Umm, then how does the GPL work? How can Linus impose a continuing obligation on anyone who re-distributes Linux, or *any* derived work? (Remember, BTW, that the creator of a derived work owns the copyright on it. So if I build some new kernel modules based on Linux, I own the copyright, but Linus can still tell me that I have to GPL them!)
Copyright law says, at root, that if I own create/own a work, then nobody else can use it without my permission. That's all it says, and that's all it needs to say. Then, I can make my permission to use my works conditional on anything I want (barring a few specific limits), which is where the rest of the restrictions come into play.
If I want to GPL it, the restriction is that you agree to give GPL rights to anyone to whom you re-distribute my code, or your own derived code. If I want to make my work proprietary, I can forbid you to make copies of it at all.
I hate to be the one to break it to you, but you don't own the games you buy. You own a license to use those games. Big difference.
To the extent that you have personal values of ownership (why that could be so important to you is beyond me--haven't you ever rented an apartment, or stayed in a hotel room overnight?), you already HAVE compromised them. And you did it because either you didn't know any better (?), or because you knew on some level and didn't care--the enjoyment of playing the game was worth your $50, regardless of the technicalities of why you're allowed to possess and run the code.
Oh, geez... You guys are just throwing analogies around, and that's about as productive as debating by producing contradicting scriptural quotes.
The grandparent poster was correct: copyright law in the US (which brings contract law to bear) permits the copyright owner a hell of a lot of leeway in making demands on the user--there are limits, but they're waaaay out there. If Bill Gates wanted, he could include a clause in the Windows XP EULA that requires all users to twirl in a circle three times on the request of any MS employee. That's because it's a license agreement: as your end of a contract, you agree to accept MS's conditions, while they fulfill their end of the contract by allowing you to use their software.
The lug net analogy is attractive, but it doesn't fly too far because you're talking about buying the car/tires/lug nets, not licensing them. They could, conceivably, license the car to you (instead of selling it) with a condition of the license being that you don't use any mechanics besides the dealer's.
Holy shit... that actually happens! It's called a "lease", and millions of people in the US agree to them every year! We gotta warn those poor bastards!
"Unfortunately for us, when a corporation is publicly held, things don't always go according to the morals of the people in charge. In the end, they have to answer the shareholders, and their job is to please the shareholders, and the only way to do that is by making them money."
Wait, how is a publicly-held corporation any more responsible to its shareholders (or less accountible to the public) than a privately-held corporation? Whether the stock is publicly traded, over-the-counter, blue-chip, or family-owned, the same set of incentives, ethical rules, and governing principles apply. Most (by FAR) limited-liability corporations in the US are NOT publicly traded, and are owned entirely by an individual, a group of partners, or a small group of investors. Publicly-held or not has nothing to do with ownership incentives, but with rules about transparency.
And there are FAR more transparency, reporting, and accountibility requirements in place for publicy-traded corporations. Ever heard of the SEC? They only regulare public-traded and over-the-counter corps, not the closely-held LLCs that make up the bulk of the business environment, because all of their regulations stem from the fact that public outfits make public stock offerings. You can be a perfectly valid corporation and never make a public stock offering, and never really fall within the SEC's purview. When you do, and you become a "publicly held" corporation, you are watched more carefully and you have to tread more so, as well.
With that in mind, what exactly are you getting at? Are you talking about the mindsets of people in bureacratic organizations, or what?
"Don't forget, large corporations are amoral by nature, and any morality they display is either dictated by statute or is a side effect of their applying the profit motive. If IBM thought it made better business sense to side with SCO rather than mass against it, it would. Follow the money."
Okay, this is second time I've read kind of statement today, and probably the 50th time I've read it this month. And it's NOT true--it's a broad generalization that seems like it should be right, because we see corporate behavior in the news so often that seems to follow this principle.
I have worked within several actual corporations (LLCs and INCs, mostly) in my professional career so far, and done business with many, many more. My most recent job (investigations and litigation work) gave me an incredible inside view of high-level goings-on at a number of corporations, big and small. And it's just not that simple.
Corporate actors (managers, officers, executives) make moral/ethical decisions all the time--sometimes inside their own firm, sometimes outside. Even beyond the basic "ethics means not breaking the law" issue, most of these people are enlightened human beings with consciences. These decision-makers, as a group, will turn down moneymaking opportunities more often than you think in order to stay on the right side of ethics/morals, even if there's no legal or PR risk involved.
Not everyone acts like this all the time. But since it's not particularly newsworthy when a CEO does the right thing, quietly, in the privacy of a board meeting, you don't see it. This is a classic media bias problem: you assume that the world is a worse place that it really is, because you only get told about the bad things, by and large. That isn't the whole story.
Remember: Corporations are run by people, just like you, me, Abe Lincoln, and Hitler. We can only except "corporate" behavior to be within the range of human behavior, from the good to the bad, because corporate behavior is just HUMAN behavior. It's situational, sure, but most of the peculiar corporate manifestations (e.g., Dilbert) are really just bureacratic behaviors--you'd find similar stupidity in the government, the military, or non-profit orgs.
Okay--classic slashdot dumbass response, quoting the contradiction to his response IN THE RESPONSE. Follow the bouncing ball:
I said (which he quoted): 'So for 99% of the world's problems, why do you need more than 4-way systems?'
To which Bright Boy responds: 'You fall into the same old trap as everone before you, namely "one size fits all."'
Um, dude, I specifically said that one size DOESN'T fit all. I said that (given that we're pulling statistics out of my ass), 99 out of 100 workloads, roughly, don't need efficient 4-way or 8-way or more-way CPU configurations in one box. WHICH WOULD CATEGORICALLY MEAN THAT WORKLOADS DO EXIST THAT NEED MASSIVELY PARALLEL ARCHITECTURES. It's just that there are very few such cases, overall.
If you go back and look at the orignal point, that it's silly to go off bad-mouthing the cheaper, less powerful x86 stuff because it's less powerful in singular units, it still stands. For most cases, the fact that it's so commoditized and SO MUCH cheaper means that you can still solve big problems, even with lower efficiencies, because the price drop is bigger than the efficiency drop.
So yeah, for your 1-in-10^n situation, YOU might need more CPUs in a box. You sure seem like you're proud of that fact, too.
"Don't even think about multi-processor Xeon systems. The primitive bus architecture and interprocessor communications simply does not scale well at all past 2 processors. You can just about get away with 4 processors, but after that, you might as well just put space heaters in the box."
I would submit that there aren't too many applications that require more CPU cycles than 4x 3-Ghz Xeons can spin, even with less than perfect efficiency--let's say in the neighborhood of %60-%80, which fits with my own experiences on Quad-Xeon boxes going back about two years.
If a workload is linear, multiple processors won't help you, at all, because a truly linear process (think of a cyclical function like an MD5 hash computation, with a really long input string) depends on the outputs of previous operations.
If a workload isn't entirely linear, the degree to which it can be parallelized depends on how much different parts of the process are dependent on each other. If you're talking about running an SSH server for a couple of dozen clients, its perfectly parallelizable because each SSH connection is a separate process! If you're talking about database updates, only the reads and the non-conflicting writes are parallelizable--a lot (for most db apps), but not everything.
Also, the degree to which multiprocessing can help your problem depends on how fast the CPUs can communicate with each other (main memory, really, but if we're talking about a cluster with multiple "main" memory spaces, it's really the CPU-to-CPU latency that's the issue). Some problems are incredibly latency-sensitive, but others are not--like the SSH server example. If your problem is NOT latency-sensitive, you could potentially use a networked cluster of boxes, each box with a limited number of CPUs.
