Wait... so you are trying to say that the Catholic Church is not a scam?
Hmmm... interesting.
And I always thought that saying that you'll be sent to hell by an invisible guy and his long-haired Jewish son who died but somehow isn't dead and becomes a cracker sometimes, unless you join the church and put money on the collection plate was... a bit of a scam.
IBM cannot afford to pay people much more than they contribute to the IBM bottom line, simply because if they did that, they'd lose money.
IBM cannot afford to pay people much less than they contribute to the IBM bottom line, simply because if you can contribute $X to the IBM bottom line then you can likely also contribute $X to some other company's bottom line, and you might jump ship.
This sets pretty hard boundaries on how much pay IBM can and cannot offer. True, IBM happens to have a strong market position and probably a bit more flexibility with this than other firms, but if the overtime pay really meant a raise big enough that they had to cut wages 15% just to compensate, then I suggest that it may very well be the case that they could not afford, in their competitive business, not to make that cut. They surely won't do it very lightly or just for fun, seeing as it is not likely to be good for employee morale, which is very important in a knowledge-intensive service industry like IT consulting (their core business these days.)
BTW, in case anybody wants to say that IBM makes a profit and that it could therefore pay their workers more and cut into that profit: that's not how it works. Profits are returns on capital. Capital doesn't fall out of thin air. It must be gotten on capital markets, like bond and stock markets, and investors expect a certain return. If IBM has a higher return than that, then the price of the stock will go up so as to equalize the return per dollar invested anyway.
Really, the boundaries within which a company like IBM has to maneuver this kind of thing are pretty tight. Unionizing the shop ain't gonna help much either, because much of what IBM does is in pretty competitive markets and if their input costs are raised too much, other companies will run away with their contracts.
(Admittedly IBM is also in some cash-cow near-monopoly markets, like maintaining old IBM-brand mainframes for banks and the like, where the economics of the situation may be different from the orthodox neoclassical story I have tried to summarize above. It could well be that if the OS/360 experts unionized they could get away with really pushing wages up a bit beyond the competitive level.)
Businesses do improve the lives of the poor. When Wal-Mart finds a way of selling groceries and consumer goods at cheaper prices then poor people have better access to them. Good for Wal-Mart, good for the poor people.
But what Gates wants is very different. Gates appears to want government to spend tax money on behalf of the poor by giving it to businesses like his. This is the old fallacy that business is more efficient than government and therefore government can be made more effective if only it were run by businessmen. This completely misunderstands the reason why government is typically less efficient than business, which has nothing to do with the kind of people that run it and everything with the incentives they face.
Businesses face feedback. Their decisions are reflected in their balance sheets. Businesspeople may blame their success on their own personal greatness, but often there's nothing much more fancy going on than the kind of gradual evolutionary process you get when many entrepreneurs get to go on their hunches and instincts and they receive brutal and direct and accurate feedback from the market system.
Government doesn't have that. Now there are some things that you've got to have the government do, and those will typically be administered a little wastefully because of this, and there's nothing you can do about that, live with it, try to be smart about designing government programs and whenever possible keep them really simple.
But the idea that government programs should be run "like a business"--and I repeat, that's the old idea that this smells of--are hogwash, and often an excuse for getting government money in the pockets of business.
That was my rant, hope you enjoyed it!
I'm going to have to disagree with the general sentiment here. Variable pricing is actually a good idea. The question is not whether people should be made to pay for roads. They already pay for roads; if the roads weren't paid for, they wouldn't get built.
The question is whether how much you pay for roads depends on how and when you use them, or not.
Right now you get unlimited use of all roads, even at the most congested times, for a flat fee--the taxes that you pay no matter whether you actually use the roads or not. We all know that traffic would move if only not so many people drove during rush hour, but nobody wants to be the one to give up the use of the roads for the greater good of traffic flow, not even those who could most easily miss it.
