And worse yet- there's nothing India wants anymore... what the hell could we make here to sell to India that China can't make for 1/100th the labor cost?
Right now, China and India cannot make CPUs (Intel and AMD), large airplanes (Boeing), pharmaceuticals (Pfizer, Merck, etc), fabrication equipment, large multiway servers (Sun, IBM, HP, Fujitsu), windows-compatible OSes, capital goods of various kinds, or even adequate food to feed their populations, which is why the U.S. is the world's largest agricultural exporter. Note that all the items on that list (except the last one) are much more profitable and pay much higher wages than what China and India export to us.
what the hell could we make here to sell to India that China can't make for 1/100th the labor cost?
If China and India didn't want to buy anything from us at our inflated prices then they would stop exporting to us. They're not just giving us products for free--exports exist to pay for imports. Even when Indians and Chinese buy U.S. assets or debt, it just means they're delaying (and slightly increasing) our exports to them.
At the same time, federal policy has been to do the same thing at a national level. Because our so-called "free trade" policies have left us with an over $700 billion annual trade deficit, other countries are sitting on huge piles of the dollars we gave them to buy their stuff (via Wal-Mart and other "low cost" retailers). But we no longer manufacture anything they want to buy with those dollars.
If foreigners have exported products to us and don't buy anything from us in return, then they've given us free products.
Exports exist to pay for imports. Foreigners export products to the U.S. because they wish to buy capital-intensive American products. Foreigners do not export products because they wish to give us things for free.
Even when foreigners buy debt (instead of our products), that means only that our exports of products to them are delayed. And when foreigners buy American companies, that means they're investing in our economy rather than their own.
Today, foreigners awash with our consumer dollars are on a two-decades-long buying spree. The UK's BP bought Amoco for $48 billion - now Amoco's profits go to England. Deutsche Telekom bought VoiceStream Wireless, so their profits go to Germany, which is where most of the profits from Random House, Allied Signal, Chrysler, Doubleday, Cyprus Amax's US Coal Mining Operations, GTE/Sylvania, and Westinghouse's Power Generation profits go as well. Ralston Purina's profits go to Switzerland, along with Gerber's; TransAmerica's profits go to The Netherlands, while John Hancock Insurance's profits go to Canada. Even American Bankers Insurance Group is owned now by Fortis AG in Belgium.
As many European and Japanese companies have been bought by American firms. For example, Volvo, Saab, Mazda, Jaguar, Lambourghini, Maserati, and other car manufacturers are now owned by U.S. companies--and that's limiting our attention just to car companies.
Just because a company is owned by a Swiss firm doesn't mean its profits go to Switzerland.
So what if Europeans are investing here. Phrased another way, there has been a massive capital flight from Europe to America. Investment is leaving Europe at a rapid pace and setting up shop in the U.S!
Is it really such a problem if foreigners would rather invest in our economy than in theirs?
Actually the money can just disappear. Money only has value that people assign to it. When you place large sums of money in the hands of people who are so rich that it has no real value, all of the value of the money just disappears like magic. The same would be true for any party getting money without a process that makes them view it as valuable.
Rich people can either spend their money or invest it. If they spend it then presumably they enjoy the consumption (otherwise why spend it?) while providing jobs, and if they invest it then they contribute to the construction of factories, companies, etc. In neither case does the money just "disappear", solely because the rich value a dollar less than the poor.
The saddest reality of the "current order" is that it is systematically destroying more wealth than can be imagined. Our world is being impoverished by this process and the desperate people are what we see in the "war on terror."
World economic growth during the 1990s was almost 5%, which was the highest growth rate per decade ever. So we have not "systematically destroyed more wealth than can be imagined" or "impoverished the world."
Our world is being impoverished by this process and the desperate people are what we see in the "war on terror." There are billions of desperate people and it is getting much more so and is spreading over entire nations.
Studies of terrorists and suicide bombers indicate they they're overwhelmingly from upper middle class households in oil-rich countries like Saudi Arabia. None of the Sept 11th hijackers came from poverty, and their stated reasons for the hijackings did not include poverty.
There are billions of desperate people and it is getting much more so and is spreading over entire nations.
Median per capita real income is higher in the 3rd world than it has been at any time previously, and it's growing faster than at any time previously.
How many of you are making more money because of all the people in China, India, and other cheap-labor locales, who buy stuff that you produce? To vote, Click here... Now, how many of you know somebody who lost their job because of overseas competition? To vote, Click here
Um, you forgot one poll: how many of you are wealthier in real terms because the prices of the products you buy have been reduced by overseas labor? The answer would be: far more than have lost their jobs to outsourcing.
After all, the $150 Japanese DVD players are now competing against $35 Chinese players. And the cost of clothing in real terms has declined by 40% over the last decade.
Unfortunately, if you had such poll asking if people had their prices reduced, then everyone would have to answer in the affirmative if they were being honest. (And don't say "prices go up because of inflation" since we're talking about prices in real terms, and your salary goes up with inflation).
I hear this protest being regurgitated by everyone who advocates unbalanced trade.
Heh. Trade is always balanced. Imports==exports--it's the result of an accounting identity. The only reason we hear of a "trade gap" is because the people compiling those statistics do not include T-bond purchases by foreigners as exports.
So I go back to school (that's assuming I can afford it...most Americans have to suport families and probably can't) and learn $New_Thing. Before I can even get hired with the skills I learned for $New_Thing, Cambodia starts doing $New_Thing at wage-slave rates, and a new cycle of unbalanced trade starts. I go to the Slashdot forums and read a post by someone who thinks he's an economic genius that reads: People who can't find jobs doing $New_Thing go on to find other kinds of jobs, those that are more profitable.
If you can't afford to take time off from work in order to study, why do you think people in Cambodia would be able to? Do they have enormous savings to draw upon? And don't say "the Cambodian government pays for it" because the Cambodian government has very little money and could not afford to pay for any more schooling than it does at present; if it raised taxes then some of its citizens who pay those taxes would be reduced to starvation.
So I go back to school and learn $New_Thing.2... You know...in programming we learn about these things called "loops"....
The programming profession has been going strong since the 1960s. It's still going strong: right now there's a 95% employment rate among techies and they make ~$80,000/yr.
At this rate the "loop" you describe requires about 100 years between iterations, and everyone's wages are raised by going through the iteration.
On the contrary, from what I've gathered from history, proper socialism has never been given a chance to sustain itself because the frightened forces of the upper classes in capitalist countries have always fought to sabotage and destabilize the governments and economies of countries that try socialist experiments.
That tired excuse has been used so many times that it's difficult to count them. "Proper socialism" has been tried countless times--the result is always disaster, and the blame is always placed on outsiders (Trotskyites!) who are allegedly sabotaging the socialist experiment. What about Cuba and North Korea? They're cut off from the world entirely, and they have nobody to blame but themselves.
And don't just claim that all examples so far haven't been "true" socialism. If no true socialism has ever prevailed, then you must wonder why every single socialist experiment ends up degenerating into something else.
