.
Yes, you will always have bugs. That does not mean you can always use the "programming is hard" cop-out. When your competitors make games that run twice as fast on half the hardware -- when your game crashes twice a day and theirs never does -- something is seriously wrong.
It could be the programmers, it could be the management, but it sure as hell isn't "Waaaah! It's too complex!"
Here's hoping OpenRT beats them to it.
I don't think that would help for games, though.
Or are you saying it's like a BSP tree, Octree, something like that?
In other words, does changing the camera angle require you to rebuild that tree? Does changing the light source? What about moving objects around?
Question, though -- will it look better or worse than current games? The walls and everything looked about as good as Doom 3. About the only things that looked better were curved surfaces, but I don't think that measures up to the fact that -- just look at the models there. You're getting 20 fps on a model that seems to have less than 100 polys, with a 32 ghz cluster. I can get 60 fps on a real computer, on Quake 4, with lighting that looks almost as good, and models that look ridiculously better.
So it's an important question: If the GUI is part of the OS, then is running "startx" from a commandline effectively switching OSes?
That's not how I was measuring. I'd assumed DOS didn't do any kind of multitasking/multiplexing/whatever, because I've never seen it available to the user except through Windows 3.1 -- and one of the major problems there was that unless a DOS program was written specifically for Windows, it tended to assume it had the machine to itself, and never bothered to do things like release resources it wasn't using, cut down on CPU usage, or otherwise behave itself. Things tended to crash, hard, but instead of fixing the DOS apps, it was more common to modify your app to detect Windows and refuse to run under it. The exception was if you wanted to actually build a Windows GUI.
I think OS becomes a meaningless distinction if you could argue that any scriptable/pluggable app is an OS. That would include word processors, IM programs, VLC (audio/video plugins are "programs"), any scripting language you like... If a web browser is an OS, I'd find it difficult to define what is NOT an OS.
Perhaps.
For a long time, I'd taken the approach towards OSes that's the classic approach towards pornography. I can't define it, but I know it when I see it. (Winston Churchill?)
It really seems like if Firefox is an OS, then we are changing an existing definition, vague as that definition may be. But that's just a gut feeling.
But otherwise, where to draw the line?
Grandparent claims the GUI is part of an operating system. I claim I can have an operating system without a GUI -- not only can I run the GUI as a separate application, but I can still do plenty of useful things with the GUI disabled.
That much we agree on, at least until you look at microkernels. And what's DOS, then? It doesn't even do multiplexing...
You see? Even "kernel" is pretty fuzzy.
Maybe I should clarify: Grandparent was claiming an OS is the environment, and they seemed to be referring to a programming environment. I'm suggesting that if any of this is true, then Java, as it is used on modern OSes, is itself an OS -- it is a self-contained programming environment. So is a web browser, by the way.
We could try to redefine OS to fit anything we like, but I think we can all agree that a web browser is not, in itself, and operating system.
Caffine is a stimulant. Alcohol is a depressant. How does that work out? Do they cancel each other out, or do you just get jittery AND stupid?
Either you're sarcastic, or you're just bad at GUIs. I can understand "Linux for Dummies" or "Commandline for Dummies". But "Google for Dummies"?
All that went wrong is that there are now more dummies, period (proof: 2004 re-election), and more dummies using more computers (and games), more often.
That's fair. While I appreciate a challenge, not being able to beat a game -- especially a game with a decent plot -- will usually mean I don't want to buy that game (or genre). Example: Midnight Club II. Either I suck at driving games, or other people find it fun to play a particular race 30, 40, even 50 times to get it just right. And no saving mid-race, of course -- you have to get the entire race just right from the beginning.
I don't think they are actually manuals. Most of the ones I've seen are walkthroughs, and they include information on how to find various secrets, what the correct way through this area is, and so on. And strategy. The ones I've seen spend maybe 2 pages on the controls and basic gameplay -- the same kind of stuff you'll probably learn on your own with an in-game tutorial level.
I have seen some recent games with incredibly thick manuals -- Final Fantasy XI, for one -- because the game itself was way too complex in how it handled some simple things. Took me long enough just to get the hang of moving around. But in the same genre, there are also games like Nexus TK, which requires memorizing quite a few keystrokes to be reasonably competent. More can be done with the mouse now, but you don't want to be using the mouse in the heat of battle. I'd have appreciated a manual here, except that I was taught by my roommate -- so, Nexus doesn't need a manual, because if you can make it through the training area, you'll find plenty of people willing to help you, especially because there's an in-game reward for helping newbies.
Your statement would be better founded if you would bother to look it up first. I'll save you the trouble:
I don't see anything in there about UIs.
That is the operating system.
I would argue that explorer.exe is an application. It is not required -- you can replace it with another shell, and still have Windows, and still run many Windows programs unmodified.
This just seems ridiculous to me. Well-designed operating systems do not require any applications at all.
Look at it this way: I can install or uninstall Linux UIs, or switch between them, often on the fly -- without ever rebooting. By that I mean desktop environments (gnome/kde), window managers (sawfish/metacity/fluxbox/windowmaker), window servers (xfree/x.org/xgl/ywindows), standalone GUI framebuffer apps (links2), commandlines (bash/csh/ksh/ruby/python/mysql)... The only thing that requires a reboot is a kernel update or a crash. Everything else only requires variosu services to be started/restarted. I can run most of these simultaneously.
So, I've come to think of "operating system" in a fairly loose way. It's useful to refer to Windows as a whole to be an "operating system", so that one can contrast it with Linux or Mac OS X, without constantly having to say "Windows and Windows applications vs Linux and various open source applications" to make my point.
But it also helps to remember that most of userspace should be regarded as an application and/or an API, so as not to be confused about things like... well, like what you just said. Take Debian/Ubuntu. Ubuntu tracks Debian very closely -- I believe you can even "upgrade" a Debian installation to Ubuntu by changing your sources.
Indeed? You could always take the Windows approach -- id seems to be able to distribute a single binary package for Quake 4 that works, out-of-the-box, on any x86 or amd64 Linux. It can't be taki
Maybe not, but I was just quoting, and trying to avoid quoting the entire 6-page Courtney Love rant.
We can still theorize about how it might work, or how Apple might implement something similar. Or you can download the Vista beta.
Yes, I understand it.
...wha?
Presumably you have to allow Firefox access to the entire Internet, right? Or at least to port 80 to everywhere. It would be quite annoying getting a popup every time you went to a new domain, or an existing one on a new round-robin address.
There are other differences besides admin.
What about programs?
An example of the simplest of problems is: Quake 4, running on my Linux, only needs read/write access to ~/.quake4, and read access to /opt/quake4 and one or two things in /lib. It may need to read one or two other things -- OpenAL configs, OpenGL libs, and sufficient access to the X server and /dev/sound -- but these are all well-defined things.
I should be able to easily set up this or any other game to be easily restricted to these areas. I'd be willing to trust the game not to screw up its own config and savegames, and I can easily backup/restore those anyway.
I know only vaguely of ways to make this possible, and I do not know of any way to make it easy/efficient enough to be practical.