DLP's generally have a sucky side-viewing angle. If a color-wheel is in use, it can sound like a jet-engine. Moving parts can be volitile. Many people can see a "rainbow effect" due to the strobed color consolidation of a color-wheel.
That being said, DLP's are the only "high end" option for people on a budget. Unless you're willing to go w/ Westinghouse/scepter/maxtent LCD's. E.g. the zenith of our day.
And the term organic has a chemical definition. So by merely putting a chain or two of the appropriate molecule into silicon, then for marketing purposes, I could label my computer:
A clever going-places mid-level management kind of guy.
It would be technically correct, but wholely loathsome and misleading.
You don't write 'I'll be back in two hours' memos in ancient egypt and stick them to the fridge, if the intended recipient never had shown any interest in archeology.
Except that the woman is not trying to convey information via her body language. Women recognize that both Men & Women can recognize "Anger", "happiness", etc. Thus when a woman recognizes subtle discomfort, interest, lack-thereof her experience gives her a subconcious reinforcement that the man recognizes this as well.
Women interact in their own circles and have the benifit of common biological cycles and common experiences. So when they are amongst themselves, there is a symbiosis; buliding apon each other through communication and support networks (and flat out a common agenda). When interacting w/ men they don't achieve this same experience.. They are incredibly frustrated because we do not react the way they intuitively expect (and by intuitivly, I mean empirically; there is no mysticism going on here). It's as if you were riding a bike and expect minute weight shifts to cause the bike to operate in a certain way, but when you ride someone else's bike, the body motion had completely unexpected results... Because this is internally learned communication, you are completely disoriented because you can't retrain your body properly nor do you fully underestand WHAT isn't working.
Notice the concept of a tom-boy.. Let's suppose for simplicities sake that male-ness and female-ness (e.g. the propensity for transgender attributes) does not apply. That we are (as the above has suggested) highly affected by our empirical understanding of the world. A girl that does not fraternize in homogeneous female circles will not have such a dramatic communication problem with men. Their expectations are consistently met, and there is little disconnect w/ men. The problem, however, is that they are not disconnected with female circles. Generally rejected as they no longer know how to communicate properly nor express acceptible behavior.
This doesn't even take into account the petty content of all-male or all-female circles (cars, toys, sports, science, sci-fi for boys, and boys, celebrities, fashion, etc for girls). The lack of common interest only exascerbates the disparity. But is no moreso dramatic than between geeks and jocks.
But the common attraction between men & women at least brings the two separate worlds together unlike geeks and non-geeks. (or even between different sects of geek). Each gender looks down on the other gender because they can recognize obvious defficiencies. "He can't communicate", or "She can't reason". "He is insensitive", or "She is overly emotional".
I've just finished a great book called "Getting to the Yes" which describes how to come to agreements w/ people more effectively. The key element in the book is that you have to recognize that people are emotional creatures. Both men & women have pride, can feel threatened, feel insignificant. There is a definitive pecking order in both male and female circles, and if you feel like you should be equal in a relationship but the pecking order says otherwise, there can be depression (for either party). So when intereacting inder-genderly, you MUST recognize these differences and encorporate them into your communication.. For men interacting with women, there are obvious strategic implications here. For women, they have to recognize that there are ramifications to their tone and actions. Women are good and adapting the focus of a communicay based on changing emotional needs. They need to recognize that this can be a good thing when they "feel" that the conversation is not going to go forward successfully. Their feeling is correct; their ability to natively read the body language gives them the tools to steer them; they need only the resolve to bring about the communicay in a direction devoid of the uncomfortable feeling. If this means changing their position or delaying the conversation (by switchign to a happier
You do know that the entire currency system these days is just based on "good faith"
Bartering has always been a sense of good-faith. People would bite coins to semi-verify their authenticity. They would be suspicious of the quality of the animals that were used to make purchases.
There have been numerous issues with counterfitting gold, and the enormous issue of shipping massive amounts of gold from country to country; piracy, insider theft, organized harbor crime, etc. Thoughout the centuries, the management of gold has been an issue.. And once a cache of gold was found in a country, the country would become artificially wealthy. Look at the rise & fall of spain.
If money grew on trees, then economies would collapse, and the core intent of efficiently allocating scarse resources (such as food, water, land, etc) would no longer have meaniful trade.
When you standardize an open commodity like gold, you have to protect it. When you standardize an artificial and hard to reproduce fiat like coin/paper, countries are suddenly being valued by their assets, not their luck of finding a very specific type of natural resource. Being lighter and more compact, it's easier to protect and transport.