Capitalism only raised the standard of living for the upper classes in America. The rest came from social struggles from the lower classes (unions, strikes, etc.) and from the threat of true economic revolution. To keep the masses happy, small reforms were allowed so as to undercut the power of radical groups that sought to actually destroy the system.
Capitalism drastically raised the standard of living for everyone. Unions had little or nothing to do with it. Union participation has been decreasing since the 1930s and the standard of living for the middle class has increased dramatically since then.
Unions and strikes caused little or no increase in the standard of living for the middle classes in America. The vast majority of the increases in wages have happened to professions which are not unionised and which are not affected by wage legislation.
Our country's massive and obscene wealth is built upon our businesses preying upon the blood and sweat of the poor in third-world countries.
Our country was rich before it traded with the 3rd world in any substantial form. The vast majority of the trade with the 3rd world we currently conduct is a result of reductions in trade barriers since the 1980s. Trade with the 3rd world was very slight during the first half of the 20th century--about all we imported was fruit from central america.
Our political and economic power acts like a siphon that sucks raw wealth from weak countries into our deep pockets.
3rd world countries are poor because they're capital-depleted and their labor productivity is very low. They have extremely little wealth to siphon off. They were poor countries before the U.S. showed up.
...The phenomenon of 3rd world nations trading with wealthier nations, has always and invariably increased the real wages of 3rd world workers. As an example, the real median wage in China has more than quadrupled in the last 25 years since they abandoned the foolish prescriptions which you espouse.
But since China's minimum wage is 1/100th of ours-
China's minimum wage is not 1/100th of ours. In fact the PPP-adjusted median income in China is around $6,000 compared to around $40,000 in the U.S. Even non-PPP adjusted, the figures are nowhere near what you indicate.
But- here's the big one- China has a billion workers to feed. They have a virtually unlimited supply of labor.
China does not have a billion workers. A large proportion of its population consists of children, the elderly, students, dependents, housewives, and so on. But that's beside the point...
So they can make all the computers and cars THEY need, PLUS enough for export to swamp the markets of the United States and Europe with computers and cars, and drive the native manufacturers in the US and Europe out of business entirely, of both computers and cars.
The purpose of exports is to pay for imports. The Chinese people export cheap goods to the United States because they wish to purchase American capital-intensive goods (like Boeing airplanes, Intel chips, pharmaceuticals, capital goods, fabrication equipment, etc) which they cannot produce indigenously.
If China were to "flood the US markets with computers and cars" without buying anything from the US in return, then the Chinese have given the US free computers and cars without getting anything in return.
Ricardo was an idiot and if he were alive today I'd shoot him just for proposing such a dangerous piece of sheer propaganda and lies.
You wish to shoot Ricardo because he wrote something you disagree with? You certainly live up to your handle, Mr Marxist! Perhaps you should read what Ricardo wrote before shooting him.
(I'm continually surprised by which posts get modded up...)
The oil and natural gas we use to generate electricity to power devices that require copper will become too expensive to use long before we run out of the copper we use in the construction of these devices.
Oil and gas aren't used for generating electricity in the United States. Oil is used for powering automobiles (and for other uses), and gas is used mostly for heating homes.
Electricity in this country is generated primarily from coal (60%), nuclear (20%), and hydroelectric (>10%). The U.S. has enough coal to last for centuries (environmental damage notwithstanding), and enough Uranium to last for millenia (if breeder reactors are used).
Even if both of those resources were completely exhausted, or were nearing exhaustion, there are many alternatives which are only modestly more expensive. Wind farms, geothermal plants, solar panels, tidal energy plants, methane hydrates, and others.
Even if all sources of power were to become scare (an impossibility), the first thing sacrificed would be air conditioning for the home, which consumes more electricity by far than computers and electronic equipment.
...But let's suppose that all coal and uranium were exhausted, and all alternatives became scarce, and home air conditioning had already been abandoned, and still there was not enough electricity to power computers. Then we could build desktop and server computers with power-saving technology similar to that used in notebooks today. At 45nm fabbing, we could make processors that are faster than today's but which consume less than 1W. 1W could be easily generated by power plants which harvest methane from pig farts, just like in "Mad Max: Beyond Thunderdome." If we used the Thunderdome approach, the only remaining possible source of scarcity would be: if pigs learned to fly.
As a result, it seems unlikely that power will become scarce enough in the near future to render cabling unimportant.
Relying on nuclear power in the light of dwindling fossil fuel reserves is a very short-sighted approach. At the current rate of consumption, there is only enough Uranium on the planet for the next 50 years -- somewhat more if you start using more expensive, lower-quality reserves. So the problem is really just shifted into the future by a very small number of years, compared to human history or the history of the planet as a whole.
That's not exactly correct. Although there's only enough Uranium-235 for 50 years of production, U-235 constitutes only 0.7% of naturally occurring Uranium. Uranium-238, on the other hand, constitutes 99.3% of naturally occurring Uranium and would last for thousands of years.
Granted, using U-238 as fuel would require breeder reactors, which are virtually non-existent right now. But breeder reactors are feasible and proven; France ran a commercial 1200MW breeder reactor for 15+ years. Although France's breeder reactor had various technological glitches, it did work, and it was the very first commercial breeder reactor. The first iteration of anything has technological glitches.
Also bear in mind, that the estimates of Uranium availability only take into account the conventional, terrestrial reserves. Those conventional reserves constitute only a tiny part of global Uranium--for example, the amount of Uranium dissolved in ocean water is many times higher than the conventional reserves. Although Uranium dissolved in ocean water is too diffuse to be profitably extracted right now, we could probably figure out better ways of extracting it in the next few thousand years. Who knows, by then, 5000 years from now, we may even be able to mine Uranium from the asteroid belt, which is right next door, cosmically speaking.
With breeder reactors, we may not ever run out of Uranium. Even if we did run out, we'd have a very long time beforehand to consider next steps.
Please don't associate this retard and Sun's opinions on "very poor language design" with "academics."
I'm afriad you're the only retard. In fact, after reading your complaints, I'm amazed how retarded you are.
The verdict of poor language design was not Sun's (nor did my post imply that) but rather a consensus of language designers. That consensus is so widely shared that Stroustrup has to issue apologies on occasion...
Java lacked descriminating unions.
You have to be joking. Those are desirable? Also, you should learn how to spell better than a 3rd grader...
Java lacked parametric polymorphism.
Java has parametric polymorphism. That topic has been discussed already....
Java had "native types" and classes
C++ has native types and classes.
But I assume you're complaining about the split between primitive types and their object wrappers in Java (int vs Integer, etc). That was the most elegant way of solving various problems without having an "everything is an object" language.
Java had constant references to mutable instances, but no constant instances.
You mean it lacks the hideously overloaded "const" keyword of C++?
Java had a completely superfluous 'new' keyword. Java forced manifest typing even when local type inference was tractable. Java lacked unsigned types.