So the problems that need lots of CPUs in one box are the problems that are A) very parallellizable, and B) highly sensitive to latency between the CPUs. This is a pretty limited set of problems, though still substantial. I couldn't say exactly what portion of the worldwide computing workload falls into that category, but it ain't much.
So for 99% of the world's problems, why do you need more than 4-way systems? You might a need a lot of them, and maybe more boxes than if you had an efficient 8-way (or bigger) solution, but being as the price-per-machine scales up faster than processing power when you get past the 2-way point, your more efficient solution will probably be more expensive, too.
Taking potshots like this at x86 chips is such bullshit. So what if it's not as optimal an architecture as the Alpha, or if the EV7 bus is pretty neat? The biggest advantage of using x86 systems over anything else isn't that they're the fastest chips, cycle-for-cycle, or that they're a particularly elegant solution. It's that they're CHEAP and FAST ENOUGH.
Think about how many Intel Xeons you could get, on 9xx chipset mobos, for $30,000. If you built them yourself, probably 15-20. Is one (or four) 1.5 GHz Alphas are more useful than a cluster of 20 Xeons? Hell no!
See, ever since Intel lost their de facto monopoly on powerful x86 chips (thank you, AMD!), their prices have dropped far enough that it's hard to beat x86 solutions on a price vs. performance basis. Even if you have to stack up more boxes in a rack to do it. Hell, Quad-CPU Xeons can still go for less than $6,000, if you build them from parts, so rackspace isn't really an issue.
I think we're conflating "manufacturer" with "architecture", here. AMD's 32-bit offerings are basically software-compatible with Intel's 32-bit stuff (the exceptions would be SSE2 and such).
I guess the poster's point was that there aren't any widely-used architectures out there besides the x86 stuff, which was originally developed by Intel, was a solely Intel offering for a very long time (close to 15 years, I think), and which is still synonymous with Intel. Despite the fact that AMD, VIA, and a couple of other outfits make x86 CPUs.
This reads like a response to one of MS's most common attacks on OSS, especially when pitching to governments, about how increasing demand for OSS => decreasing demand for proprietary software, which causes the loss of jobs becasue OSS people do it for free, not as a career.
Which is bullshit.
1) Many, MANY OSS programmers work for traditional companies, which may or may not be primarily software companies. Really, it's not a case of some unpaid commie hippie stealing an MS programmer's job, it's a case of a well-paid IBM programmer stealing an MS programmer's job. Which is fine by me--the market at work.
2) The OSS development model seems to have lower overall costs associated with it--open-source projects can give you the same functionality and features, but the total cost of developing all that software is much less than the total cost of developing the congruent proprietary product. This is GOOD, because it means that less people are doing more work, which means we're more productive and efficient. MS hurts because they're not able to compete with the more efficient (and therefore cheaper to the consumer) OSS product, and they lose revenue. Again, fine all around.
What this is REALLY about is that OSS is a different management model for building software, and it's a model that's based on a different understanding of how best to profit from your ownership of intellectual property (copyright on software you've written/had written by others). That's why MS has started an internal drive to study the development process used by the kernel coders and others--they want to see if they can take some of the techniques and processes that are OSS and apply them to help MS become more competitive.
I actually have a variation on this scheme running on (almost) all my machines, at home. I'm working on the management interfaces to make it more useful in production, but the concept works:
I have an install system worked out based on a modified CRUX installer ISO. I added some packages and a default 2.6.8 kernel with NFSv4 and Kerberos support, so that when you boot the ISO to install you can mount an NFSv4 share as the root in which to install your system.
I install the distro normally--CRUX is slimmed-down and optimized for size, BTW--into the NFSv4 share. Then, I mount a temp filesystem from the NFSv4 server and run the "mkcramfs" script to turn all the directories and their contents (except for the contents of/var and/boot) into a Cramfs image, which cuts down the disk usage by about 60-70%. I make a gzipped tarball out of the contents of/var.
Then, I fdisk a device (which shows up as normal IDE) into three partitions:/boot, which is just big enough for two kernel images;/, which is just big enough for me to write the Cramfs image; and the remainder of the CF disk.
The/boot partition gets formatted EXT2, and the/boot files copied from NFS to the CF disk. Then the Cramfs image gets written to the / partition. Finally, I format the 3rd partition as EXT2 and copy the tarball of the/var contents to it. Then, I run lilo on the CF disk, using the lilo.conf file in NFS.
The rc scripts have to be modified so that after mounting/, they will temporarily mount that 3rd partition with the/var contents, mount a tmpfs RAM-based FS on/var, and un-tar the/var contents into the tmpfs volume. The rc.shutdown script has to be modified to do the reverse, so that on shutdown the current/var contents get tarred up and written back to the 3rd partition.
This is REALLY nice because it makes it easy to do package and kernel upgrades. Since you have the NFSv4 mount, holding the current state of the filesystem, you can do an upgrade by mounting the NFSv4 volume, chrooting into it, running your upgrades, and then re-doing the installation process.
This is all automated into a couple of shell scripts. Everything except my desktop machine (KDE, right?) runs inside of a 128 MB CF disk. I've even got one machine that runs a RAID 0 array of two CF cards, on on each IDE channel, so that it can keep running and alert me if either card goes south.
It sound complicated, but once you have the NFSv4 stuff set up, its really, really straightforward. And even that part can be done with less hassle by using NFSv3 without Kerberos, or just using something like SHFS (which I did, initially, because it only needs one simple package for the clients and nothing for the server besides OpenSSH).
I gotta put up a web page about this, at some point. It's pretty frickin' cool.
A large part of the reason why the internet has been such a big and fast-growing success is because the propensity to get online has been linked to education and sophistication in the user base. If you put 500 million cavemen and -women in front of these nifty terminals, I think the best you could hope for is some kind of "infinite monkeys" outcome.
But maybe I'm wrong--if you put halfway intelligent people online, even if they're not all totally literate at the beginning, they will probably get a lot more opportunities and incentives to climb the education curve. I guess that must be where all of these "cheap terminals for the 3rd world" are going.
And now that I think about it, one of big concentrations of unleashed education, intellect, and technical sophistication on the Internet is Slashdot. You can make up your own punchline on that.
Ok, Ok. Maybe you're not trolling--would a troll ever put this much effort into it? But you missed the point that I made in my first response. Let me restate it very clearly:
"LIMITED CORPORATE LIABILITY" HAS NO RELATION TO CRIMINAL OR CIVIL LIABILITY FOR A CORPORATION'S BEHAVIOR!!! It is an expression that ONLY relates bankruptcy and responsibility for debts.
Dude, this isn't personal--I have a very specific beef with you because you keep ranting on about how vile corporations are, but you don't seem to grasp the basics of what a corporation is, legally, which would seem to be a pretty important part of arguing about it! You probably took some intro economics classes, or read a bunch of econ blogs, without ever taking any formal introductions to basic business concepts--which is a pretty common disease on the web.
Here's the other point you missed: SUPPORTERS OF TERRORIST ORGANIZATIONS ONLY GO THE JAIL IF THEY *KNOW* THAT THEY WERE SUPPORTING A TERRORIST ORGANIZATION!! If a Muslim American gave money to an Islamic charity that was secretly diverting funds to Al-Qaeda, and the FBI shuts down the charity when they figure it out, the random Muslim doesn't go to jail!!
In America, at least, you can't prosecute the local Ford dealer when some bank robbers use an Escort Sedan they bought from him as a getaway car. If the dealer gave them in car in exchange for a cut of the robbery take, THEN he goes to jail because he knew he was helping with an illegal plan.