There are other economic issues involved with using an intrinsicly valueable commodity for currency. The "value" of gold fluctuates. This means, no matter what the state labels the value of gold, the consumers (who wear and decorate their homes with it because it's pretty) will have their own aggregate value. When there is a disconnect between the state sanctioned value and the consumer value, arbitrage can occur. This would be something similar to if the marginal cost of producing a penny or dollar were 100 times greater than the value of it. With gold, WHEN this occurs, people take their pay, melt it down, and resell it until the supply-demand equilizes ; lowering the price to match the state's stated rate. Or until the state wises up and raises the value of gold. When gold is too low relative to it's value, people will melt down their jewlry and fashion coins.
When two countries have different values for gold (e.g. I can buy a chicken with fewer gold in one country than another), then people will trade gold until their effective prices are equilized (arbitrage). The negative is that improperly set gold prices will form a massive egress of their stockpiles of gold.
In the old days, the cost of transporting something over-seas was VERY high.. You had piracy, heavy slow ships.. You had to pay the crew well to get honest enough shippers.. All sorts of things.
Now you can put it in an anonymous box for the most part and have a pretty high degree of trust.. Moreover you can ship billions of dollars worth of assets. Arbitrage is almost instantaneous in the modern world.. We no longer require physical presense of our assets.. We purchase "orange juice futures" by the billions while the commodity actually sits in another country... Likewise with gold.
A country in our modern world would very quickly find it's "currency" dramaticlly shifted due to minor fluctuations in the stock-market. And currency is extremely important to the economic welfare of a country.
If we were still on the gold standard, then inflation would NOT be controllable by the fed (who's job it is is to manage our currency). Instead of buying/selling US treasuries, the fed would have to directly buy and sell gold. But in econimicly tight times, they would eventually deplete the gold stock-piles, and surely congress would limit how low the stockpiles could be depleted. Again, remember, they never have to take the money out of fort-knox to sell it.. A serial-numbered pilon is merely designated as owned by someone other than the US government.
The 1999 forward recession has been dramatically avoided because the fed has been pumping virtual dollars into the economy like mad.... Inflating the economy. Whether you believe this is a smart move or not, the fed would NOT have this
The alarmists aren't happy unless they're running around screaming "the sky is falling!".
In positional arguing, one side needs to over-exaggerate their position such that the "compromise" is closer to their intended goal.
Forget global warming, and other death-to-earth claims. (Nature will have the final word here.) What about industrial pollution of drinking water and the rise in asma and lung disease in cities due to direct human-affecting pollution or freon leakage..
There are so many more direct problems associated with cost-effective pollution. Maybe people care about it, maybe not. But by extending the problem to killing off innocent people that don't live/work in urban areas, it becomes a moral crisis.. The only problem is that these people go too far in their plausibility and their arguments are thereby dismissed.
costco online is about to have a $1,700 full resolution 1900x1080p 37" LCD w/ full inputs (DVI, multiple HDMI, component, VGA, composite, s-vid).
We're not talking about supporting current hardware. We're talking about supporting a content format; essentially the software. The cost of the DVD player is NOTHING.. We're talking about replacing thousands of dollars of DVD collections. Are you going to gamble this collection upgrade on one format when another one might win.
I'm assuming we're not going to see dual releases of the same titles either.. It's like XM v.s. sireus (sp?). Or HBO v.s. Showtime.. Different content which you'd have to pay double to the full range.
No, he means the player. HDMI supports a specific sets of resolutions; standard DVD is one of them.
Currently you can get hdmi-capable DVD players. It passes both audio and video to your hdmi TV. I believe there are even HDMI audio recievers; though I don't know if DVD players have dual HDMI outputs thus that seems kind of silly.
Not all new TV's have HDMI inputs either. Many have DVI which can have an HDMI adaptor applied (if it doesn't ship with the TV).
From my most recent research, what I've found is that the optimal cost-effective way to get 37+ inches of flat-panel is to get the clearest-picture TV that you can afford and get a DVD player with high quality conversion to one of the several HDMI signals.. A more expensive DVD player is going to be much cheaper than a high quality conversion device on the TV.
The end result is that HDMI is a preferable solution going forward. Most likely it is going to win out; we're just going to have to accept the associated DRM.
I'll leave for further discussion the techincal merits of displaying full resolution 1080i or 1080p in analog or even through DVI.
Maybe not stupid, but definitely powerless. Boycotts don't exist anymore; they have to be forced on people; e.g. international embargos, regulations, militarily-enforced cartels. Facing a lack of options, most modern people (especially now-minded Americans) will take the best of what is available to them. As a result, anything short of fearing for safety, isn't sufficient to maintain a boycott.
I'm not aware of flash memory handling 150Meg / second random-seek time. And definitely not for writing.