Clearly, we need some mechanism to make sure that during rush hour, the people who absolutely need to use the road at that point in time get to drive. The congestion by itself scares some people away during rush hour, but by definition not enough. Now if we moved to a system where you had to pay a fee that was adjust to be just high enough to keep scare off enough people to keep traffic flowing, we could allow those people to use the roads that have demonstrated by their willingness to pay that they most need it. Obviously this is distorted a bit by the fact that some people are wealthier than others, but that is a problem of each and every good that you pay for, from toilet paper to chewing gum.
If you still aren't convinced, let me explain it differently. Imagine what would happen if we paid for gasoline the way we pay for roads. We would all get "free" gasoline at the gas stations, paid for from our taxes. There would still be only a limited amount of gasoline available, and a limited number of gas stations to get it at, so you'd have to get up early and stand in line for a long time if you wanted to get any gas. But there would be no reason for you to drive a fuel-efficient car, because once you got to the front of the line you could just fill up.
Then somebody might come up with the plan to charge for gas in proportion to how much people use, and snl2587 would exclaim "Genius!" and "Now there's a way to simulate the sagging economy! Have them pay more for commuting to work!"
Another one bites the dust! You can try to run a business just making things that people want and taking their money for it, but at some point somebody with power will smell your money and want a cut, if not of the money itself then at least of the influence that comes with it.
David Boaz wrote a wonderful little editorial when Google set up a lobbying office, and he provocatively called it "Parasite Economy Latches onto New Host." A little shrill, maybe, but not really all that bad an analogy.
And it is good to remember that the problem comes from two sides: companies lobby government for goodies, but at the same time government will do bad things to companies that fail to set up shop in DC (or Brussels, or wherever the nearest concentration of political power is.)
This is not true. Unions have power (in the US system) because when a simple majority of the workers that work for a given firm vote to have a union, the government then forces the company to negotiate with that union "in good faith," even if it could easily find workers elsewhere, and even if 49% of the workers don't want a union. There is a big difference between unions that are voluntary organizations for bargaining, and forced collective bargaining when a majority votes for it.
Arrangements like mandatory collective bargaining do not exist because society values them. They exist because some politicians passed a law at some point in the past and got away with it.
To me, the most important reason that I would like to see companies like Dell sell computers with Linux pre-installed is that this would force them to make sure the hardware is Linux compatible. I usually update to the new Ubuntu whenever it comes out, and I have learned to do so by just wiping the whole thing (I once fubar'ed everything with dist-upgrade); so a pre-installed computer would only save me from having to install Linux once. Besides, if the hardware is right, installing Linux only takes an hour or so. The main PITA is that computers ship with annoying pieces of Windows-only hardware (modems, WiFi cards, those sorts of things) or have bugs that happen to go unnoticed because Windows (or MacOS) doesn't use that particular feature (like faulty tables in ROMs of various sorts). Little things like that can turn installing Linux from a one hour operation into a one month operation, and I would rather like to know ahead of time that it is actually an operation that is possible at all. The manufacturer shipping the box with Linux is about as much assurance of that fact as I'm gonna want.
Fair use in the US is not defined as a percentage, or as anything else simple and definite. It's what lawyers call a "multi-prong balancing test": there are some factors that make a use more fair, some that make them less fair, and each of those have different "weights". That way, you can virtually always put up an argument either way, and the arguments sit on a long gliding scale from well-established doctrine to slightly avant-guarde speculation to outright crackpottery. The annoying thing is that, of course, this means you're never quite sure of your rights unless you actually get sued. As Larry Lessig has said about a proposal to introduce a similar fuzzy test for orphan works, lawyers love 5-factor balancing tests, because they invariably mean more (well-paid) work for the lawyers than simple rules. But my point was, please do not spread rumors about US copyright law allegedly allowing for N pages, or M minutes, or anything else that sounds like it actually makes sense, unless you have actually found a precedent (not just a statutory provision; those are regularly re-interpreted beyond recognition).
That's not always true. If the rules for when you do and do not get punished for wrongdoing are clear and the courts predictable and reliable (big if!), then punishing people after the fact for wrongdoing is a way of preventing wrongdoing. After all, if a you know that your wrongdoing will cost you money in the end, you'd be stupid to do it. (Unless, of course, the fines or damages aren't set high enough.)