This all comes down to whether the person supporting the organization KNOWS that the organization is doing something illegal. If you knew about it beforehand and gave support anyway, you're a co-conspirator and you go to jail, too. If you didn't know about it beforehand, you don't have to answer for anything.
Now, ask yourself why, with so many corporate bankruptcies and so much bad behavior, why aren't investors going to jail, too? (HINT: IT HAS NOTHING TO DO WITH "LIMITED CORPORATE LIABILITY".)...Time's up! The answer is that the bulk of investors in a big corporation have no idea that the corporation in which they've invested has been doing anything illegal, because the actual perpetrators (like Ken Lay, Andrew Fastow, etc.) have been HIDING their illegal acts from everybody. If the investors could know, the SEC could know, and the bad guys would be caught.
Note, also, that the vast majority of corporations in the US never do anything significantly or intentionally illegal. This is important--you imply that investors ought to be suspicious of corporations in general, but there's no reason to be, because most of them by far are totally above-board. The few cases that surprise us get a lot of press, and probably have fooled you (and others) into the impression that MOST corporations act like that.
They don't. You assumed that they did, and looked for an explanation as to why investors don't go to jail, as they should know going in that corporations are inherently bad, greedy, evil groups that need to be watched closely. Misunderstanding the concept of limited corporate liability entirely, you leapt to a conclusion.
Your idea of scaring employees and investors into incredible paranoia by suddenly sending everyone to jail is plain nuts--it's trying to solve a problem that isn't that widespread with a sledge-hammer response. Do you want to throw every Catholic in the country in jail, too, for their collection-plate contributions that paid the wages of child molesters? Or how about black-listing people in the 1950s who believed in Communism and bought the Daily Worker, just because of COMINTERN's activities?
BTW, this thread is eventually going to get locked. If you want to continue this argument, we need to find another place to do it.
Hmm... Having read most of those links (and heard suspiciously similar ranting elsewhere), I'm gonna call this guy a troll.
In case he's not, well, he's a idiot. Here's a good one:
"Why should investors have limited liability? If people support a bad cause, shouldn't they too go to jail? It is happening now with people who supposedly support "terrorism", so why should corporate investors get a free pass when they support pollution, habitat destruction, sweatshop practices, employee boredom, and so on?"
The concept of "limited corporate liability" has absolutely NOTHING to do with criminal or financial liability of investors for corporate actions. Liability here refers purely to the financial obligations of the corporation--if the corp has outstanding debts that it can't pay, the lenders can't go after the investors' personal bank accounts to fulfill those debts. As an investor, your financial liability is limited to the amount that you invested in the corporation. Nothing more is meant, and nothing less.
Take the counter-example: If you're a member of or a benefactor (donator) to the KKK, and a bunch of other KKK members go out and burn down a church, you're not liable or responsible unless you:
A) participated in the illegal act,
B) knowingly gave support or aid to the illegal act,
C) helped plan the illegal act, or
D) knew about the illegal plan and failed to turn that knowledge over to the authorities (or stop the plan in some way).
If an investor knows that a corporation is doing something illegal, that particular investor might have criminal or civil liability for not doing something. But when Enron collapsed, did they send all the fucking secretaries and office managers to jail? No, because those people didn't have any fucking clue that bad shit was going on. The same thing is true for investors: with a few notable exceptions (board members/executives who were both shareholders and employees), the vast majority of Enron shareholders knew absolutely nothing about what was going on.
And so you, troll-boy, think that somehow this should be changed, so that be nature of an association to a corporation, you get an additionaly legal liability for its actions? This seems a little bit unfair--it's guilt by association, instead of guilt by actual responsibility. It's also prejudice against people who invest in incorporated companies, instead of in partnerships, limited-liability partnerships, non-limited liability corporations, or sole-proprietarships. Which is more than unfair--it just doesn't make any goddamned sense!
But now that I'm re-reading your last paragraph, I'm pretty sure that you're just a troll and that, but for the practice, this is wasted typing.
In order for any market predictions (or market advice, for that matter) to be useful data, it must be tied to a specific time period in which is can be used to make a profit. "Buy Enron" doesn't make a lot of sense if you get in for $50K right before it tanks, but it does make sense if you get in at the IPO and sell a month before the big drop.
So you could differentiate between self-fulfilling behaviors and independent behaviors: if the predictions says "SCOX will go from $5 to between $7 and $8 during the time period [1100-1400] tomorrow", and this news gets leaked the day before, then SCOX's price will start to climb right after the news comes out. By the time tomorrow at 11 AM rolls around, the price will already be well above $5, probably somewhere between $7 and $8, if most of the market deems the information trustworthy (lower if it doesn't, unless the stock is remarkably low volume).
Note that a model like this doesn't really even allow for a test case of the independent behaviors unless you tightly control the distribution of your information prior to the period when it becomes useful--or the market doesn't trust the data, or some similar kind of mechanism limits the impact that the predictions have on moneymakers.
But you can bet your ass: if the predictions are useful and accurate, word WILL get out that you have a magic trick. And then, EVERYBODY will immediately start following your buying/selling/shorting decisions, because they assume you're going to be right. Eventually, someone will probably figure out a way to beat your edge--tapping your phone calls to your broker, or bribing/blackmailing your broker--or just stealing the models from you. So this is neither a perfect edge, not an eternal one, because there is too much money to be made in not letting you keep it intact.
You avoid the tech terms because you don't know them. This is such a broken explanation that it's not even funny. I'll just take the obvious one, and let someone else pick on the rest:
"A single CPU computer can execute ONE instruction at the time. Meaning one program thread running at the time. But wait you say, my OS can run multiple programs at the same time. WRONG. It can't. It is a trick. It is running one program at the time but it is switching the program it is running really fast. There is however a problem with this. When it has switched to a program all the other programs are effectevily at the the mercy of the program now running INCLUDING the OS. Wich is why DOS and Windows and Linux and Mac OS and all the others had "hangups". With an extremely well written OS these hangups (when a program doesn't switch back to the OS) can be avoided but it still remains a case that all the programs and the OS are fighting for time on 1 single cpu."
NO. The ability to have multiple processes at once in a "runnable" state, with the CPU switching rapidly between them (called "multitasking"), is NOT the same beast across all those OSs that you just lumped together. MS-DOS was never capable of multi-tasking--it's a single-user, single-process OS. Mac OS=9 and Windows 95/98 are multi-tasking OSs, behaving somewhat as you described--one misbehaving program can prevent the kernel from asserting control of the hardware. But every modern OS, including Linux, Windows NT/2K/XP/2K3, the BSDs, and OSX, has a dependable scheduler that will maintain control in the face of any non-kernel process--the kernel determines when the currently-executing process will let go, not the process itself. On Linux and the BSDs, a mechanism that will let an arbitrary user program crash the machine in this fashion is usually considered a bug in its own right. It is possibly for user processes to tie up system resources (like hard disks) indefinitely, which can be hella inconvenient, but you can (barring bugs) always get a shell in edgewise to issue a "umount" or "shutdown -h now".
Also, I have run and maintained Windows 2000/XP on dozens of workstations, for years, and I think that the problems you describe are best attributed to flaky hardware or unstable drivers. There is no reason on a WinNT 5.x kernel why adding a second processor would cause less BSODs.
Also, I'm assuming you're running 2K or XP, because 98/ME can't even use a second processor. That means that you had to have re-installed Windows to make use of that second processor at all (to rebuild the hardware abstraction layer). If you did, re-installing Windows probably cleaned up some driver issues (maybe got rid of some malware you had lying around, too), which in turn explains why your machine is more "stable". If you didn't re-install Windows, or if you're running a non-NT kernel (95/98/ME), you aren't even using your second processor, ever. The OS is ignoring it completely.