As for 16 hour time, I always hook $80 UPS's to my machines, so if your machine is off but you're hooked up to a UPS, you'll have more than a week of life if the power goes out. Yes if your machine itself is on when the power goes out and you're not there to shut the machine off (thereby letting the UPS feed the memory drive) you'll have a problem, but even then, most power outages are less than a couple hours.
Price wise, if you're talking about a $500 machine, then the price is of course impractical.. But if you have a $5,000 server, a $2,500 work-station or a $3,000 gaming machine (including monitor + speakers + else), then the price isn't bad; just looking at it as one additional performance enhancement.
If you're a dot-com company and only have 10 gig of database data to worry about, you could set up several drives and partition your databases across them getting tremendous performance improvements (continuous random seeks). Not sure about mysql, but postgres does NOT make use of memory directly, it's memory footprint never seems to get above 9Meg per process. You can up various numbers, but they're dangerous if you're not an expert. Simply throwing a fast ram-drive is easy to deal with (with the stated above exception of power-outage). It's better than a system RAM-drive solution because you have more reliable fail-over (no need to copy to-from physical disk space on start-up shutdown/crash).
The consolidation to 939 was probably nothing to do with having a "foreward looking standard" as having fewer current generation parts to test.. Every quarter, there is a new technolog, either driven by the video card industry or the CPU industry. The common denominator is the motherboard.. So in essence, choice of a MB is more critical than the other periferals like CPU / vid-card, because it limits your choices.. Now imagine that you are a MB manufacturer and there are two major CPU makers, and EACH maker requires 6 different configurations. THEN, you have 4 different classes of chipset makers (SiS, Via, AMD/Intel(native), nVidia, and soon ATi). To provide a complete suite of choices for the customer (so nobody can say, hey, ASUS, I like you, but I needed this particular configuration).. If as an MB maker, you picked the wrong horse (ATi, v.s. nVidia for your high end chipset), and your competition chooses the opposite, then there is a generation of customers that have cozied up to another name-brand.
As MB makers produces the dozens of possible combinations, they have to test these guys, and invariably, some have problems.. That's why you have different versions of the same motherboard (though they never advertise which version you're purchasing). You generally never want a rev1, but by the time you get to a stable rev, the next generation is out with dramatic enhancements.
Thus, in a given period of time, it is in the interest of CPU makers, motherboard makers, and chipset makers alike to limit the number of possible combinations of motherboards needed. If any one component is unstable, then the user experience will be negative.
BUT, as we generally require new types of features (such as DDR2 v.s. DDR1, PCI-x v.s. AGP), you absolutely require a new MB from time to time.. But, you want this new generation of parts to have as few combinations.
Wasn't the whole point of the Kyoto protocol to pay large sums for pollution? Be it $500 million or 1 billion - that's just the short-term effect on economy, the idea is that by taxing polution NOW
growth == greater consumption in future. enhanced efficiency == less consumption per capital enhanced efficiency ~ technological advancement growth costs entail:
cost of status quo
cost of adding new people and virtual people (such as businesses)
cost of doing newer things (flying v.s. driving, disposing dipers instead of hand wringing cloth)
cost of research to advance technology... others
Thus efficiency is part of growth cost.
Moreoever, taking a step back often reduces the value of existing efficiencies (economies of scale , excess profits due to consumerism which are often allocated to research or upgrades of machinery/processes to existing technological levels).. Thus ending growth, or worse, enducing a recession counter-acts efficiency, so the reduction in consumption means less activity AND greater cost per unit of activity.
Thus, take a small European country that is currently growing and consuming a lot of resources (where clean air is considered one consumable resource). Assume it pollutes heavily, but produces a lot of goods (energy, merchandise, etc). Let's say its overall efficiency is above average is it uses the most dollar-efficient technology (the cleanest technology usually isn't the best dollar value technology; you pay for cleanliness). Since it's a small country, it isn't operating at peek efficiency, but it isn't practical for it to grow much more. In other words, the country is full of industry with good management; making appropriate trade-offs.
Now we introduce a new dynamic.. The value of clean-air today has increased. Lets say pollution is starting to leave visible (though not yet repulsive) negative effects on the immediate surroundings. Moreoever, a Kyoto-like treaty is being heavily adopted everywhere and the fear that the local government will implement it is on the horizon.
There will therefore be a shift in the optimal operating environment. However, since there aren't sufficient economies of scale, we don't see a graceful transition. It is staggard, and may leave some companies bankrupt (through letigation, existing competative markets not leaving room for increased costs etc). The general effect will be that the country will move further away from the optimal operating environment; often being cheaper to simply produce less than to upgrade equipment to cleaner (and thereby higher operating-cost) varients.