Those are EXTREMELY minor concerns. You honestly have difficulty typing the 'new' keyword? And your complaint is that it could theoretically have been omitted from the language?
Java lacked a full tower of numeric types, but had a limited number that can't be used with normal math operators.
C++, C, and most languages lack a complete set of numeric types primitive to the langauge. C++ has recently added additional numeric types like complex (they weren't there for the first 15+ years), but they're in class libraries, and similar libraries are available for Java.
Java had undead objects revived by finalizers.
This is an obscure concern that almost never comes up in practice.
Trust me, I can go on.
I must say, there are a great many legitimate criticisms of Java. But you haven't hit on even one of them. Your list was a compilation of incorrect and trivial criticisms.
Trust me, I can go on.
I have no doubt that you could "go on" and demonstrate your ignorance still further.
Java is, from a language design standpoint, a joke.
Java is widely recognized among language designers as being one of the cleaner mass-deployed languages. It's a well-designed though imperfect language. However it has defects, of course....
It was designed by cutting chunks off of it's predecessors and not supplying any viable substitutes.
Those chunks were removed precisely because they were not orthogonal. Those chunks/features were simply overlapping. As such other viable substitutes already existed. Nothing was removed from Java which would prevent any operation.
Bear in mind that most of the "features" not included from C++ were examples of very poor language design.
There are some omissions from Java which are unfortunate, like pre-conditions and post-conditions, however most of the omissions were not present in Java's predecessors.
The only features of C++ which are lacking in Java and which arguably should be included are: operator overloading, and some variant of the const keyword (although not exactly like C++ const, which is hideously overloaded).
People always brag about their million-line.net1.1 and Java2 projects, claiming that such million-line monstrocities are proof the language is maintainable.
No serious programmer ever brags about million-line programs, or claims that lines of code somehow correlates positively with maintainability. You're attributing a point of view to "people" which they don't possess. Perhaps some friend of yours made a comment like that, which you now attribute to "people"?
What they don't tell you is that 900 000 of those million lines are spent on simple get/set wrappers, typecasting containers, re-implementing containers missing from the library, and recoding functions for different datatypes, giant class-based switch statements
Typecasting containers is not done in Java, since it has generic containers. Re-implementing missing containers is very rare since the Java API has a wide variety of containers already (some even argue the API is too big). Recoding functions for different datatypes is unnecessary since Java has always been polymorphic. And there's nothing wrong with switch statements; they're syntactic sugar for if/else (I assume you're not opposed to if/else).
I'll grant that the repetitive get/set wrappers are unfortunate and unnecessary. I don't understand why the designers of Java don't just borrow ideas from Eiffel on this matter. I realize the Java designers are opposed to syntactic sugar in general, but in this case it would be justified, since a huge proportion of Java code wasted is on statements like
"public Foo getFoo() { return this.foo; }".
Wrong, because it's the normal coder and user that makes the choice, not the malicious coder. He is limited to exploit whatever is being used. If my mail server is implemented in Java, the attacker has to find a hole in that Java code or Java/VM itself to gain entry via the mail server. (And there is a mail server written in Java, BTW.)
No, for several reasons.
If my mail server is implemented in Java, the attacker has to find a hole in that Java code or Java/VM itself to gain entry via the mail server.
But the malicious coder can gain entry through any service running on the system, not just the mail server. All services running on the machine must be written in verifiable bytecode to prevent intrusion.
So perhaps the programmers writing daemons should use only Java, and perhaps users should run only Java services. That is half way to the solution.
What about buggy or malicious drivers, protocl stacks, or loadable kernel modules? Users install those programs from CDs or from net downloads, and those programs run in ring 0 which means they can overwrite any memory on the system--they could even corrupt the JVM which previously had been secure. Is it really safe to have the operating system accept modifications to itself, without being able to verify that those modifications won't corrupt the rest of the operating system? Indeed, Microsoft claims that >85% of system crashes are caused by bad drivers written by 3rd party companies...
No. To prevent intrusions, all programs running on the system must not have these problems, which means it must be enforced by the operating system.
> > These facilities make it impossible for any application to have > > buffer overruns, segfaults, or overruns of other apps' data > > -- as a result, all applications can run in ring 0 and virtual > > memory is not required. > > Which is complete nonsense, because there are many other classes of > security problems, some of them application-specific, other than buffer > overruns and memory access.
The original paragraph did not claim that bytecode verification would prevent all security problems. Rather, it specifically claimed that verification would prevent "buffer overruns, segfaults, overruns of other apps' data". (See above).
> Or, tell me how Singularity helps me to protect cases where I want to > right-click on an email address in the body of an email, but garantee(!) > that there is no hole that would allow the sender to anyhow access my > address book.
The original paragraph did not claim that Singularity would prevent you from right-clicking something, or would prevent an app from sending out your address book. It claimed that bytecode verification could prevent buffer overruns, segfaults, overwriting of other apps' data, etc...
> Right. They are orthogonal. That's what he tried to say.
No, that's not what he tried to say. The parent post said: "They [Singularity] aren't the first... Check out EROS [eros-os.org] for an implementation that exists now" which clearly (and wrongly) implied that EROS was an implementation of the same idea, which came before. And the subsequent post claimed "Singularity is a Microkernel, EROS is a Microkernel" which is clearly intended to point out the supposed similarity.
This is simply impossible. The Halting Problem demonstrates that it's impossible to construct a general algorithm that can determine whether some other general algorithm will halt...It's impossible to prove that applications written in Sing#/Spec#C# are correct simply because the compiler cannot divine the intentions of the programmer...
When I said "provably correct or bounded at runtime" I didn't mean that the verifier could prove correctness according to human requirements. I meant that the verifier could prove a more limited kind of correctness: that buffer overruns, illegal casts, improper dereferences, and so on, cannot occur in a given program, regardless of its state. Obviously, it would be preferable to prove correctness in the broader sense, but that's impossible for another program to do. Still, that doesn't mean we should abandon the attempt to prove certain kinds of correctness automatically.
All IPC is necessarily bounded at run-time (limited by machine addressable limits).
Yes, but that kind of run-time bound is useless. We need to prevent a program from overwriting something in actual memory, like another program, the operating system, another region of memory, etc.
If verification exists, and it makes your programs more correct simply by turning it on, why wouldn't people use it?
Some people wouldn't use verification because they're malicious and write programs which are purposefully incorrect according to the verifier. For example, rootkits, trojans, worms, viruses, etc. Nobody writes those programs in Java. Instead, malicious programs are written in C (or other unsafe languages), but your computer, with its OS, will still gleefully execute those malicious programs written in the unsafe language. Wouldn't it be better if the operating system stopped the execution of a malicious program, regardless of where it came from or what language it was written in?