And lastly, be careful with your definition of "stable"--there's a big difference between responsiveness in the UI and problems with the system actually crashing.
I think you're talking about two different mass die-offs. The Yucatan crater theoretically caused the Late Cretaceous die-off (approx. 65 million years ago) that made the dinosaurs go extincet. The Australian crater has been linked to the Late Permian die-off, which happened about 250 million years ago.
So, Racer X, the scientific community would appear to have two consensuses (consenses? WTF?), one on each of the two issues.
Mass extinctions are a fairly regular event in the Earth's geologic history. There are at least two more, besides the Permian and Cretaceous catastrophes, with which I'm familiar. Most people only get taught about the Cretaceous one in high school, though, so they never hear about the others.
Kind of like the Ice Age. Up until I was 16, I only thought there was one. Turns out there were a shitload of them.
You know, I saw this and said "fucking cool, man!" Package management has always been one of my least favorite things, and it sucks when you build something new only to realize that you've totally broken it. This is 10x worse when the package that breaks is, oh, say, RPM. (Fucking RedHat--Goddamn 8 and 9 were such crap.)
But I digress. Point is, I thought about using this on my laptop, which runs Crux/Slackware in a dual boot setup. But HTF am I supposed to use this intelligently when I've already got a package management system?
And wouldn't this mean that you have to either symlink all the binaries in each directory into/usr/bin, or add each app directory indivually to the path? Neither seems particularly cool.
I've been wondering where skyrocketing malpractice insurance premiums would eventually lead. This sucks, but something like this had to happen, eventually.
We want a perfect medical system where mistakes are minimized as much as possible, which lawsuits will encourage. But the cost adds up in terms of the risk that this system exposes individual doctors to--basically, being sued out of business. Every doctor will make a mistake at some point in his/her career, and that mistake might cost him/her everything.
Strangely, though, the availability of insurance screws this up. Those huge punitive awards are meant to pressure doctors not to screw up, but since virtually every practicing doctor has insurance, the cost of a lawsuit is spread over all of the doctors in terms of high insurance premiums. Since the pressure isn't specifically directed to punish the doctor that screws up (more so than any other doctor), its impact is limited.
And actually, those huge damage amounts are also a side-effect of insurance. You can't impose a $50-million judgement on a doctor who might be worth $1-3 million or so. Juries get a lot more open to imposing huge awards when they realize that the direct payee of the award is a faceless insurance company. Of course, everybody gets hurt on the back end, but that rarely occurs to anyone.
Honestly, it makes a lot more sense to cap/eliminate punitive awards in these cases, and to impose mandated penalties on doctors who lose malpractice cases: revoke medical licenses, ban from practice for a specified period. It's not perfect, but it won't end up being as expensive as the current mechanisms.
Actually, during the blackout last August, my cellphone service in NYC didn't drop at all. I mean, it was almost impossible to get a call through to anybody, but Verizon kept service the whole time--and I believe most of the other providers did, too.
A cell tower may or may not have battery backup or generator hook-ups, but I know that many cell towers in big cities are placed on buildings, and are hooked into the power system of the host building. Many large office buildings these days have redundant power backups available, and a prudent tower operator would contract with the building management to have his tower supported by the building backup system.
I've never seen USB 2.0 devices that could do sustained transfer rates of more than ~13 MB/sec. IEEE1394 using the Oxford 911 bridge maxes out at around 20 MB/sec.
Compare that to IDE on a PATA/133 bus, which can do > 34 MB/sec. at the beginning of a drive, trailing off to ~25 MB/sec. or so at the end.
Strangely enough, though, I HAVE noticed a declining transfer rate as you approach the end of the drive on the FireWire devices.
Have you seen transfer rates that approach the peak bus rates (in the 34 MB/sec range for USB 2, or 40 MB/sec for FW)? If so, I'd love to get my hands on the bridges that you're using!
This shit is getting out of hand. If it goes on long enough, we might start seeing a significant chunk of the economic benefit gained by having patents (it does exist, come on, guys!) getting swallowed up by the costs in time, money, and chilling effects on REAL entrepreneurs. At which point the US Supreme Court should be required to submit to whippings before reversing the god-awful decision that opened the door to business-method patents.
The problem with patenting business methods is that you cease to talk concretely about an actual implementation of a way of doing something. Take telephones: I could have invented cellular telephones a year after Bell was awarded his patent, and it wouldn't have infringed because it's a different way of doing the same thing. Now, though, I'm sure Bell would have patented "voice communications at a distance", or some such shit.
That's where the costs emerge--patents become a way of grabbing and extorting licenses for stuff that hasn't been invented yet, and which the patent holder couldn't have even invented.
Because that's what makes this possible, right? This isn't an invention--as another poster said, it's just an arrangement of existing systems. Now, some kind of actual implementation of a machine that would DO this might not be a business method patent, but I'd rather not get into a software patents discussion just now.
This kind of crap didn't happen as often before that goddamned decision--not as often, anyway. And I don't think that it's just the Internet that's making it happen.
What do you call it when someone is intentionally putting loud, obnoxious statements into circulation, but they aren't exactly trolling? Or is that trolling, too? I don't know--listen to this guy:
"Yes, that's hard but you must admit that I have a point."
Everything about this is calculated to piss off the Slashdot reader--the overt arrogance (Mensa membership!?? Are you kidding!??), the attempt to seem like an authority on the subject by making overly specific assertions, even down to the bad punctuation! On top of that, he starts throwing down about PHP and Perl--yeah, when I want to make a well-reasoned argument, the first thing I do is start flaming a religious OSS obsession. Good strategy. Now, I don't want to be too judgemental, because I can get pretty snippy myself (it's good to blow off steam by smacking someone around), but I think this goes further--it's a calculated attempt to piss people off for the purpose of pissing people off.
His argument doesn't even make any sense--the *right* tools for a job, ANY job, are the tools that fit that particular job. For small, non-scaling apps that need quick and easy-to-maintain/modify structures, PHP/Perl and Apache work wonderfully. Why the fuck should anyone start fucking around with Corba for building a web forum? Especially when the skills to work in PHP are much more common, and therefore cheaper?
I'm not trying to point out that he's wrong--I'm just trying to show how totally ungrounded the post is.
I've seen this account do this before--not always, so maybe it's not a straight-up troll account--but he IS just being an asshole.
Your last comment about people with multiple soundcards and special player apps isn't necessarily true. Anybody with multiple sound cards and a player app that can run in parallel can already read a single file to two outputs at once.* The grandparent poster's point about not changing anything about how the music can actually be played stands.
* But since a CD isn't a file, audio tracks would have to be read by a cd player app and the files would have to be read by the OS after mounting the file system. Since you can't have two apps colliding with each other, each assuming that it exclusively controls the CD-ROM device (or failing on a "busy" response when called for), you can't read both the audio tracks and the data files at the same time. Although you could have two separate processes reading the files simultaneously. Or, you could write an abstraction layer that mediates multiple control requests to a CD-ROM device by creating multiple virtual CD-ROM devices, and then serializing the incoming read requests and passing them on to the physical device, kind of like a file system does--of course, this wouldn't be very useful without a CD-ROM device that can physically read and seek fast enough to still permit the virtual devices to be read in real time.
Slashdot Mirror
Umm, then how does the GPL work? How can Linus impose a continuing obligation on anyone who re-distributes Linux, or *any* derived work? (Remember, BTW, that the creator of a derived work owns the copyright on it. So if I build some new kernel modules based on Linux, I own the copyright, but Linus can still tell me that I have to GPL them!)