For instance, the government can prohibit all the things that might cause traffic accidents, like talking with passengers, not being well rested, talking on cell phones, fiddling with the radio, and so on. It's enormously expensive to enforce those rules, and people will only stick to them if you do actually enforce them a lot. You need lots of police to patrol the streets all the time, and everybody has an incentive to be as uncooperative with them as they can.
On the other hand, if you just make people liable for crashing into each other, you only need to enforce that law in cases where it's very obvious that a crash happened, and you don't need oodles of police patrols just to check whether people are using cell phones. And moreover, you have victims of accidents who voluntarily are going to put time, money, and effort into prosecuting the cases, so you barely need to involve a police force at all. But best of all, just the threat of having to pay (or of losing eligibility for cheap insurance) will prevent people from driving recklessly in the first place.
And the sad thing is that CL is nowhere near NEW. In fact, I would say that the most interesting programming languages to study if you're interested in how using a better language can really improve software engineering, are Lisp (60s), Scheme (70s), OCaml/SML (80s), and Haskell (90s). Features in the so-called "academic" languages take time to trickle down in diluted forms into the mainstream, but much effort is now being made by their supporters to create workable and accessible development environments and library bindings, so that you can use metaprogramming, higher-order functions, polymorphic types and what-have-you to your advantage right now.
The idea of combining a unification of user and machine interfaces with a more expressive data model than byte-streams, thus removing the crufty and bug-prone conversion to and from text of every piece of data, is not new, let alone particular to Microsoft. It was quite fundamental to the Lisp Machine, and early Smalltalk based workstations, as well as to Project Oberon.
It is not just that it just so happens to be much like Linux or UNIX or BSD what makes OS X so much more palatable to the average "geek" than Windows; it is more that it adheres to some basic design principles reminiscent of UNIX systems. These include the everything-is-a-bitstream concept that enables you to interact automatically with any program that you can interact with interactively.
There are other ways of achieving that goal, but after the great OS shakedown of the 90s we are left with two basic models: *NIX, which is at times oversimplified, but can with a little trouble be automated arbitrarily, and therefore ends up displaying the kind of flexibility that "geeks" like; and Windows, which only lets itself be talked to through either a broken version of the old Mac "caveman interface" (point-and-grunt) or through immensely complicated library calls involving lots of weird structs and callbacks.
So to sum up, it's the basic ideas of extensibility and the common interface for interactive and automated use that define UNIX, much more so than the specific tools and scripts and commands, almost all of which are in continuous evolution and would be unrecognizable to someone familiar with a 1970s-era UNIX.
Apple has up till now had a pretty decent history of exposing any new functionality in OS X in the traditional, extensible (if cryptic) UNIX way (e.g. disktool), as well as in the traditional, inflexible (if intuitive) Mac way.
That is actually not true if by "the hardware half" you mean the procedures that are actually burned into a ROM (the program called the BIOS on IBM PC derivatives).
You can boot OS X in "verbose" mode (it's something like holding Ctrl-Super-Alt-Meta-Cokebottle-V, or maybe just Cmd-Opt-V, during the boot chime) and see the console output that the grey apple logo hides from you. It does come from OS X, and it is not the same as what you get if you run, say, Linux or BSD.
I do think that very likely the grey->blue transition you speak of marks the transition between the initialization of the kernel proper and the running of the init scripts, but I am not 100% sure of that.
And that is exactly what Apple was missing in OS X. Old versions of Mac OS always booted pretty quickly, and their customers are used to that. I once had that discussion with a Windows user who complained Linux took so long to boot; the average user does not like to leave their computer on, for whatever reason.
Besides, what I'd be interested in is how this compares to Sun's SMF (System Managament Facility), which is also supposed to be a replacement for the SysV init system.
Well, then DO RTFA.
Ada, as well as PL/1, COBOL, Java and similar atrocities, is "specified" by a several hundred page document designed by committee. In its case, a military committee. Its flaws were manifold and pointed out before the spec was even ready (by, among others, Edsger Dijkstra). I would not hesitate for a moment to agree with you that Ada did not fulfill its promise; nor, for that matter, did PL/1, COBOL, or Java.