Security constraints must be enforced by the operating system, not by languages or compilers! Otherwise, those constraints could be trivially bypassed by writing the malicious code in some other language (even assembly!) which your OS will happily execute anyway! Take Unix permissions as an example. Suppose they were enforced only by languages. Suppose they were enforced in the C standard library, for example, or suppose they were enforced at compile time using correctness checking. That wouldn't prevent a malicious cracker from modifying the output of that compiler (assembly language) to bypass those constraints.
There is no reason you cannot build this functionality on EROS...We have this functionality running on Linux right now with Java, mono, etc...why should the OS care how an IPC data string is structured?.. These are all guarantees you can have in any other OS... Other microkernels merely use the MMU to enforce isolation instead.
The idea is that all applications running on the operating system (including their IPC) are either provably correct or bounded at run time.
It's important that these restraints be enforced by the operating system. By enforcing these restraints in the operating system, you guarantee that things like buffer overruns and segfauls cannot occur. That's why IPC must be verified by the OS: it guarantees that an application couldn't possibly corrupt some other application. If these things were done using a VM in userspace (Java?) then the security and reliability holes would remain--as long as code verification is optional (by writing in Java rather than C), the malicious coder will opt not to use it.
Also, the isolation enforced by code verification is much more stringent than the isolation enforced by an MMU. Virtual memory enforces isolation per process, whereas verification enforces isolation per buffer which is much more granular. Thus, not only are inter-process violations prevented, but buffer overruns are prevented as well. In addition, enforcing isolation through code verification does not impose the same run-time performance penalties as MMUs, which thus far have limited the adoption of microkernels.
Singularity is built on a microkernel. EROS is built on a microkernel.
You completely misunderstood the point of Singularity. The point of Singularity is that all code (except OS code) is subject to verification, and any code that isn't verifiable is runtime bounds-checked. Furthermore, in Singularity, inter-process communication is structured, such that the OS can verify IPC traffic. Furthermore, the languages for Singularity are strongly typed at the object code level, and garbage collection is performed by the OS--explicit deallocation is impossible for any application. These facilities make it impossible for any application to have buffer overruns, segfaults, or overruns of other apps' data--as a result, all applications can run in ring 0 and virtual memory is not required.
All that has nothing whatsoever to do with Eros. The two projects are not even similar.
Of course a verifier could be written as an application for Eros (or for DOS, for that matter). That statement is like saying that C++ is no different from assembly, because they're both built atop similar hardware and can be used to implement the same things.
Not that you know what developing on an EROS-like system is like, considering it's a completely revolutionary architecture comparable only to KeyKOS from which it's derived.
The Coyotos OS is based on Eros and is quite similar. Additionally, Eros is not completely revolutionary. From the eros web page, What's new about Eros?: "Each of these faclities is...essential to providing scalable reliability, and all of them have appeared in prior systems. No prior system, however, has... this particular combination... in quite the same way.".
Not that you know what developing on an EROS-like system is like
Suse is unimportant. It's yet another linux distro, one among many, and it isn't even that different. If Suse disappeared, its users could just switch to another distro.
Mono, however, is vastly cooler and far more important than Suse, for two reasons. First, there is no open source alternative to Mono--they're actually writing something new instead of just putting together yet another distro. Second, consider Mono's impact--Mono is an implementation of the.NET CLR and C# compilers that will allow future applications written for the Microsoft platform to run seamlessly on Linux (!!). It would be difficult to overstate how important that is. Mono is a major contribution.
Novell funds both Suse and Mono. But only Mono matters. If Novell has problems (and I'm not sure they even are having problems), it shouldn't be Suse that concerns us.
An easy way to discourage driving is to end the enormous government subsidies on oil. Right now, the federal and state governments spend huge sums for things like roads, freeways, stoplights, military interventions in the middle east and elsewhere to protect the oil supply, and other oil-related expenses. Gas taxes are nowhere near high enough to cover the real cost of oil, so in essence, the government is taxing the citizenry to subsidize oil.
Fixing the problem is conceptually simple: internalize the true cost of oil in the price of gasoline. Simply lower income taxes, and raise gas taxes, until gas taxes are high enough to pay for all the costs of oil, including military interventions to assure its supply. If that were done, taxes would be much lower, and gasoline would probably cost more than $14/gallon this year. Alternatives would quickly be found.
Quite a long tirade. I would suggest you skip the presumption that I am uneducated. You're quite quick to jump on that claim, with absolutely no evidence. I made no claim about your level of education, and my "excuse" was made for your benefit. Anyone can claim to be literate, but your lack of linguistic sophistication betrays itself. If you really think there is an English language outside of its usage, you are deluded. There are no rules of English outside of how words are used by English speakers. Everything is fair game if it is understood. Even "manifest" -- the employees in your store have coined a new meaning for that symbol, and it is now a part of the language game played in your store. If you don't like it, tough. That's the way languages work.
I never suggested that there was an English language outside of its usage, or that there were rules of English outside of those used by English speakers. In fact, I repeatedly said just the opposite. What I did suggest, was that a person cannot make up his own rules just because he has his "own linguistic history."
I assume that when you speak of "language games" you're referring to Wittgenstein's later work. In Wittgenstein's later work, he claims that private language is impossible, and that speakers who make up their own usages will be unintelligible even to themselves. Furthermore, he claims that something isn't a rule unless someone can be wrong about it and be corrected. Those are much stronger claims than I was making. I simply claimed that people who make up their own rules aren't speaking English.
It's clearly not the case that everything is fair game if it's understood. That's not how language works. The acquisition of language takes the form of training, which is a point raised repeatedly by Wittgenstein. If a child asks his parent "what does the word 'atheist' mean," the parent doesn't reply "let's make up a definition between ourselves."
My friends who used the word "manifest" incorrectly would often use the term in that way with customers. "Should we manifest it to you?" The customer would respond with a confused, blank stare. I'd have to interject: "He means we can ship it to your house." Only then could the transaction proceed.
...Although I found your remarks ill-considered, I wouldn't have been indignant if your post had been more polite and appropriate.
Here is the entire definition of "agnostic" taken from m-w.com:
a person who holds the view that any ultimate reality (as God) is unknown and prob. unknowable; broadly : one who is not committed to believing in either the existence or the nonexistence of God or a god
The second definition which you cited from answers.com doesn't fit either:
#2: One who is doubtful or noncommittal about something.
Sorry, but the application is neither doubtful nor noncommittal, since computer applications cannot harbor doubts or commitments.
Note that the second definition which you cited was recently added to certain dictionaries only because of persistent misuse of the term. I realize that words can be redefined because of persistent misuse, and that the new definition becomes as legitimate as the old. But that often causes an enormous loss of precision and exactitude in the language. If "agnostic" can mean any kind of uncertainty, do you still know what it means when I say "I'm an agnostic?" Perhaps I'm having trouble deciding where to go for lunch. Bear in mind, there's already a word in the dictionary for uncertainty (namely "uncertainty"), and another one for doubt ("doubt"). Steve Jobs could have used either term, or he could have said "This program is database independent." But he was trying to be clever (as you are, I'd imagine).