Copyright law says, at root, that if I own create/own a work, then nobody else can use it without my permission. That's all it says, and that's all it needs to say. Then, I can make my permission to use my works conditional on anything I want (barring a few specific limits), which is where the rest of the restrictions come into play.
If I want to GPL it, the restriction is that you agree to give GPL rights to anyone to whom you re-distribute my code, or your own derived code. If I want to make my work proprietary, I can forbid you to make copies of it at all.
I hate to be the one to break it to you, but you don't own the games you buy. You own a license to use those games. Big difference.
To the extent that you have personal values of ownership (why that could be so important to you is beyond me--haven't you ever rented an apartment, or stayed in a hotel room overnight?), you already HAVE compromised them. And you did it because either you didn't know any better (?), or because you knew on some level and didn't care--the enjoyment of playing the game was worth your $50, regardless of the technicalities of why you're allowed to possess and run the code.
Oh, geez... You guys are just throwing analogies around, and that's about as productive as debating by producing contradicting scriptural quotes.
The grandparent poster was correct: copyright law in the US (which brings contract law to bear) permits the copyright owner a hell of a lot of leeway in making demands on the user--there are limits, but they're waaaay out there. If Bill Gates wanted, he could include a clause in the Windows XP EULA that requires all users to twirl in a circle three times on the request of any MS employee. That's because it's a license agreement: as your end of a contract, you agree to accept MS's conditions, while they fulfill their end of the contract by allowing you to use their software.
The lug net analogy is attractive, but it doesn't fly too far because you're talking about buying the car/tires/lug nets, not licensing them. They could, conceivably, license the car to you (instead of selling it) with a condition of the license being that you don't use any mechanics besides the dealer's.
Holy shit... that actually happens! It's called a "lease", and millions of people in the US agree to them every year! We gotta warn those poor bastards!
"Unfortunately for us, when a corporation is publicly held, things don't always go according to the morals of the people in charge. In the end, they have to answer the shareholders, and their job is to please the shareholders, and the only way to do that is by making them money."
Wait, how is a publicly-held corporation any more responsible to its shareholders (or less accountible to the public) than a privately-held corporation? Whether the stock is publicly traded, over-the-counter, blue-chip, or family-owned, the same set of incentives, ethical rules, and governing principles apply. Most (by FAR) limited-liability corporations in the US are NOT publicly traded, and are owned entirely by an individual, a group of partners, or a small group of investors. Publicly-held or not has nothing to do with ownership incentives, but with rules about transparency.
And there are FAR more transparency, reporting, and accountibility requirements in place for publicy-traded corporations. Ever heard of the SEC? They only regulare public-traded and over-the-counter corps, not the closely-held LLCs that make up the bulk of the business environment, because all of their regulations stem from the fact that public outfits make public stock offerings. You can be a perfectly valid corporation and never make a public stock offering, and never really fall within the SEC's purview. When you do, and you become a "publicly held" corporation, you are watched more carefully and you have to tread more so, as well.
With that in mind, what exactly are you getting at? Are you talking about the mindsets of people in bureacratic organizations, or what?
"Don't forget, large corporations are amoral by nature, and any morality they display is either dictated by statute or is a side effect of their applying the profit motive. If IBM thought it made better business sense to side with SCO rather than mass against it, it would. Follow the money."
Okay, this is second time I've read kind of statement today, and probably the 50th time I've read it this month. And it's NOT true--it's a broad generalization that seems like it should be right, because we see corporate behavior in the news so often that seems to follow this principle.
I have worked within several actual corporations (LLCs and INCs, mostly) in my professional career so far, and done business with many, many more. My most recent job (investigations and litigation work) gave me an incredible inside view of high-level goings-on at a number of corporations, big and small. And it's just not that simple.
Corporate actors (managers, officers, executives) make moral/ethical decisions all the time--sometimes inside their own firm, sometimes outside. Even beyond the basic "ethics means not breaking the law" issue, most of these people are enlightened human beings with consciences. These decision-makers, as a group, will turn down moneymaking opportunities more often than you think in order to stay on the right side of ethics/morals, even if there's no legal or PR risk involved.
Not everyone acts like this all the time. But since it's not particularly newsworthy when a CEO does the right thing, quietly, in the privacy of a board meeting, you don't see it. This is a classic media bias problem: you assume that the world is a worse place that it really is, because you only get told about the bad things, by and large. That isn't the whole story.
Remember: Corporations are run by people, just like you, me, Abe Lincoln, and Hitler. We can only except "corporate" behavior to be within the range of human behavior, from the good to the bad, because corporate behavior is just HUMAN behavior. It's situational, sure, but most of the peculiar corporate manifestations (e.g., Dilbert) are really just bureacratic behaviors--you'd find similar stupidity in the government, the military, or non-profit orgs.
Okay--classic slashdot dumbass response, quoting the contradiction to his response IN THE RESPONSE. Follow the bouncing ball:
I said (which he quoted): 'So for 99% of the world's problems, why do you need more than 4-way systems?'
To which Bright Boy responds: 'You fall into the same old trap as everone before you, namely "one size fits all."'
Um, dude, I specifically said that one size DOESN'T fit all. I said that (given that we're pulling statistics out of my ass), 99 out of 100 workloads, roughly, don't need efficient 4-way or 8-way or more-way CPU configurations in one box. WHICH WOULD CATEGORICALLY MEAN THAT WORKLOADS DO EXIST THAT NEED MASSIVELY PARALLEL ARCHITECTURES. It's just that there are very few such cases, overall.
If you go back and look at the orignal point, that it's silly to go off bad-mouthing the cheaper, less powerful x86 stuff because it's less powerful in singular units, it still stands. For most cases, the fact that it's so commoditized and SO MUCH cheaper means that you can still solve big problems, even with lower efficiencies, because the price drop is bigger than the efficiency drop.
So yeah, for your 1-in-10^n situation, YOU might need more CPUs in a box. You sure seem like you're proud of that fact, too.
"Don't even think about multi-processor Xeon systems. The primitive bus architecture and interprocessor communications simply does not scale well at all past 2 processors. You can just about get away with 4 processors, but after that, you might as well just put space heaters in the box."
I would submit that there aren't too many applications that require more CPU cycles than 4x 3-Ghz Xeons can spin, even with less than perfect efficiency--let's say in the neighborhood of %60-%80, which fits with my own experiences on Quad-Xeon boxes going back about two years.
If a workload is linear, multiple processors won't help you, at all, because a truly linear process (think of a cyclical function like an MD5 hash computation, with a really long input string) depends on the outputs of previous operations.
If a workload isn't entirely linear, the degree to which it can be parallelized depends on how much different parts of the process are dependent on each other. If you're talking about running an SSH server for a couple of dozen clients, its perfectly parallelizable because each SSH connection is a separate process! If you're talking about database updates, only the reads and the non-conflicting writes are parallelizable--a lot (for most db apps), but not everything.
Also, the degree to which multiprocessing can help your problem depends on how fast the CPUs can communicate with each other (main memory, really, but if we're talking about a cluster with multiple "main" memory spaces, it's really the CPU-to-CPU latency that's the issue). Some problems are incredibly latency-sensitive, but others are not--like the SSH server example. If your problem is NOT latency-sensitive, you could potentially use a networked cluster of boxes, each box with a limited number of CPUs.
So the problems that need lots of CPUs in one box are the problems that are A) very parallellizable, and B) highly sensitive to latency between the CPUs. This is a pretty limited set of problems, though still substantial. I couldn't say exactly what portion of the worldwide computing workload falls into that category, but it ain't much.