BitC, on the other hand, as far as I can tell, turns out to be a tiny language, with a syntax based on Scheme and a semantics based on ML with a whiff of Pascal for the low-level details. It's description is terse and clear, and it seems to focus on fixing glaring problems with previous languages rather than making the programmer treat their work like a form in triplicate.
They mention explicitly that the syntax is just a surface-level thing, and that they may well support alternative "sugared" syntax later on.
Aside from that, there is one very great advantage to LISP-style syntax, which is that any program finds an obvious representation in a data structure that is primitive in the language's runtime system. Therefore, you can very easily build powerful macro systems and other source-code manipulating tools.
That may sound arcane, but anyone who has done some serious programming in a LISP dialect can tell you that real source-to-source macros (as opposed to C's textual-substitution macros) can greatly simplify programs, and eliminate the need for bloating the language standard with control structures that could easily be defined by the user.
To see that the need for macros in operating systems is apparent, one only needs to look at the Linux kernel, which bends gcc macrology over backwards in many places.
No, Multics did not have orthogonal persistence. Also, as it says in the linked page, Coyotos is different from its predecessor EROS exactly in that it does not have orthogonal persistence.
I am a foreign student. The Dept. of Homeland Security maintains a database called SEVIS. My school is required to keep them updated with the following information (from ice.gov):
* Name
* Place and date of birth
* Country of Citizenship
* Address
* Status (full-time, part-time)
* Date of commencement of studies
* Degree program and field of study
* Practical training, beginning and ending dates
* Termination date and known reasons
* I-20 and application for I-20
* Number of credits completed per year
The information needs to updated pretty much instantaneously whenever it changes.
No, it can't be automated. However, many things that can't be automated can still be eliminated by automation.
Repairing typewriters requires pretty much the same amount of skill and time now as it did 30 years ago. Typewriter repairmen nonetheless lost their jobs. Nobody used typewriters.
Well-designed computer systems are easier to figure out and fail less often, so they require fewer calls to the helpdesk, whose employees are not helped much by the fact that their job can't be automated.
Now to reassure those who work at help desks: an extremely small amount of progress has been made on the question of how to build the kind of systems that would require fewer help-desk calls.
It is not even clear that the market encourages such systems. The problem is that the only known technique proven to make systems more reliable and easier to use is to Keep It Simple. Corporations have a very strong tendency to instead Keep It Cheap.
For instance, modems used to work. They were simple, self-contained and very well-debugged systems with a simple and well-defined interface consisting of an RS232 cable and a straightforward text-based command protocol. Then, winmodems were introduced, which cut cost ever so minimally by doing all the digital signal processing in software. The damage has still not been completely repaired.
Caltech security is one of those things you need a lot of Zen and zzzzzzzz to understand. Are you allowed inside the steam tunnels as and undergrad student? Does a dog have the Buddha nature? Mu. It's not so much about being _allowed_, that is a dichotomy that is determined by the evil "General Counsel", a.k.a. the Institute Lawyers, who are afraid of asbestos suits. The question is whether anyone cares, and the answer is no. Campus security will politely ask you to leave if they ever see you in there, which is very unlikely because they rarely go in themselves, and when they do, it is understood that you can walk right back in. Such is doublethink, such is The Institute.
Uhuh. If more than one explanation is consistent with the facts, then there is good reason to go with the simplest. In this experiment, they supposedly want to gather some of these facts. As long as we can describe what we observe equally well using standard GR as using all the fancy string stuff, I agree there is little reason to use much beyond standard GR. In order, however, to think of facts that might possibly disprove standard GR, we need people to come with those fancy theories and see where their predictions are different.
If that's truly a South Park episode, then they ripped it off of Rowan Atkinson, the British comedian known for his Mr. Bean show. Except in his version, the jews were right: Atkinson's "Hell" sketch
Slashdot Mirror
Wait... so you are trying to say that the Catholic Church is not a scam?
Hmmm... interesting.
And I always thought that saying that you'll be sent to hell by an invisible guy and his long-haired Jewish son who died but somehow isn't dead and becomes a cracker sometimes, unless you join the church and put money on the collection plate was... a bit of a scam.