> There is no metaphor
You must be kidding. It's obviously a methaphor.
>I do find it humourous how you just assumed that your misunderstanding was due to Steve Job's lack of education when it was obviously your own lack of modern literacy that caused it.
I find it ridiculous that you would so completely misunderstand the issues of grammar and word choice, especially on a thread discussing precisely those issues. And I find it pathetic that you would invoke the perennial excuse of people who don't know the rules of English: "I have my own history, and my own rules." I (unlike you) am literate, which is why I stick to the "narrow dogma" that words have specific meanings.
We may "each have our own linguistic history" but that obviously doesn't imply that everything in that history is correct. Suppose I have a history of spelling "lose" as "loose?" Or "nuclear" as "nukyoolar?" According to your view, the second spelling is just as correct, right?
The difficulty is not my mistunderstanding, but Steve Job's lack of education (and yours).
> There will certainly be uses we've never heard. But assuming that someone is uneducated because they didn't fit into our limited mold is foolish.
True, it doesn't fit into my "limited mold." Neither does spelling "lose" as "loose," or "definitely" as "defunately." My mold is limited, because language must be conducted according to shared rules otherwise it will become impossible for us to understand one another. If language is conducted solely according to one's own rules, then it's not English but gibberish.
The difficulty here is one of excuses. People who have never learned to speak beyond the 4th grade level often invoke your absurd excuse: "I have my own history, and my own rules." When students invoked that excuse in one of my math classes, the teacher had an excellent response: "You can do math according to your own rules, when you're at home by yourself." Of course, English isn't math, because the rules of English aren't independent of everybody. But the rules of English are independent of you.
Indeed, the problem is worse than I've implied. Everyone knows there are rules to English. Everyone knows there's a definition of "agnostic" independent of "his own history." People who misuse words like "agnostic" are people who wish they knew the rules of English, so they try throwing around esoteric words to impress people. They could just have used "unsure," but they want to show people that they have a mastery of mo
If someone's written work is devoid of some common rules of grammar and usage, does it matter if you completely and unambiguously understand what they are saying/writing?.. I try to use the rules, but if I understand you, what else matters?
You make a strong point. It's for that reason that I tend not to be too pedantic when criticizing others' grammar.
But the point you're making has two flaws, as I see it. 1) People who spell 'lose' as 'loose' come across as never having read a book in their entire lives. This makes a bad impression even if the meaning of the sentence is successfully conveyed. 2) Some violations of grammar and word choice are so egregious, that I have difficulty parsing the sentences in which I find them. In some cases I can guess the meaning, as with sentences that don't respect basic parts of speech ("Fill smart! Think different! I'm doing good!"). However it takes additional effort on my part--I must rewrite the sentence mentally into something intelligible.
There are other cases, however, where I have no fucking idea what the speaker is trying to convey. For example, I recently worked at a bookstore, and when the employees there saw a shipping manifest, they inferred that "to manifest" meant "to ship" or "to move." As a result they would often say "Should we manifest this to another store" or "Please manifest this over there." Sometimes they would say such things to customers. The customers would just stare in confusion.
I was genuinely confused when Steve Jobs, a man I'd assume was educated, first said that one of his products was "Operating System agnostic." Since then, that weird usage has taken off. For example, applications can now be "database agnostic." Of course, applications can't really be agnostic. To a person who actually speaks the English language, those phrases are meaningless. Even if we were to use 'agnostic' as a metaphor, it still wouldn't work. Is the application in doubt about the existece of God? Is the application in doubt about the existence of the database? Can the application even harbor doubts? What happens if the application becomes an atheist? Perhaps it will be a "database atheist," meaning the app is not database-driven? In fact, my application is a "database pantheist." See if you can figure out what the fuck that means.
In which universe? It sure as hell isn't this one. Java 1.5 start up time is appalling.
I just compiled a small Java command line application and executed it, using Java 1.5 on my 2.8 ghz p4, in this universe. Startup time for the Java app was 51 milliseconds, which is virtually imperceptible to the user.
It also quickly consumes huge amounts of RAM and sends the machine into swap even with small Swing apps.
As a test, I just ran a minor swing app, and the RAM consumed was 9M (excepting share). This would not "send the machine into swap" unless I were using a 386 sx/16 with 8MB of RAM from 15 years ago, or unless I were running so many other things that I had virtually no RAM remaining, which would be my fault and not Java's.
You call them "misconceptions"... they aren't. I'd encourage anyone to actually *try* modern Java/Swing apps...
You may consider taking your own advice. Several of the things you claimed were true can be proved false. Either you didn't try the things you claimed, or you're making stuff up, or you did something wrong.
...I routinely use the IntelliJ IDE/Case Tool, which is an enormous, heavyweight application written using Swing. Although the memory usage is worse than a comparable C++ app, I don't detect any difference in GUI responsiveness. In fact, GUI responsiveness is so good that Swing apps on Windows are perceptibly more responsive than natively compiled C++ apps on XF86, since X still imposes noticeable performance penalties.
Granted, memory usage is worse with Java. This is for two reasons: 1) a generational compacting collector takes about twice the heap space, and 2) libraries are not shared between different Java apps. The second problem is going away with Java 1.6. The first problem can be gotten rid of by changing garbage collection parameters, but it's a size vs. speed tradeoff.
nuclear decay is a completely spontaneous process. the only way to get more beta particles is to have more radioactive material. long lasting does not mean lots of power.
They weren't claiming you could recharge the nuclear battery. They were claiming that the battery would still emit sufficient power to run a laptop, even after 10 years. That's obviously possible; radioactive half-lives are often measured in decades; it's just a matter of how much nuclear material the battery would have to contain.
this reminds me of an essay I read by a second year physics student that nanotechnology would allow us to run 10GHz computers for 10 years off a watch battery. it's BS but you don't need to look at the technology to see that, it's just basic thermodynamics
Neither the nuclear battery, nor the watch battery, violate the basic laws of thermodynamics.
First of all, your criticism is pedantic and irrelevant to the topic being discussed.
Second, your criticism is incorrect. The subjunctive mood was not used in my sentence, nor was it called for. The subjunctive mood is used when the state of affairs being described is contingent or hypothetical. Since my sentence described an actual fact, the subjunctive mood wasn't called for, and wasn't used.
As an aside, your posts are foolish. Your intial post was incorrect and ill-informed. When this was pointed out, you responded with a pedantic, irrelevant, and incorrect criticism of grammar. I realize this is slashdot and most people don't put great effort into their posts. But you could try reading the linked article before commenting upon it, or looking up the word "subjunctive" before incorrectly critizing someone for violating the grammatical construct to which it refers.
This paper would fail the first rule, if you don't understand it, reject it. Either it is drivel, or it is submitted to the wrong conference/journal/whatever or you should not be a referee for this.
The article clearly states that the random paper was submitted to a fake journal/conference that doesn't even bother to read submissions.