So for 99% of the world's problems, why do you need more than 4-way systems? You might a need a lot of them, and maybe more boxes than if you had an efficient 8-way (or bigger) solution, but being as the price-per-machine scales up faster than processing power when you get past the 2-way point, your more efficient solution will probably be more expensive, too.
Taking potshots like this at x86 chips is such bullshit. So what if it's not as optimal an architecture as the Alpha, or if the EV7 bus is pretty neat? The biggest advantage of using x86 systems over anything else isn't that they're the fastest chips, cycle-for-cycle, or that they're a particularly elegant solution. It's that they're CHEAP and FAST ENOUGH.
Think about how many Intel Xeons you could get, on 9xx chipset mobos, for $30,000. If you built them yourself, probably 15-20. Is one (or four) 1.5 GHz Alphas are more useful than a cluster of 20 Xeons? Hell no!
See, ever since Intel lost their de facto monopoly on powerful x86 chips (thank you, AMD!), their prices have dropped far enough that it's hard to beat x86 solutions on a price vs. performance basis. Even if you have to stack up more boxes in a rack to do it. Hell, Quad-CPU Xeons can still go for less than $6,000, if you build them from parts, so rackspace isn't really an issue.
I think we're conflating "manufacturer" with "architecture", here. AMD's 32-bit offerings are basically software-compatible with Intel's 32-bit stuff (the exceptions would be SSE2 and such).
I guess the poster's point was that there aren't any widely-used architectures out there besides the x86 stuff, which was originally developed by Intel, was a solely Intel offering for a very long time (close to 15 years, I think), and which is still synonymous with Intel. Despite the fact that AMD, VIA, and a couple of other outfits make x86 CPUs.
This reads like a response to one of MS's most common attacks on OSS, especially when pitching to governments, about how increasing demand for OSS => decreasing demand for proprietary software, which causes the loss of jobs becasue OSS people do it for free, not as a career.
Which is bullshit.
1) Many, MANY OSS programmers work for traditional companies, which may or may not be primarily software companies. Really, it's not a case of some unpaid commie hippie stealing an MS programmer's job, it's a case of a well-paid IBM programmer stealing an MS programmer's job. Which is fine by me--the market at work.
2) The OSS development model seems to have lower overall costs associated with it--open-source projects can give you the same functionality and features, but the total cost of developing all that software is much less than the total cost of developing the congruent proprietary product. This is GOOD, because it means that less people are doing more work, which means we're more productive and efficient. MS hurts because they're not able to compete with the more efficient (and therefore cheaper to the consumer) OSS product, and they lose revenue. Again, fine all around.
What this is REALLY about is that OSS is a different management model for building software, and it's a model that's based on a different understanding of how best to profit from your ownership of intellectual property (copyright on software you've written/had written by others). That's why MS has started an internal drive to study the development process used by the kernel coders and others--they want to see if they can take some of the techniques and processes that are OSS and apply them to help MS become more competitive.
"At least the caveman has shown that they have the one thing that's important. The ability to survive and propogate."
I don't know about that--you don't see many cavemen walking around these days, do you?
Unless you buy into the theory that they all bought Win98 PCs and got online, where they grunt along as ACs on slashdot.
Is that you, Ungk? Get a goddamned user account.
I actually have a variation on this scheme running on (almost) all my machines, at home. I'm working on the management interfaces to make it more useful in production, but the concept works:
/var and /boot) into a Cramfs image, which cuts down the disk usage by about 60-70%. I make a gzipped tarball out of the contents of /var.
/boot, which is just big enough for two kernel images; /, which is just big enough for me to write the Cramfs image; and the remainder of the CF disk.
/boot partition gets formatted EXT2, and the /boot files copied from NFS to the CF disk. Then the Cramfs image gets written to the / partition. Finally, I format the 3rd partition as EXT2 and copy the tarball of the /var contents to it. Then, I run lilo on the CF disk, using the lilo.conf file in NFS.
/, they will temporarily mount that 3rd partition with the /var contents, mount a tmpfs RAM-based FS on /var, and un-tar the /var contents into the tmpfs volume. The rc.shutdown script has to be modified to do the reverse, so that on shutdown the current /var contents get tarred up and written back to the 3rd partition.
I have an install system worked out based on a modified CRUX installer ISO. I added some packages and a default 2.6.8 kernel with NFSv4 and Kerberos support, so that when you boot the ISO to install you can mount an NFSv4 share as the root in which to install your system.
I install the distro normally--CRUX is slimmed-down and optimized for size, BTW--into the NFSv4 share. Then, I mount a temp filesystem from the NFSv4 server and run the "mkcramfs" script to turn all the directories and their contents (except for the contents of
Then, I fdisk a device (which shows up as normal IDE) into three partitions:
The
The rc scripts have to be modified so that after mounting
This is REALLY nice because it makes it easy to do package and kernel upgrades. Since you have the NFSv4 mount, holding the current state of the filesystem, you can do an upgrade by mounting the NFSv4 volume, chrooting into it, running your upgrades, and then re-doing the installation process.
This is all automated into a couple of shell scripts. Everything except my desktop machine (KDE, right?) runs inside of a 128 MB CF disk. I've even got one machine that runs a RAID 0 array of two CF cards, on on each IDE channel, so that it can keep running and alert me if either card goes south.
It sound complicated, but once you have the NFSv4 stuff set up, its really, really straightforward. And even that part can be done with less hassle by using NFSv3 without Kerberos, or just using something like SHFS (which I did, initially, because it only needs one simple package for the clients and nothing for the server besides OpenSSH).
I gotta put up a web page about this, at some point. It's pretty frickin' cool.
A large part of the reason why the internet has been such a big and fast-growing success is because the propensity to get online has been linked to education and sophistication in the user base. If you put 500 million cavemen and -women in front of these nifty terminals, I think the best you could hope for is some kind of "infinite monkeys" outcome.
But maybe I'm wrong--if you put halfway intelligent people online, even if they're not all totally literate at the beginning, they will probably get a lot more opportunities and incentives to climb the education curve. I guess that must be where all of these "cheap terminals for the 3rd world" are going.
And now that I think about it, one of big concentrations of unleashed education, intellect, and technical sophistication on the Internet is Slashdot. You can make up your own punchline on that.
Ok, Ok. Maybe you're not trolling--would a troll ever put this much effort into it? But you missed the point that I made in my first response. Let me restate it very clearly:
...Time's up! The answer is that the bulk of investors in a big corporation have no idea that the corporation in which they've invested has been doing anything illegal, because the actual perpetrators (like Ken Lay, Andrew Fastow, etc.) have been HIDING their illegal acts from everybody. If the investors could know, the SEC could know, and the bad guys would be caught.
"LIMITED CORPORATE LIABILITY" HAS NO RELATION TO CRIMINAL OR CIVIL LIABILITY FOR A CORPORATION'S BEHAVIOR!!! It is an expression that ONLY relates bankruptcy and responsibility for debts.
Dude, this isn't personal--I have a very specific beef with you because you keep ranting on about how vile corporations are, but you don't seem to grasp the basics of what a corporation is, legally, which would seem to be a pretty important part of arguing about it! You probably took some intro economics classes, or read a bunch of econ blogs, without ever taking any formal introductions to basic business concepts--which is a pretty common disease on the web.
Here's the other point you missed: SUPPORTERS OF TERRORIST ORGANIZATIONS ONLY GO THE JAIL IF THEY *KNOW* THAT THEY WERE SUPPORTING A TERRORIST ORGANIZATION!! If a Muslim American gave money to an Islamic charity that was secretly diverting funds to Al-Qaeda, and the FBI shuts down the charity when they figure it out, the random Muslim doesn't go to jail!!