But I might be wrong.
IBM cannot afford to pay people much more than they contribute to the IBM bottom line, simply because if they did that, they'd lose money.
IBM cannot afford to pay people much less than they contribute to the IBM bottom line, simply because if you can contribute $X to the IBM bottom line then you can likely also contribute $X to some other company's bottom line, and you might jump ship.
This sets pretty hard boundaries on how much pay IBM can and cannot offer. True, IBM happens to have a strong market position and probably a bit more flexibility with this than other firms, but if the overtime pay really meant a raise big enough that they had to cut wages 15% just to compensate, then I suggest that it may very well be the case that they could not afford, in their competitive business, not to make that cut. They surely won't do it very lightly or just for fun, seeing as it is not likely to be good for employee morale, which is very important in a knowledge-intensive service industry like IT consulting (their core business these days.)
BTW, in case anybody wants to say that IBM makes a profit and that it could therefore pay their workers more and cut into that profit: that's not how it works. Profits are returns on capital. Capital doesn't fall out of thin air. It must be gotten on capital markets, like bond and stock markets, and investors expect a certain return. If IBM has a higher return than that, then the price of the stock will go up so as to equalize the return per dollar invested anyway.
Really, the boundaries within which a company like IBM has to maneuver this kind of thing are pretty tight. Unionizing the shop ain't gonna help much either, because much of what IBM does is in pretty competitive markets and if their input costs are raised too much, other companies will run away with their contracts.
(Admittedly IBM is also in some cash-cow near-monopoly markets, like maintaining old IBM-brand mainframes for banks and the like, where the economics of the situation may be different from the orthodox neoclassical story I have tried to summarize above. It could well be that if the OS/360 experts unionized they could get away with really pushing wages up a bit beyond the competitive level.)
Businesses do improve the lives of the poor. When Wal-Mart finds a way of selling groceries and consumer goods at cheaper prices then poor people have better access to them. Good for Wal-Mart, good for the poor people. But what Gates wants is very different. Gates appears to want government to spend tax money on behalf of the poor by giving it to businesses like his. This is the old fallacy that business is more efficient than government and therefore government can be made more effective if only it were run by businessmen. This completely misunderstands the reason why government is typically less efficient than business, which has nothing to do with the kind of people that run it and everything with the incentives they face. Businesses face feedback. Their decisions are reflected in their balance sheets. Businesspeople may blame their success on their own personal greatness, but often there's nothing much more fancy going on than the kind of gradual evolutionary process you get when many entrepreneurs get to go on their hunches and instincts and they receive brutal and direct and accurate feedback from the market system. Government doesn't have that. Now there are some things that you've got to have the government do, and those will typically be administered a little wastefully because of this, and there's nothing you can do about that, live with it, try to be smart about designing government programs and whenever possible keep them really simple. But the idea that government programs should be run "like a business"--and I repeat, that's the old idea that this smells of--are hogwash, and often an excuse for getting government money in the pockets of business. That was my rant, hope you enjoyed it!
I'm going to have to disagree with the general sentiment here. Variable pricing is actually a good idea. The question is not whether people should be made to pay for roads. They already pay for roads; if the roads weren't paid for, they wouldn't get built.
The question is whether how much you pay for roads depends on how and when you use them, or not.
Right now you get unlimited use of all roads, even at the most congested times, for a flat fee--the taxes that you pay no matter whether you actually use the roads or not. We all know that traffic would move if only not so many people drove during rush hour, but nobody wants to be the one to give up the use of the roads for the greater good of traffic flow, not even those who could most easily miss it.
Clearly, we need some mechanism to make sure that during rush hour, the people who absolutely need to use the road at that point in time get to drive. The congestion by itself scares some people away during rush hour, but by definition not enough. Now if we moved to a system where you had to pay a fee that was adjust to be just high enough to keep scare off enough people to keep traffic flowing, we could allow those people to use the roads that have demonstrated by their willingness to pay that they most need it. Obviously this is distorted a bit by the fact that some people are wealthier than others, but that is a problem of each and every good that you pay for, from toilet paper to chewing gum.