Of course, this kind of scam works on the reluctance of accademics to just say they don't understand something.
In my job I constantly interact with professors of computer science, and I find no reluctance among them to admit they don't understand something.
The randomly generated paper would never have been accepted to any serious journal of computer science. A mediocre grad student would realize the paper was meaningless upon reading the first four sentences.
The program for generating random papers may be funny, but it doesn't produce plausible CS papers. I read the two papers linked from the site and you can tell they're obviously fake from reading the abstracts only.
Slashdot Mirror
Exports exist to pay for imports. Foreigners export products to the U.S. because they wish to buy capital-intensive American products. Foreigners do not export products because they wish to give us things for free.
Even when foreigners buy debt (instead of our products), that means only that our exports of products to them are delayed. And when foreigners buy American companies, that means they're investing in our economy rather than their own.
As many European and Japanese companies have been bought by American firms. For example, Volvo, Saab, Mazda, Jaguar, Lambourghini, Maserati, and other car manufacturers are now owned by U.S. companies--and that's limiting our attention just to car companies.Just because a company is owned by a Swiss firm doesn't mean its profits go to Switzerland.
So what if Europeans are investing here. Phrased another way, there has been a massive capital flight from Europe to America. Investment is leaving Europe at a rapid pace and setting up shop in the U.S!
Is it really such a problem if foreigners would rather invest in our economy than in theirs?
After all, the $150 Japanese DVD players are now competing against $35 Chinese players. And the cost of clothing in real terms has declined by 40% over the last decade.
Unfortunately, if you had such poll asking if people had their prices reduced, then everyone would have to answer in the affirmative if they were being honest. (And don't say "prices go up because of inflation" since we're talking about prices in real terms, and your salary goes up with inflation).
At this rate the "loop" you describe requires about 100 years between iterations, and everyone's wages are raised by going through the iteration.
And don't just claim that all examples so far haven't been "true" socialism. If no true socialism has ever prevailed, then you must wonder why every single socialist experiment ends up degenerating into something else.
Capitalism drastically raised the standard of living for everyone. Unions had little or nothing to do with it. Union participation has been decreasing since the 1930s and the standard of living for the middle class has increased dramatically since then. Unions and strikes caused little or no increase in the standard of living for the middle classes in America. The vast majority of the increases in wages have happened to professions which are not unionised and which are not affected by wage legislation. Our country was rich before it traded with the 3rd world in any substantial form. The vast majority of the trade with the 3rd world we currently conduct is a result of reductions in trade barriers since the 1980s. Trade with the 3rd world was very slight during the first half of the 20th century--about all we imported was fruit from central america. 3rd world countries are poor because they're capital-depleted and their labor productivity is very low. They have extremely little wealth to siphon off. They were poor countries before the U.S. showed up....The phenomenon of 3rd world nations trading with wealthier nations, has always and invariably increased the real wages of 3rd world workers. As an example, the real median wage in China has more than quadrupled in the last 25 years since they abandoned the foolish prescriptions which you espouse.
If China were to "flood the US markets with computers and cars" without buying anything from the US in return, then the Chinese have given the US free computers and cars without getting anything in return.
You wish to shoot Ricardo because he wrote something you disagree with? You certainly live up to your handle, Mr Marxist! Perhaps you should read what Ricardo wrote before shooting him.(I'm continually surprised by which posts get modded up...)
Electricity in this country is generated primarily from coal (60%), nuclear (20%), and hydroelectric (>10%). The U.S. has enough coal to last for centuries (environmental damage notwithstanding), and enough Uranium to last for millenia (if breeder reactors are used).
Even if both of those resources were completely exhausted, or were nearing exhaustion, there are many alternatives which are only modestly more expensive. Wind farms, geothermal plants, solar panels, tidal energy plants, methane hydrates, and others.
Even if all sources of power were to become scare (an impossibility), the first thing sacrificed would be air conditioning for the home, which consumes more electricity by far than computers and electronic equipment.
...But let's suppose that all coal and uranium were exhausted, and all alternatives became scarce, and home air conditioning had already been abandoned, and still there was not enough electricity to power computers. Then we could build desktop and server computers with power-saving technology similar to that used in notebooks today. At 45nm fabbing, we could make processors that are faster than today's but which consume less than 1W. 1W could be easily generated by power plants which harvest methane from pig farts, just like in "Mad Max: Beyond Thunderdome." If we used the Thunderdome approach, the only remaining possible source of scarcity would be: if pigs learned to fly.
As a result, it seems unlikely that power will become scarce enough in the near future to render cabling unimportant.
Granted, using U-238 as fuel would require breeder reactors, which are virtually non-existent right now. But breeder reactors are feasible and proven; France ran a commercial 1200MW breeder reactor for 15+ years. Although France's breeder reactor had various technological glitches, it did work, and it was the very first commercial breeder reactor. The first iteration of anything has technological glitches.
Also bear in mind, that the estimates of Uranium availability only take into account the conventional, terrestrial reserves. Those conventional reserves constitute only a tiny part of global Uranium--for example, the amount of Uranium dissolved in ocean water is many times higher than the conventional reserves. Although Uranium dissolved in ocean water is too diffuse to be profitably extracted right now, we could probably figure out better ways of extracting it in the next few thousand years. Who knows, by then, 5000 years from now, we may even be able to mine Uranium from the asteroid belt, which is right next door, cosmically speaking.
With breeder reactors, we may not ever run out of Uranium. Even if we did run out, we'd have a very long time beforehand to consider next steps.
The verdict of poor language design was not Sun's (nor did my post imply that) but rather a consensus of language designers. That consensus is so widely shared that Stroustrup has to issue apologies on occasion...
You have to be joking. Those are desirable? Also, you should learn how to spell better than a 3rd grader... Java has parametric polymorphism. That topic has been discussed already.... C++ has native types and classes.But I assume you're complaining about the split between primitive types and their object wrappers in Java (int vs Integer, etc). That was the most elegant way of solving various problems without having an "everything is an object" language.
You mean it lacks the hideously overloaded "const" keyword of C++? Those are EXTREMELY minor concerns. You honestly have difficulty typing the 'new' keyword? And your complaint is that it could theoretically have been omitted from the language? C++, C, and most languages lack a complete set of numeric types primitive to the langauge. C++ has recently added additional numeric types like complex (they weren't there for the first 15+ years), but they're in class libraries, and similar libraries are available for Java. This is an obscure concern that almost never comes up in practice. I must say, there are a great many legitimate criticisms of Java. But you haven't hit on even one of them. Your list was a compilation of incorrect and trivial criticisms. I have no doubt that you could "go on" and demonstrate your ignorance still further.Bear in mind that most of the "features" not included from C++ were examples of very poor language design.
There are some omissions from Java which are unfortunate, like pre-conditions and post-conditions, however most of the omissions were not present in Java's predecessors.
The only features of C++ which are lacking in Java and which arguably should be included are: operator overloading, and some variant of the const keyword (although not exactly like C++ const, which is hideously overloaded).