In America, at least, you can't prosecute the local Ford dealer when some bank robbers use an Escort Sedan they bought from him as a getaway car. If the dealer gave them in car in exchange for a cut of the robbery take, THEN he goes to jail because he knew he was helping with an illegal plan.
This all comes down to whether the person supporting the organization KNOWS that the organization is doing something illegal. If you knew about it beforehand and gave support anyway, you're a co-conspirator and you go to jail, too. If you didn't know about it beforehand, you don't have to answer for anything.
Now, ask yourself why, with so many corporate bankruptcies and so much bad behavior, why aren't investors going to jail, too? (HINT: IT HAS NOTHING TO DO WITH "LIMITED CORPORATE LIABILITY".)
Note, also, that the vast majority of corporations in the US never do anything significantly or intentionally illegal. This is important--you imply that investors ought to be suspicious of corporations in general, but there's no reason to be, because most of them by far are totally above-board. The few cases that surprise us get a lot of press, and probably have fooled you (and others) into the impression that MOST corporations act like that.
They don't. You assumed that they did, and looked for an explanation as to why investors don't go to jail, as they should know going in that corporations are inherently bad, greedy, evil groups that need to be watched closely. Misunderstanding the concept of limited corporate liability entirely, you leapt to a conclusion.
Your idea of scaring employees and investors into incredible paranoia by suddenly sending everyone to jail is plain nuts--it's trying to solve a problem that isn't that widespread with a sledge-hammer response. Do you want to throw every Catholic in the country in jail, too, for their collection-plate contributions that paid the wages of child molesters? Or how about black-listing people in the 1950s who believed in Communism and bought the Daily Worker, just because of COMINTERN's activities?
BTW, this thread is eventually going to get locked. If you want to continue this argument, we need to find another place to do it.
Hmm... Having read most of those links (and heard suspiciously similar ranting elsewhere), I'm gonna call this guy a troll.
In case he's not, well, he's a idiot. Here's a good one:
"Why should investors have limited liability? If people support a bad cause, shouldn't they too go to jail? It is happening now with people who supposedly support "terrorism", so why should corporate investors get a free pass when they support pollution, habitat destruction, sweatshop practices, employee boredom, and so on?"
The concept of "limited corporate liability" has absolutely NOTHING to do with criminal or financial liability of investors for corporate actions. Liability here refers purely to the financial obligations of the corporation--if the corp has outstanding debts that it can't pay, the lenders can't go after the investors' personal bank accounts to fulfill those debts. As an investor, your financial liability is limited to the amount that you invested in the corporation. Nothing more is meant, and nothing less.
Take the counter-example: If you're a member of or a benefactor (donator) to the KKK, and a bunch of other KKK members go out and burn down a church, you're not liable or responsible unless you:
A) participated in the illegal act,
B) knowingly gave support or aid to the illegal act,
C) helped plan the illegal act, or
D) knew about the illegal plan and failed to turn that knowledge over to the authorities (or stop the plan in some way).
If an investor knows that a corporation is doing something illegal, that particular investor might have criminal or civil liability for not doing something. But when Enron collapsed, did they send all the fucking secretaries and office managers to jail? No, because those people didn't have any fucking clue that bad shit was going on. The same thing is true for investors: with a few notable exceptions (board members/executives who were both shareholders and employees), the vast majority of Enron shareholders knew absolutely nothing about what was going on.
And so you, troll-boy, think that somehow this should be changed, so that be nature of an association to a corporation, you get an additionaly legal liability for its actions? This seems a little bit unfair--it's guilt by association, instead of guilt by actual responsibility. It's also prejudice against people who invest in incorporated companies, instead of in partnerships, limited-liability partnerships, non-limited liability corporations, or sole-proprietarships. Which is more than unfair--it just doesn't make any goddamned sense!
But now that I'm re-reading your last paragraph, I'm pretty sure that you're just a troll and that, but for the practice, this is wasted typing.
Well, this may not be such a good example.
In order for any market predictions (or market advice, for that matter) to be useful data, it must be tied to a specific time period in which is can be used to make a profit. "Buy Enron" doesn't make a lot of sense if you get in for $50K right before it tanks, but it does make sense if you get in at the IPO and sell a month before the big drop.
So you could differentiate between self-fulfilling behaviors and independent behaviors: if the predictions says "SCOX will go from $5 to between $7 and $8 during the time period [1100-1400] tomorrow", and this news gets leaked the day before, then SCOX's price will start to climb right after the news comes out. By the time tomorrow at 11 AM rolls around, the price will already be well above $5, probably somewhere between $7 and $8, if most of the market deems the information trustworthy (lower if it doesn't, unless the stock is remarkably low volume).
Note that a model like this doesn't really even allow for a test case of the independent behaviors unless you tightly control the distribution of your information prior to the period when it becomes useful--or the market doesn't trust the data, or some similar kind of mechanism limits the impact that the predictions have on moneymakers.
But you can bet your ass: if the predictions are useful and accurate, word WILL get out that you have a magic trick. And then, EVERYBODY will immediately start following your buying/selling/shorting decisions, because they assume you're going to be right. Eventually, someone will probably figure out a way to beat your edge--tapping your phone calls to your broker, or bribing/blackmailing your broker--or just stealing the models from you. So this is neither a perfect edge, not an eternal one, because there is too much money to be made in not letting you keep it intact.
As the Libertarians say, "The market will solve."
You avoid the tech terms because you don't know them. This is such a broken explanation that it's not even funny. I'll just take the obvious one, and let someone else pick on the rest:
"A single CPU computer can execute ONE instruction at the time. Meaning one program thread running at the time. But wait you say, my OS can run multiple programs at the same time. WRONG. It can't. It is a trick. It is running one program at the time but it is switching the program it is running really fast. There is however a problem with this. When it has switched to a program all the other programs are effectevily at the the mercy of the program now running INCLUDING the OS. Wich is why DOS and Windows and Linux and Mac OS and all the others had "hangups". With an extremely well written OS these hangups (when a program doesn't switch back to the OS) can be avoided but it still remains a case that all the programs and the OS are fighting for time on 1 single cpu."
NO. The ability to have multiple processes at once in a "runnable" state, with the CPU switching rapidly between them (called "multitasking"), is NOT the same beast across all those OSs that you just lumped together. MS-DOS was never capable of multi-tasking--it's a single-user, single-process OS. Mac OS=9 and Windows 95/98 are multi-tasking OSs, behaving somewhat as you described--one misbehaving program can prevent the kernel from asserting control of the hardware. But every modern OS, including Linux, Windows NT/2K/XP/2K3, the BSDs, and OSX, has a dependable scheduler that will maintain control in the face of any non-kernel process--the kernel determines when the currently-executing process will let go, not the process itself. On Linux and the BSDs, a mechanism that will let an arbitrary user program crash the machine in this fashion is usually considered a bug in its own right. It is possibly for user processes to tie up system resources (like hard disks) indefinitely, which can be hella inconvenient, but you can (barring bugs) always get a shell in edgewise to issue a "umount" or "shutdown -h now".
Also, I have run and maintained Windows 2000/XP on dozens of workstations, for years, and I think that the problems you describe are best attributed to flaky hardware or unstable drivers. There is no reason on a WinNT 5.x kernel why adding a second processor would cause less BSODs.