If you still aren't convinced, let me explain it differently. Imagine what would happen if we paid for gasoline the way we pay for roads. We would all get "free" gasoline at the gas stations, paid for from our taxes. There would still be only a limited amount of gasoline available, and a limited number of gas stations to get it at, so you'd have to get up early and stand in line for a long time if you wanted to get any gas. But there would be no reason for you to drive a fuel-efficient car, because once you got to the front of the line you could just fill up.
Then somebody might come up with the plan to charge for gas in proportion to how much people use, and snl2587 would exclaim "Genius!" and "Now there's a way to simulate the sagging economy! Have them pay more for commuting to work!"
Another one bites the dust! You can try to run a business just making things that people want and taking their money for it, but at some point somebody with power will smell your money and want a cut, if not of the money itself then at least of the influence that comes with it. David Boaz wrote a wonderful little editorial when Google set up a lobbying office, and he provocatively called it "Parasite Economy Latches onto New Host." A little shrill, maybe, but not really all that bad an analogy. And it is good to remember that the problem comes from two sides: companies lobby government for goodies, but at the same time government will do bad things to companies that fail to set up shop in DC (or Brussels, or wherever the nearest concentration of political power is.)
This is not true. Unions have power (in the US system) because when a simple majority of the workers that work for a given firm vote to have a union, the government then forces the company to negotiate with that union "in good faith," even if it could easily find workers elsewhere, and even if 49% of the workers don't want a union. There is a big difference between unions that are voluntary organizations for bargaining, and forced collective bargaining when a majority votes for it.
Arrangements like mandatory collective bargaining do not exist because society values them. They exist because some politicians passed a law at some point in the past and got away with it.
To me, the most important reason that I would like to see companies like Dell sell computers with Linux pre-installed is that this would force them to make sure the hardware is Linux compatible. I usually update to the new Ubuntu whenever it comes out, and I have learned to do so by just wiping the whole thing (I once fubar'ed everything with dist-upgrade); so a pre-installed computer would only save me from having to install Linux once. Besides, if the hardware is right, installing Linux only takes an hour or so. The main PITA is that computers ship with annoying pieces of Windows-only hardware (modems, WiFi cards, those sorts of things) or have bugs that happen to go unnoticed because Windows (or MacOS) doesn't use that particular feature (like faulty tables in ROMs of various sorts). Little things like that can turn installing Linux from a one hour operation into a one month operation, and I would rather like to know ahead of time that it is actually an operation that is possible at all. The manufacturer shipping the box with Linux is about as much assurance of that fact as I'm gonna want.
Fair use in the US is not defined as a percentage, or as anything else simple and definite. It's what lawyers call a "multi-prong balancing test": there are some factors that make a use more fair, some that make them less fair, and each of those have different "weights". That way, you can virtually always put up an argument either way, and the arguments sit on a long gliding scale from well-established doctrine to slightly avant-guarde speculation to outright crackpottery. The annoying thing is that, of course, this means you're never quite sure of your rights unless you actually get sued. As Larry Lessig has said about a proposal to introduce a similar fuzzy test for orphan works, lawyers love 5-factor balancing tests, because they invariably mean more (well-paid) work for the lawyers than simple rules. But my point was, please do not spread rumors about US copyright law allegedly allowing for N pages, or M minutes, or anything else that sounds like it actually makes sense, unless you have actually found a precedent (not just a statutory provision; those are regularly re-interpreted beyond recognition).
That's not always true. If the rules for when you do and do not get punished for wrongdoing are clear and the courts predictable and reliable (big if!), then punishing people after the fact for wrongdoing is a way of preventing wrongdoing. After all, if a you know that your wrongdoing will cost you money in the end, you'd be stupid to do it. (Unless, of course, the fines or damages aren't set high enough.)
For instance, the government can prohibit all the things that might cause traffic accidents, like talking with passengers, not being well rested, talking on cell phones, fiddling with the radio, and so on. It's enormously expensive to enforce those rules, and people will only stick to them if you do actually enforce them a lot. You need lots of police to patrol the streets all the time, and everybody has an incentive to be as uncooperative with them as they can.