We're talking well managed companies in this country, so the prices sold to customers is based on well researched consumer-preferences, so an increase in cost can not merely be fully thrown into higher prices - instead there MUST be both an increase and price and decrease in production effecting to a decrease in profits. As stated above, the effect is a reduction in research and implementation of existing better technologies.
So now a year later, if we compare the two possible futures of the same country.. One had a normal progression in technology and GDP, the other had a reduction in GDP, technology and consumption. Kyoto would merely look at the reduction in consumption and crowned this as a victory. But the country is now operating less efficiently. Cleaner technology will be implemented non-uniformly throughout the country, but at the expense of other technological innovations AND at the expense of higher operating costs (furthering future year's decline).
What does the lack of efficiency mean? Well, for one, a continued reduction in innovation, which includes a reduction in efficiency generating techniques.. Thus the two future paths (one Kyoto influenced, one not) will ironically have a reduced potential for clean technologies a decade down the road. Ironic, because one would assume that with such emphasis on pollution, technology would foc
Just goes to show you that when our "elected representatives" look at us, the electorate, all they see are pockets to be picked. Whose idea was it to concentrate all that power in the hands of the very few, anyways?
I dispise politicians just like I dispise lawyers; not the people, but in a manner similar to the movie "Dogma", where you know the friend sitting next to you will eventually become your mortal enemy.
That being said. Lawyers and politicians are essentially performing a single duty, making visible gains for their client. Much like a corporate president or CEO, the person performing the duty is irrelevant, only the results they are able to achieve. Who-so-ever can achieve the greatest results will be put into the position. Obviously, those that are willing to take risks (e.g. avoid morality in so far as they don't get caught or more importantly bad-publicity), have an advantage over those that don't.. The optimal politician is one that is imoral, backstabbing, coniving, lieing, yet appears to be wholesome to the public.
Most people that get in to politics are either idealists, or lobyists at heart. There is very little difference between an idealist and a lobyist.. They have a vision or a desire that they want to achieve at the cost of lesser important things. An abortion lobyist honestly believes their cause is more important than things such as fiscal viability. A bigotted lobyist honestly believes that suppressing deviant life-sytes is more important in the long run to their country than which country we're going to war with. As controvercial as a lobyists views are, they are the central point of their negotiations. What is an idealist if not someone that has a set of values that they hope to promote at the expense of things not part of their values?
I like McCain because he seems to value meta-issues.. things you don't ledgislate on.. The particulars such as abortion, war, etc are topics to be individually debated using the framework of good moral conduct.. The outcome, therefore is less important that the restoration of legitimate debate (which has been absent in this country).
Now with the positive and negatives of politics expressed, taxes lie at the heart. Politics unfortunately is the art of trade in peoples lives. It is inherently evil in my opinion for it commoditizes us in ways [evil] corporate America never could. The "I'll scratch your back if you scratch my back" is VERY important in a senatorial body. The cost of giving away something one's constituents don't want to give up (higher taxes, no new roads this year, nuclear waste in our back yards), but isn't likely to get one unelected, for something one personally needs or their constituency predominantly needs (such as war, no war, abortion, no abortion, hurricane-relief, etc). You strategize finding the key people that you need to make trades with to minimize your district's costs, while maximizing the probability of attaining both your and your disttricts goals (hopefully the two are compatible).
This tit-for tat usually involves the following logical argument:
Case 1: district 1: nominal services district 2: nominal services tax in each district: $100
Case 2: district 1: high local services local tax: $50 fed tax: $100 district 2: low local services fed tax: $100
district 1 looks bad for residents.
Case 3: district 1: high service bundle1 district 2: high service bundle2 fed tax: $200
Now both districts are relatively equally off, so residents can not complain that it was THEIR representative that is failing them.. Both have services to show to their locals. Everyone is sharing in the cost burden (though some districts will get more utility than others'). The politicians win all around.
So case 3 almost alawys is favored.. So you're essentially LOOKING to add pork to ledgislation, because society is "better off" all around.. It's like the MS tax,, by having EVERYONE pay an extra $20 for the
I don't know the situation in Sweeden, but it occurs to me the following:
There are certain fixed expenses for any country no matter the size. If you publicly build a road, a court-house, etc. Then there is a minimum public cost. Thus the smaller the country, the greater the burden per-capita for such infrastructure.
In the US, Social Security and infrastructure needs are relatively small portions of the tax pie (no more than 30% combined)
The other issue is that things like social-security aren't really taxes, but social engineering. Basically a government is doing 3 things: 1) wealth-redistribution, giving poorer people a better deal than wealthy on retirement 2) forced savings (which runs at the cost of reduced GDP) which alleviates welfare on retirement 3) reducing inflation by removing the public's ability to spend on gambling institutions such as speculative financial markets and real-estate. The US, for example, requires S.S. dollars to go into US Treasuries (which of course is convinient for itself).