No serious programmer ever brags about million-line programs, or claims that lines of code somehow correlates positively with maintainability. You're attributing a point of view to "people" which they don't possess. Perhaps some friend of yours made a comment like that, which you now attribute to "people"? Typecasting containers is not done in Java, since it has generic containers. Re-implementing missing containers is very rare since the Java API has a wide variety of containers already (some even argue the API is too big). Recoding functions for different datatypes is unnecessary since Java has always been polymorphic. And there's nothing wrong with switch statements; they're syntactic sugar for if/else (I assume you're not opposed to if/else).I'll grant that the repetitive get/set wrappers are unfortunate and unnecessary. I don't understand why the designers of Java don't just borrow ideas from Eiffel on this matter. I realize the Java designers are opposed to syntactic sugar in general, but in this case it would be justified, since a huge proportion of Java code wasted is on statements like "public Foo getFoo() { return this.foo; }".
So perhaps the programmers writing daemons should use only Java, and perhaps users should run only Java services. That is half way to the solution.
What about buggy or malicious drivers, protocl stacks, or loadable kernel modules? Users install those programs from CDs or from net downloads, and those programs run in ring 0 which means they can overwrite any memory on the system--they could even corrupt the JVM which previously had been secure. Is it really safe to have the operating system accept modifications to itself, without being able to verify that those modifications won't corrupt the rest of the operating system? Indeed, Microsoft claims that >85% of system crashes are caused by bad drivers written by 3rd party companies...
No. To prevent intrusions, all programs running on the system must not have these problems, which means it must be enforced by the operating system.
> > These facilities make it impossible for any application to have
> > buffer overruns, segfaults, or overruns of other apps' data
> > -- as a result, all applications can run in ring 0 and virtual
> > memory is not required.
>
> Which is complete nonsense, because there are many other classes of
> security problems, some of them application-specific, other than buffer
> overruns and memory access.
The original paragraph did not claim that bytecode verification would prevent all security problems. Rather, it specifically claimed that verification would prevent "buffer overruns, segfaults, overruns of other apps' data". (See above).
> Or, tell me how Singularity helps me to protect cases where I want to
> right-click on an email address in the body of an email, but garantee(!)
> that there is no hole that would allow the sender to anyhow access my
> address book.
The original paragraph did not claim that Singularity would prevent you from right-clicking something, or would prevent an app from sending out your address book. It claimed that bytecode verification could prevent buffer overruns, segfaults, overwriting of other apps' data, etc...
> Right. They are orthogonal. That's what he tried to say.
No, that's not what he tried to say. The parent post said: "They [Singularity] aren't the first... Check out EROS [eros-os.org] for an implementation that exists now" which clearly (and wrongly) implied that EROS was an implementation of the same idea, which came before. And the subsequent post claimed "Singularity is a Microkernel, EROS is a Microkernel" which is clearly intended to point out the supposed similarity.
Security constraints must be enforced by the operating system, not by languages or compilers! Otherwise, those constraints could be trivially bypassed by writing the malicious code in some other language (even assembly!) which your OS will happily execute anyway! Take Unix permissions as an example. Suppose they were enforced only by languages. Suppose they were enforced in the C standard library, for example, or suppose they were enforced at compile time using correctness checking. That wouldn't prevent a malicious cracker from modifying the output of that compiler (assembly language) to bypass those constraints.
It's important that these restraints be enforced by the operating system. By enforcing these restraints in the operating system, you guarantee that things like buffer overruns and segfauls cannot occur. That's why IPC must be verified by the OS: it guarantees that an application couldn't possibly corrupt some other application. If these things were done using a VM in userspace (Java?) then the security and reliability holes would remain--as long as code verification is optional (by writing in Java rather than C), the malicious coder will opt not to use it.
Also, the isolation enforced by code verification is much more stringent than the isolation enforced by an MMU. Virtual memory enforces isolation per process, whereas verification enforces isolation per buffer which is much more granular. Thus, not only are inter-process violations prevented, but buffer overruns are prevented as well. In addition, enforcing isolation through code verification does not impose the same run-time performance penalties as MMUs, which thus far have limited the adoption of microkernels.
All that has nothing whatsoever to do with Eros. The two projects are not even similar.
Of course a verifier could be written as an application for Eros (or for DOS, for that matter). That statement is like saying that C++ is no different from assembly, because they're both built atop similar hardware and can be used to implement the same things.
The Coyotos OS is based on Eros and is quite similar. Additionally, Eros is not completely revolutionary. From the eros web page, What's new about Eros?: "Each of these faclities is...essential to providing scalable reliability, and all of them have appeared in prior systems. No prior system, however, hasSuse is unimportant. It's yet another linux distro, one among many, and it isn't even that different. If Suse disappeared, its users could just switch to another distro.
Mono, however, is vastly cooler and far more important than Suse, for two reasons. First, there is no open source alternative to Mono--they're actually writing something new instead of just putting together yet another distro. Second, consider Mono's impact--Mono is an implementation of the .NET CLR and C# compilers that will allow future applications written for the Microsoft platform to run seamlessly on Linux (!!). It would be difficult to overstate how important that is. Mono is a major contribution.
Novell funds both Suse and Mono. But only Mono matters. If Novell has problems (and I'm not sure they even are having problems), it shouldn't be Suse that concerns us.
An easy way to discourage driving is to end the enormous government subsidies on oil. Right now, the federal and state governments spend huge sums for things like roads, freeways, stoplights, military interventions in the middle east and elsewhere to protect the oil supply, and other oil-related expenses. Gas taxes are nowhere near high enough to cover the real cost of oil, so in essence, the government is taxing the citizenry to subsidize oil.
Fixing the problem is conceptually simple: internalize the true cost of oil in the price of gasoline. Simply lower income taxes, and raise gas taxes, until gas taxes are high enough to pay for all the costs of oil, including military interventions to assure its supply. If that were done, taxes would be much lower, and gasoline would probably cost more than $14/gallon this year. Alternatives would quickly be found.
I assume that when you speak of "language games" you're referring to Wittgenstein's later work. In Wittgenstein's later work, he claims that private language is impossible, and that speakers who make up their own usages will be unintelligible even to themselves. Furthermore, he claims that something isn't a rule unless someone can be wrong about it and be corrected. Those are much stronger claims than I was making. I simply claimed that people who make up their own rules aren't speaking English.
It's clearly not the case that everything is fair game if it's understood. That's not how language works. The acquisition of language takes the form of training, which is a point raised repeatedly by Wittgenstein. If a child asks his parent "what does the word 'atheist' mean," the parent doesn't reply "let's make up a definition between ourselves."
My friends who used the word "manifest" incorrectly would often use the term in that way with customers. "Should we manifest it to you?" The customer would respond with a confused, blank stare. I'd have to interject: "He means we can ship it to your house." Only then could the transaction proceed.