On the other hand, if you just make people liable for crashing into each other, you only need to enforce that law in cases where it's very obvious that a crash happened, and you don't need oodles of police patrols just to check whether people are using cell phones. And moreover, you have victims of accidents who voluntarily are going to put time, money, and effort into prosecuting the cases, so you barely need to involve a police force at all. But best of all, just the threat of having to pay (or of losing eligibility for cheap insurance) will prevent people from driving recklessly in the first place.
And the sad thing is that CL is nowhere near NEW. In fact, I would say that the most interesting programming languages to study if you're interested in how using a better language can really improve software engineering, are Lisp (60s), Scheme (70s), OCaml/SML (80s), and Haskell (90s). Features in the so-called "academic" languages take time to trickle down in diluted forms into the mainstream, but much effort is now being made by their supporters to create workable and accessible development environments and library bindings, so that you can use metaprogramming, higher-order functions, polymorphic types and what-have-you to your advantage right now.
The idea of combining a unification of user and machine interfaces with a more expressive data model than byte-streams, thus removing the crufty and bug-prone conversion to and from text of every piece of data, is not new, let alone particular to Microsoft. It was quite fundamental to the Lisp Machine, and early Smalltalk based workstations, as well as to Project Oberon.
It is not just that it just so happens to be much like Linux or UNIX or BSD what makes OS X so much more palatable to the average "geek" than Windows; it is more that it adheres to some basic design principles reminiscent of UNIX systems. These include the everything-is-a-bitstream concept that enables you to interact automatically with any program that you can interact with interactively.
There are other ways of achieving that goal, but after the great OS shakedown of the 90s we are left with two basic models: *NIX, which is at times oversimplified, but can with a little trouble be automated arbitrarily, and therefore ends up displaying the kind of flexibility that "geeks" like; and Windows, which only lets itself be talked to through either a broken version of the old Mac "caveman interface" (point-and-grunt) or through immensely complicated library calls involving lots of weird structs and callbacks.
So to sum up, it's the basic ideas of extensibility and the common interface for interactive and automated use that define UNIX, much more so than the specific tools and scripts and commands, almost all of which are in continuous evolution and would be unrecognizable to someone familiar with a 1970s-era UNIX.
Apple has up till now had a pretty decent history of exposing any new functionality in OS X in the traditional, extensible (if cryptic) UNIX way (e.g. disktool), as well as in the traditional, inflexible (if intuitive) Mac way.
That is actually not true if by "the hardware half" you mean the procedures that are actually burned into a ROM (the program called the BIOS on IBM PC derivatives).
You can boot OS X in "verbose" mode (it's something like holding Ctrl-Super-Alt-Meta-Cokebottle-V, or maybe just Cmd-Opt-V, during the boot chime) and see the console output that the grey apple logo hides from you. It does come from OS X, and it is not the same as what you get if you run, say, Linux or BSD.
I do think that very likely the grey->blue transition you speak of marks the transition between the initialization of the kernel proper and the running of the init scripts, but I am not 100% sure of that.
And that is exactly what Apple was missing in OS X. Old versions of Mac OS always booted pretty quickly, and their customers are used to that. I once had that discussion with a Windows user who complained Linux took so long to boot; the average user does not like to leave their computer on, for whatever reason.
Besides, what I'd be interested in is how this compares to Sun's SMF (System Managament Facility), which is also supposed to be a replacement for the SysV init system.
Yes, the US has a minimum wage, although it is not very high. The state of California sets its own, higher minimum wage.
:
From http://www.dir.ca.gov/dlse/FAQ_MinimumWage.htm
Beginning January 1, 2002, the minimum wage in California is $6.75 per hour.
For sheepherders, however, [...]
It also seems that usually the employee has to file a complaint before the state starts to care.