As for gas tax, thank green-peace. Our taxed vice is smoking.
On the one hand, you could argue that it's not fair for smaller countries, but you could always aggregate your infrastructure costs with neighboring countries. Otherwise, you're getting value for your high taxes. Surely not every dollar's worth, but definitely moreso than if you had no taxes at all (and thus no civil infrastructure).
There is no reason why a JVM couldn't release memory that isn't needed anymore to the OS. However, since memory resources are always in a state of flux with Java, the likelihood that hitting such a high water mark again will occur is high, so it generally isn't worth doing. Java only claims extra memory from the OS IF the programmer was dumb enough to hang on to too much memory.. This is a notorious problem with Java programmers.. They think it's too expensive to construct new objects, so they cache them ALL. Even the stupidest little ones like lists of data. So each of the thousands of classes each holds on to little caches of data, which point to other caches of data, bla bla bla. The end result is that the GC winds up doing 2 or three full sweeps and doesn't recover enough memory, so it asks the OS for more..
The way it's supposed to work is that object allocation is embarrasingly fast (compared to most other architectures), so you allocate what you need and release it as soon as you don't need it. You have thousands of very fast mini-garbage collections this way, and never get above a 32Meg foot-print. (Yes, I know 32Meg is large compared to say perl). I often run full web applications that never grow to more than 32Meg.. However, when doing full caching DB's I grow to over a gig in size. But the performance is worth it in that case. It's only when some thread-locals dont' release their wears that I find memory "leaks", which aren't really leaks, since the next time that thread is re-used, that memory will be freed and thereby recovered.. In the mean time slow (2 second) full GCs need to be performed. But it's still bad programmer design, not the VM itself. A c program written in the same way would leak just as badly, and lets not forget the propensity of c++ programs to never even free objects it's not done with.
It's a proportionality thing.. Often, the difficulty of programming i18n is less dramatic than the problem of actually generating all the different language sets.. Often such management means an advanced GUI (maybe not advanced, but certainly more than just raw field entries for a DB-backed widget).
i18n is generally token language-set in a 1:n relationship.. Which maps nicely to table layouts, thus I don't see any need to create i18n support in the DB itself.
If you want some degree of abstraction, java provides the ResourceBundle for which you can easily write your own DB-backed loaders. You instantiate your bundle, then pass it around to whatever device needs to render the actual text.
There is lots of i18n support in the ASP/JSP environment (I assume in PHP as well). jstl, struts, webwork have nice end-to-end support for i18n (error messages accepting tokens instead of raw text, etc).
For manipulation of static sets of text, there are generally plugins for your editors which allow you to manage a suite of bundleName_locale.properties with color highlighting, missing-term warnings, etc. I personally use the properties plugin for idea in the java environment.
Anymore, you should conciously be evaluating if anything is ever being displayed to the end-user, and organizing that material into i18n bundles of some sort. Standardizing on whatever your platform natively supports is critical because you can leverage the tonns of tools that are here and are bound to come.
Yeah, some are security consultants, white hat, etc., but I don't think these are what Ranum was primarily talking about.
But, according to my argument, it is irrelevant whether they are indeed white-hat or not.. Someone hacking into your site is always bad, if nothing else, there is an exchange of ego between the hackor and hackee.
My point is that any hack that does not involve a massive financial or legal upset is a good hack, since it adjusts the priorities of everyone involved so as to avoid a future occurance. Such a "lock-down" serves, if nothing else, as an example to the company's peers... The IT guy that has been FIGHTING to get past the beuracracy to lock down the local network now has evidence to support his claims... He is likely to get better funding, and will be able to coorse employees by removing admin rights, etc.. Something that is VERY hard to do if not already in place. This is especially true in Universities where people feel like open-access is a requirement of their job.
Finally to re-iterate, my thesis is that a slow migration of hacks is not only good, but essential for survival. Not having an ever increasily sophisticated competitor will leave you unable to cope with serious challenges ahead. It is only when the competition advances faster than you that there is a problem; rapid changes against your favor. This is how species die out, and this is how industries collapse.
Though I haven't read Wealth of Nations yet, I don't suspect that Adam smith ever said that greed was moral, but the existance of greed needs to be recognized and factored in to decision making.. Likewise, I make no claim that hacking is moral.. I do not necessarily encourage new would-be-hackers out there, but what I do say is that competition [to an otherwise altruistic and trusting society] exists as a fact of life.. Prosecute the financial obstructors, but learn to embrace the environment of natural distrust; this is the animal kingdom as it has survived for millions of years (without serious interruption).