...Although I found your remarks ill-considered, I wouldn't have been indignant if your post had been more polite and appropriate.
The second definition which you cited from answers.com doesn't fit either:
Sorry, but the application is neither doubtful nor noncommittal, since computer applications cannot harbor doubts or commitments.
Note that the second definition which you cited was recently added to certain dictionaries only because of persistent misuse of the term. I realize that words can be redefined because of persistent misuse, and that the new definition becomes as legitimate as the old. But that often causes an enormous loss of precision and exactitude in the language. If "agnostic" can mean any kind of uncertainty, do you still know what it means when I say "I'm an agnostic?" Perhaps I'm having trouble deciding where to go for lunch. Bear in mind, there's already a word in the dictionary for uncertainty (namely "uncertainty"), and another one for doubt ("doubt"). Steve Jobs could have used either term, or he could have said "This program is database independent." But he was trying to be clever (as you are, I'd imagine).
> There is no metaphor
You must be kidding. It's obviously a methaphor.
>I do find it humourous how you just assumed that your misunderstanding was due to Steve Job's lack of education when it was obviously your own lack of modern literacy that caused it.
I find it ridiculous that you would so completely misunderstand the issues of grammar and word choice, especially on a thread discussing precisely those issues. And I find it pathetic that you would invoke the perennial excuse of people who don't know the rules of English: "I have my own history, and my own rules." I (unlike you) am literate, which is why I stick to the "narrow dogma" that words have specific meanings.
We may "each have our own linguistic history" but that obviously doesn't imply that everything in that history is correct. Suppose I have a history of spelling "lose" as "loose?" Or "nuclear" as "nukyoolar?" According to your view, the second spelling is just as correct, right?
The difficulty is not my mistunderstanding, but Steve Job's lack of education (and yours).
> There will certainly be uses we've never heard. But assuming that someone is uneducated because they didn't fit into our limited mold is foolish.
True, it doesn't fit into my "limited mold." Neither does spelling "lose" as "loose," or "definitely" as "defunately." My mold is limited, because language must be conducted according to shared rules otherwise it will become impossible for us to understand one another. If language is conducted solely according to one's own rules, then it's not English but gibberish.
The difficulty here is one of excuses. People who have never learned to speak beyond the 4th grade level often invoke your absurd excuse: "I have my own history, and my own rules." When students invoked that excuse in one of my math classes, the teacher had an excellent response: "You can do math according to your own rules, when you're at home by yourself." Of course, English isn't math, because the rules of English aren't independent of everybody. But the rules of English are independent of you.
Indeed, the problem is worse than I've implied. Everyone knows there are rules to English. Everyone knows there's a definition of "agnostic" independent of "his own history." People who misuse words like "agnostic" are people who wish they knew the rules of English, so they try throwing around esoteric words to impress people. They could just have used "unsure," but they want to show people that they have a mastery of mo
But the point you're making has two flaws, as I see it. 1) People who spell 'lose' as 'loose' come across as never having read a book in their entire lives. This makes a bad impression even if the meaning of the sentence is successfully conveyed. 2) Some violations of grammar and word choice are so egregious, that I have difficulty parsing the sentences in which I find them. In some cases I can guess the meaning, as with sentences that don't respect basic parts of speech ("Fill smart! Think different! I'm doing good!"). However it takes additional effort on my part--I must rewrite the sentence mentally into something intelligible.
There are other cases, however, where I have no fucking idea what the speaker is trying to convey. For example, I recently worked at a bookstore, and when the employees there saw a shipping manifest, they inferred that "to manifest" meant "to ship" or "to move." As a result they would often say "Should we manifest this to another store" or "Please manifest this over there." Sometimes they would say such things to customers. The customers would just stare in confusion.
I was genuinely confused when Steve Jobs, a man I'd assume was educated, first said that one of his products was "Operating System agnostic." Since then, that weird usage has taken off. For example, applications can now be "database agnostic." Of course, applications can't really be agnostic. To a person who actually speaks the English language, those phrases are meaningless. Even if we were to use 'agnostic' as a metaphor, it still wouldn't work. Is the application in doubt about the existece of God? Is the application in doubt about the existence of the database? Can the application even harbor doubts? What happens if the application becomes an atheist? Perhaps it will be a "database atheist," meaning the app is not database-driven? In fact, my application is a "database pantheist." See if you can figure out what the fuck that means.
I just compiled a small Java command line application and executed it, using Java 1.5 on my 2.8 ghz p4, in this universe. Startup time for the Java app was 51 milliseconds, which is virtually imperceptible to the user.
As a test, I just ran a minor swing app, and the RAM consumed was 9M (excepting share). This would not "send the machine into swap" unless I were using a 386 sx/16 with 8MB of RAM from 15 years ago, or unless I were running so many other things that I had virtually no RAM remaining, which would be my fault and not Java's.
You may consider taking your own advice. Several of the things you claimed were true can be proved false. Either you didn't try the things you claimed, or you're making stuff up, or you did something wrong.
...I routinely use the IntelliJ IDE/Case Tool, which is an enormous, heavyweight application written using Swing. Although the memory usage is worse than a comparable C++ app, I don't detect any difference in GUI responsiveness. In fact, GUI responsiveness is so good that Swing apps on Windows are perceptibly more responsive than natively compiled C++ apps on XF86, since X still imposes noticeable performance penalties.
Granted, memory usage is worse with Java. This is for two reasons: 1) a generational compacting collector takes about twice the heap space, and 2) libraries are not shared between different Java apps. The second problem is going away with Java 1.6. The first problem can be gotten rid of by changing garbage collection parameters, but it's a size vs. speed tradeoff.
They weren't claiming you could recharge the nuclear battery. They were claiming that the battery would still emit sufficient power to run a laptop, even after 10 years. That's obviously possible; radioactive half-lives are often measured in decades; it's just a matter of how much nuclear material the battery would have to contain.
Neither the nuclear battery, nor the watch battery, violate the basic laws of thermodynamics.
First of all, your criticism is pedantic and irrelevant to the topic being discussed.
Second, your criticism is incorrect. The subjunctive mood was not used in my sentence, nor was it called for. The subjunctive mood is used when the state of affairs being described is contingent or hypothetical. Since my sentence described an actual fact, the subjunctive mood wasn't called for, and wasn't used.
As an aside, your posts are foolish. Your intial post was incorrect and ill-informed. When this was pointed out, you responded with a pedantic, irrelevant, and incorrect criticism of grammar. I realize this is slashdot and most people don't put great effort into their posts. But you could try reading the linked article before commenting upon it, or looking up the word "subjunctive" before incorrectly critizing someone for violating the grammatical construct to which it refers.
The randomly generated paper would never have been accepted to any serious journal of computer science. A mediocre grad student would realize the paper was meaningless upon reading the first four sentences.
The program for generating random papers may be funny, but it doesn't produce plausible CS papers. I read the two papers linked from the site and you can tell they're obviously fake from reading the abstracts only.