Well, then DO RTFA. Ada, as well as PL/1, COBOL, Java and similar atrocities, is "specified" by a several hundred page document designed by committee. In its case, a military committee. Its flaws were manifold and pointed out before the spec was even ready (by, among others, Edsger Dijkstra). I would not hesitate for a moment to agree with you that Ada did not fulfill its promise; nor, for that matter, did PL/1, COBOL, or Java. BitC, on the other hand, as far as I can tell, turns out to be a tiny language, with a syntax based on Scheme and a semantics based on ML with a whiff of Pascal for the low-level details. It's description is terse and clear, and it seems to focus on fixing glaring problems with previous languages rather than making the programmer treat their work like a form in triplicate.
They mention explicitly that the syntax is just a surface-level thing, and that they may well support alternative "sugared" syntax later on. Aside from that, there is one very great advantage to LISP-style syntax, which is that any program finds an obvious representation in a data structure that is primitive in the language's runtime system. Therefore, you can very easily build powerful macro systems and other source-code manipulating tools. That may sound arcane, but anyone who has done some serious programming in a LISP dialect can tell you that real source-to-source macros (as opposed to C's textual-substitution macros) can greatly simplify programs, and eliminate the need for bloating the language standard with control structures that could easily be defined by the user. To see that the need for macros in operating systems is apparent, one only needs to look at the Linux kernel, which bends gcc macrology over backwards in many places.
No, Multics did not have orthogonal persistence. Also, as it says in the linked page, Coyotos is different from its predecessor EROS exactly in that it does not have orthogonal persistence.
There already is such a database.
I am a foreign student. The Dept. of Homeland Security maintains a database called SEVIS. My school is required to keep them updated with the following information (from ice.gov):
* Name
* Place and date of birth
* Country of Citizenship
* Address
* Status (full-time, part-time)
* Date of commencement of studies
* Degree program and field of study
* Practical training, beginning and ending dates
* Termination date and known reasons
* I-20 and application for I-20
* Number of credits completed per year
The information needs to updated pretty much instantaneously whenever it changes.
No, it can't be automated. However, many things that can't be automated can still be eliminated by automation.
Repairing typewriters requires pretty much the same amount of skill and time now as it did 30 years ago. Typewriter repairmen nonetheless lost their jobs. Nobody used typewriters.
Well-designed computer systems are easier to figure out and fail less often, so they require fewer calls to the helpdesk, whose employees are not helped much by the fact that their job can't be automated.
Now to reassure those who work at help desks: an extremely small amount of progress has been made on the question of how to build the kind of systems that would require fewer help-desk calls.
It is not even clear that the market encourages such systems. The problem is that the only known technique proven to make systems more reliable and easier to use is to Keep It Simple. Corporations have a very strong tendency to instead Keep It Cheap.
For instance, modems used to work. They were simple, self-contained and very well-debugged systems with a simple and well-defined interface consisting of an RS232 cable and a straightforward text-based command protocol. Then, winmodems were introduced, which cut cost ever so minimally by doing all the digital signal processing in software. The damage has still not been completely repaired.
And then, if more people hack darcs, more people will know Haskell, and the world will be a better place.
Caltech security is one of those things you need a lot of Zen and zzzzzzzz to understand. Are you allowed inside the steam tunnels as and undergrad student? Does a dog have the Buddha nature? Mu. It's not so much about being _allowed_, that is a dichotomy that is determined by the evil "General Counsel", a.k.a. the Institute Lawyers, who are afraid of asbestos suits. The question is whether anyone cares, and the answer is no. Campus security will politely ask you to leave if they ever see you in there, which is very unlikely because they rarely go in themselves, and when they do, it is understood that you can walk right back in. Such is doublethink, such is The Institute.
Hell, I would hope that they have made their way out of the 1960s and use a bignum for that...
Uhuh. If more than one explanation is consistent with the facts, then there is good reason to go with the simplest. In this experiment, they supposedly want to gather some of these facts. As long as we can describe what we observe equally well using standard GR as using all the fancy string stuff, I agree there is little reason to use much beyond standard GR. In order, however, to think of facts that might possibly disprove standard GR, we need people to come with those fancy theories and see where their predictions are different.
If that's truly a South Park episode, then they ripped it off of Rowan Atkinson, the British comedian known for his Mr. Bean show. Except in his version, the jews were right: Atkinson's "Hell" sketch