To say hackers are evil is like saying germs, viruses, and carnivores in general are evil. By merely acting out Adam Smith's society being benifited best by each acting in his own best interests (adapted by John Nash to include societal interests for best outcome), we are keeping in step with mere nature.. A dog will forage for food, defend it's food, and kill it's food, so that it can stay alive. A rabbit will defend against other rabbits if need be (though they'll generally run away from anything else).
A patron is looking for a good deal, and will expend effort to maximize their deal, so sloopy wording on a sign on your store-front are invites to a natural onslaught of fiscal frustration. By natural, I mean there is no evil intent in people trying to keep you for your word in maintaining a good bargain (that you didn't intend).
If there is money on the street, it is conceivable that: a) the original owner will never find it again b) someone else will take the money
So you justify taking the money yourself.
If you are hungry, you might be inclined to take two samples at a free food-sample kiosk. It's unfair as it goes beyond the intent of "sampling" and takes away from other's (since there is usually a set amount of sample provided for the day).
In reality, those that are sheltered from such harsh survival of the fittest environments will EVENTUALLY meet with that environment.. It is impossible (short of death) to avoid it. Thus the question is not IF we will meet our challenges, but when, and how quickly will the difficulty level rise.
For those with assets we fear to loose (time,money,posessions,intellectual property, etc), it is natural for them to be saught by others. Having a public wiki is valueable advertising real-estate (or a personal repository for globally accessible content). So grafiti, being merely a primitive form of marketing, is bound to happen. Bank accounts are an obvious point of content.. If you happened to come across money on the street, you are more than likely to take it. If your ATM machine started allowing you to withdraw cash w/o deducting from your bank account, there is a better than likely chance that you'll take advantage (anonymous theft when it is considered to not overwhelmingly harm someone else - proportionate loss/gain - is often self justified). There isn't much difference from taking from that ATM machine and taking from an online bank account that you've happened by. Yes there is a greater issue of proportionality (you might be stealing from someone poorer than you), but you might think to yourself (I'm teaching them a lesson).. What-ever the cause, an otherwise moral man may find themselves tempted.. To say nothing of the mafia.
And ultimately organized crime is the tyrannasauras of our internet age. The mafia being only one form of it (unfriendly governments being an even more serious threat). The age of mafia and internet "WAR" (literally between nation-states) is only a matter of time.
So if our "evolution" through natural selection and adverse environment does not "toughen" us enough to sustain such natural phenomena, then we will die (or at least the medium will die).
So lets look again at these "evil" hackers. Many of the hackers were self-professed white-hackers, or anonymous exposers. If you are inclined to see if a WEB-INF directory or IIS-specific file-set are visible on a public site, you can either email their sys-admin who might sue you for hacking, or simply ignore you (like MS tries to do with serious security alerts so long as the general public is oblivious), or you can make it a priority for them... Deface their web site, delete lots of their database records.. Make it too expensive for them NOT to resolve the issue.
These are altruistic people. Slightly less altruistic are those that advertise themselves 3l33t hacker-names advertised here and there. As they have the fun and recognition-factor of it all (especially if they get CNN coverage).
In this interpretation, I go back in time, and kill my great grandfather when he was a baby. Well, all that happens is that I fork the universe into two possible histories.
My main problem with this theory is that in order to have these two "alternate universes", you'd have to have twice as much matter in the universe. A timeline and space-line should be continuous (my own conjecture), thus for an external observer (I know, impossible in general relativity), you can hold fixed x,y and time coordinates and trace through a z coordinate and see continuity. Now the z dimension can contract, warp, fold, or even loop (which facilitates the idea of more than 3 spacial dimensions as described in string-theory), but it should still be continuous.. Even pinch-off points like black-holes represent continuous regions of space.
So take the worm-hole representing a space-time-loop back into time (namely you present hyper-energetic force onto an asteroid causeing it to warp space-time, such that the other end point brings you back to your living room 20 years ago). By passing through this hypothetical tunnel, you are passing along a continuous path that brings you back in time.]
For an "alternate" reality to be born of this new path-forward, there must be a new continous line that passes through an entire universe that use to exist, but with an alternate path forward. It would be like the infinite spiral.. If traveling through space-time in our relality, you could pass around this worm-hole, never going through it, and never experiencing the changes relevant to it.. But by going through a couple thousand cubic-miles of the region OF the hole, you are passing through a micro-fine alternate universe.. You have essentially compacted an entire alternate copy of the universe into a tiny region of space.