Also, I'm assuming you're running 2K or XP, because 98/ME can't even use a second processor. That means that you had to have re-installed Windows to make use of that second processor at all (to rebuild the hardware abstraction layer). If you did, re-installing Windows probably cleaned up some driver issues (maybe got rid of some malware you had lying around, too), which in turn explains why your machine is more "stable". If you didn't re-install Windows, or if you're running a non-NT kernel (95/98/ME), you aren't even using your second processor, ever. The OS is ignoring it completely.
And lastly, be careful with your definition of "stable"--there's a big difference between responsiveness in the UI and problems with the system actually crashing.
I think you're talking about two different mass die-offs. The Yucatan crater theoretically caused the Late Cretaceous die-off (approx. 65 million years ago) that made the dinosaurs go extincet. The Australian crater has been linked to the Late Permian die-off, which happened about 250 million years ago.
So, Racer X, the scientific community would appear to have two consensuses (consenses? WTF?), one on each of the two issues.
Mass extinctions are a fairly regular event in the Earth's geologic history. There are at least two more, besides the Permian and Cretaceous catastrophes, with which I'm familiar. Most people only get taught about the Cretaceous one in high school, though, so they never hear about the others.
Kind of like the Ice Age. Up until I was 16, I only thought there was one. Turns out there were a shitload of them.
You know, I saw this and said "fucking cool, man!" Package management has always been one of my least favorite things, and it sucks when you build something new only to realize that you've totally broken it. This is 10x worse when the package that breaks is, oh, say, RPM. (Fucking RedHat--Goddamn 8 and 9 were such crap.)
/usr/bin, or add each app directory indivually to the path? Neither seems particularly cool.
But I digress. Point is, I thought about using this on my laptop, which runs Crux/Slackware in a dual boot setup. But HTF am I supposed to use this intelligently when I've already got a package management system?
And wouldn't this mean that you have to either symlink all the binaries in each directory into
WHy is the desktop trying to manage apps, again?
I've been wondering where skyrocketing malpractice insurance premiums would eventually lead. This sucks, but something like this had to happen, eventually.
We want a perfect medical system where mistakes are minimized as much as possible, which lawsuits will encourage. But the cost adds up in terms of the risk that this system exposes individual doctors to--basically, being sued out of business. Every doctor will make a mistake at some point in his/her career, and that mistake might cost him/her everything.
Strangely, though, the availability of insurance screws this up. Those huge punitive awards are meant to pressure doctors not to screw up, but since virtually every practicing doctor has insurance, the cost of a lawsuit is spread over all of the doctors in terms of high insurance premiums. Since the pressure isn't specifically directed to punish the doctor that screws up (more so than any other doctor), its impact is limited.
And actually, those huge damage amounts are also a side-effect of insurance. You can't impose a $50-million judgement on a doctor who might be worth $1-3 million or so. Juries get a lot more open to imposing huge awards when they realize that the direct payee of the award is a faceless insurance company. Of course, everybody gets hurt on the back end, but that rarely occurs to anyone.
Honestly, it makes a lot more sense to cap/eliminate punitive awards in these cases, and to impose mandated penalties on doctors who lose malpractice cases: revoke medical licenses, ban from practice for a specified period. It's not perfect, but it won't end up being as expensive as the current mechanisms.
Actually, during the blackout last August, my cellphone service in NYC didn't drop at all. I mean, it was almost impossible to get a call through to anybody, but Verizon kept service the whole time--and I believe most of the other providers did, too.
A cell tower may or may not have battery backup or generator hook-ups, but I know that many cell towers in big cities are placed on buildings, and are hooked into the power system of the host building. Many large office buildings these days have redundant power backups available, and a prudent tower operator would contract with the building management to have his tower supported by the building backup system.
I've never seen USB 2.0 devices that could do sustained transfer rates of more than ~13 MB/sec. IEEE1394 using the Oxford 911 bridge maxes out at around 20 MB/sec.
Compare that to IDE on a PATA/133 bus, which can do > 34 MB/sec. at the beginning of a drive, trailing off to ~25 MB/sec. or so at the end.
Strangely enough, though, I HAVE noticed a declining transfer rate as you approach the end of the drive on the FireWire devices.
Have you seen transfer rates that approach the peak bus rates (in the 34 MB/sec range for USB 2, or 40 MB/sec for FW)? If so, I'd love to get my hands on the bridges that you're using!
This shit is getting out of hand. If it goes on long enough, we might start seeing a significant chunk of the economic benefit gained by having patents (it does exist, come on, guys!) getting swallowed up by the costs in time, money, and chilling effects on REAL entrepreneurs. At which point the US Supreme Court should be required to submit to whippings before reversing the god-awful decision that opened the door to business-method patents.
The problem with patenting business methods is that you cease to talk concretely about an actual implementation of a way of doing something. Take telephones: I could have invented cellular telephones a year after Bell was awarded his patent, and it wouldn't have infringed because it's a different way of doing the same thing. Now, though, I'm sure Bell would have patented "voice communications at a distance", or some such shit.
That's where the costs emerge--patents become a way of grabbing and extorting licenses for stuff that hasn't been invented yet, and which the patent holder couldn't have even invented.
Because that's what makes this possible, right? This isn't an invention--as another poster said, it's just an arrangement of existing systems. Now, some kind of actual implementation of a machine that would DO this might not be a business method patent, but I'd rather not get into a software patents discussion just now.
This kind of crap didn't happen as often before that goddamned decision--not as often, anyway. And I don't think that it's just the Internet that's making it happen.
What do you call it when someone is intentionally putting loud, obnoxious statements into circulation, but they aren't exactly trolling? Or is that trolling, too? I don't know--listen to this guy:
"Yes, that's hard but you must admit that I have a point."
Everything about this is calculated to piss off the Slashdot reader--the overt arrogance (Mensa membership!?? Are you kidding!??), the attempt to seem like an authority on the subject by making overly specific assertions, even down to the bad punctuation! On top of that, he starts throwing down about PHP and Perl--yeah, when I want to make a well-reasoned argument, the first thing I do is start flaming a religious OSS obsession. Good strategy.
Now, I don't want to be too judgemental, because I can get pretty snippy myself (it's good to blow off steam by smacking someone around), but I think this goes further--it's a calculated attempt to piss people off for the purpose of pissing people off.
His argument doesn't even make any sense--the *right* tools for a job, ANY job, are the tools that fit that particular job. For small, non-scaling apps that need quick and easy-to-maintain/modify structures, PHP/Perl and Apache work wonderfully. Why the fuck should anyone start fucking around with Corba for building a web forum? Especially when the skills to work in PHP are much more common, and therefore cheaper?
I'm not trying to point out that he's wrong--I'm just trying to show how totally ungrounded the post is.
I've seen this account do this before--not always, so maybe it's not a straight-up troll account--but he IS just being an asshole.
Your last comment about people with multiple soundcards and special player apps isn't necessarily true. Anybody with multiple sound cards and a player app that can run in parallel can already read a single file to two outputs at once.* The grandparent poster's point about not changing anything about how the music can actually be played stands.
* But since a CD isn't a file, audio tracks would have to be read by a cd player app and the files would have to be read by the OS after mounting the file system. Since you can't have two apps colliding with each other, each assuming that it exclusively controls the CD-ROM device (or failing on a "busy" response when called for), you can't read both the audio tracks and the data files at the same time. Although you could have two separate processes reading the files simultaneously. Or, you could write an abstraction layer that mediates multiple control requests to a CD-ROM device by creating multiple virtual CD-ROM devices, and then serializing the incoming read requests and passing them on to the physical device, kind of like a file system does--of course, this wouldn't be very useful without a CD-ROM device that can physically read and seek fast enough to still permit the virtual devices to be read in real time.