So the energy required to generate this alternate universe would be the fraction of the total energy contained in the universe for the time-space-period it spans. In other words, the amount of hyper-energy required to create the warp is simply insane.
This isn't to say that such regions of space do not exist.. The universe may have sub-divided at it's creation, sending enormous fractions of the total space-time energy into such small regions of space producing mostly symmetric alternate realities. What I suggest is that to newly create such a region would be implausible.
Even if this isn't how a worm-hole that warps time would exist, I still conjecture that energy must be conserved in any medium that allows energetic material to pass from one point in the time-line to another (since general relativity would seem to hold symmetric time to the spacial axises). To alter history would require either a finite amount of energy to pass back and assimilate into the known history of the universe (as the master article suggests), or is simply not possible. But I can not fathom the synthesis of an infinite amount of energy required to provide a continus tunnel allowing material from the present to produce an infinite number of time-lines.
The only other possibility I can consider is the concept of wave interference on particles.. Namely the double-slit experiment.. A single particle travels as if there were an infinite number of universes, each having a particle travel a different path.. But the particle feels the interference forces as if they were all part of the same universe.. Namely the fact that there is a bean of light going through two slits COULD mean that there are an infinite number of photons going through both slits and an interference pattern emerges.. But when we can measurably demonstrate that only a single photon is being fired, the photon still acts the same... The master article talks about the difference between measured and unmeasured states.. The interference pattern occurs during an unmeasured portion of the photon's travel. In fact, measuring along the path would cause new interference patterns, thereby disrupting the final pattern, so it is impossible to
. Accessing a global or shared resource? Lock. It takes one line of code (& one to unlock). It becomes second nature.
While I generally agree with you.. It isn't as simple as just locking every time you access a global resource.. You have deadlock issues. So you either need a monitor thread or a global lock which is required to acquire a subset of locks. Or if you feel particularly daring, you can have a convention to always lock in a particular order. The problem comes when you have different modules which acquire different locks, and you can't control the order in which the modules are invoked for a particular event/task. This becomes increasingly difficult as your application grows appendages..
Your best bet is to minimize the need to access such global resources, and to have locks encapsulated and atomic, and more importantly, not farming out to acquire other locks. But this isn't always easy to do. Minimizing global resource access implies making more use of local variables.. Passing instances up the call stack instead of having modules search for global resources.
In servlet land, you have a synchronized application context Map and a per-user synchronized Session Map. Then you have a non-synchronized request Map passed up the call tree. Here you are explicitly defining the global-ness of operations. Moreoever servlet designers recognize this explicit scope separation and place shared resources at the appropriate level.. Anything being placed into the Request scope need no synchronization since it's single-threaded access.. When you place something in the other scopes, you are making a conscious statement that you must be mindful of MT-safety.
for (E e : input) {
F f = operation on e;
output.add(f); }
But allows the generic application of farming out to worker threads. Essentially explicit SIMD operation.
This could be the actualy hibernate re-materialization, or reporting or aggregation.
The point here is that this is a simple example of a procedure-local variables being used in an MT way.
As for concurrent access to "entities"; Objects which have uniquely identifiable properties (as opposed to value-objects like a date), systems like java provide a relatively easy way to access them.. Namely object-synchronization. This is most perfectly exemplified in Collection/Map synchornization. Maps are repeatedly used in MT environments in java, yet they are very reliable. (Rather when using Hashtable or the Collections.synchronizedMap() wrapper).
Concurrently entity modification in today's environment can easily utilize "optimistic" locking, whereby the entire transaction can be marked as repeatable.. You initiate a repeatable transaction. You load all relavent entities from a shared environment (such as hibernate or directly from the DB), then you update a version of some sort (like a modified-timestamp on each entity). Then you go to push the objects back to the share environment.. But in doing so, you check to make sure you have the latest version.. Otherwise you throw a concurrent modification exception... The transaction is rolled back as if nothing has happened.. At which point you minorly annoy the user by providing an error screen, or you intelligently catch the exception and restart the entire process.
The above obviously is most suitable when there are low probabilities of conflict, as this provides the highest concurrent execution performance. Not to mention is the least intrusive form of concurrent execution.. The only thing it requires is adherence to a transactional model.. But mysql's INNODB or postgres provide this nicely without an EJB framework.
When overlapping modification is too likely, then you can utilize explicit blocking locking methods. Often such locking occurs at the DB access point, not the application level. So assuming that you have an adaquate DB abstraction layer such as OJB or hibernate, then you're safe even here in terms of writing bug-free code.
When a mixture of performance and concurrent modification of the same items is paramount, then you'll probably have to resort to object locking (or even finer grained locking).. But I would venture to say that this is rarely needed.. As the benifits of writing the slower but bug-freer code and purchasing more expensive hardware is not to be under-estimated.
Of course there are programming models that don't exactly conform to a web servlet / mainframe (SIMD) work-flow. So your mileage may vary.
It's more than just setting up an environment as immutable.. What the parent post was talking about what procedure-local variable usage.. Statics are obviously an obsticle to MT, but so are field-variables. If a class is instantiated w/in the method and used by the method and invoked methods, then you are inherently thread-safe. For the purposes of sharing objects between method invocations, you have immutable singleton "verbs", and noun immutable "Value Objects". What's left are noun mutable "Entity" Objects. By separating design into verbs, value-objects and entities, then you know where you have to focus your synchronization efforts.. Mind you it takes WORK to maintain the value-objectness of non entities.
The typical process for "verbs" is to first call set-attributes, init, then execute, then get-attributes. The problem is that if the verb object is passed around there is a constraint of single-threadedness.
If on the other hand, the verb is written as a singleton which returns value-objects, then there is no dangerous of having multiple threads acceess the verb. But it often requires creating multiple transient objects.. For example:
State s = myFactoryObj.setupVerb(attr-list); s.execute(); s.getAttr();
But as the name implies, being an entity object, it is likely to be used in multiple contexts.
By separating verbs from shared mutable entities, your code is more thread-safe, and as a bonus, more unit-wise testable.. You can trust that the entire "login process" works independently of a user object.
The book domain-driven-design is an excellent read on such architectural separations.
Such techniques are applicatable to even non OO environments. Side-effects are evil and stand in the way of many things. Making state modification explicit, you can more correctly identify synchornization points and more importantly reduce the number of them.
Re:Intel working on silicon laser to link cores
on
AMD Quad Cores, Oh My
·
· Score: 1
Not all apps parallelize well, and parallelizing apps is difficult in any case, so Si Lasers are not likely to yield a big breakthrough in performance.
Ok, imagine this. You put on one machine, a database, a java server on another machine running with 500 threads, and an apache server on still a 3'rd machine with mixed mode multi-threaded, multi-processed operation (to facilitate a hetero-genous php, perl, static-file, java proxied, https'd environment). This is a simple e-commerse web site.
What is the advantage of putting this on one machine? Reduced latency. Tremendously reduced administration. Now what about fail-over? Ok, you build two identical machines and clustering techniques. Now you have two mirror machines (now tremendously easy to administer).. You give up some performance in clustering, but you have no single point of failure. Compare that to the usual standard of having a single database machine with a clustered web environment.
Having one (two) machines means reduces UPS, heat, administrative, power, etc requirements. This is what blade servers are all about, but even better.
Now it is entirely likely that multi CPU machine will be slower in through-put than two independent but chained computers (web + application + DB machines). But the latency is certainly traded in this situation. To compensate, it it likely that having a few unique instruction sets (httpd, JVM, DB-code) will live in common sufficiently larege cache enviornments, thus the competition for resources may be minimal enough to not be negative.. Moreoever, UNIX sockes are significantly faster than TCP sockets even on a local machine, to say nothing of the latency rich network connection.
Ok, so what about non mainframe style "small work, LOTS of jobs" operation? Take a work-station, where you have 20 major applications sitting mostly idle, but taking up enormous resources. The work-station process involves one or two major applications being active at a time, but constantly switching between them. If there is a single CPU, then there is a high probability that one will be performing background activity.. Especially for things like compilation, where you'll switch to another task while waiting. Even if a particular task has a max number of practical CPU's that it can fully utilize, having a "spare" idle CPU is GREAT for task switching.. There is no delay in contexting IN one application while either contexting out the old app, or simply keeping it active until it's finished it's current task. Granted some switching occurs 60 times a second on single CPUs, how much of that work can be avoided by not having to even switch out a task for a 32CPU environment. Taking two apps running concurrently on a single CPU always runs slower than running them sequentially. So such environments, even if only giving a small percentage gain on average can still provide a happier user experience.
Finally, take the end-user desktop environment. As other's have pointed out, innovation has typically lagged proliferation. 3D games didn't take off until we have 5 different standards in hardware video-acceleration. Eventually voodoo stood out from the pack and dramatically increased the bravery of game makers, since there was a legitimate market to show-case incredible features , while backing down to the lesser capable graphics standards.. Eventually nVidia gave validation to the DirectX standard, and after quality issues were practical to overcome, that market exploded; building upon itself; faster features, more supported software..
Today multi-threading on the desktop exists PURELY for user-responsiveness, not for performance, since 99% of all desktops are single-threaded. We now expect that the menu for an application should continue operating even when it's doing some intensive task. In fact, we feel upset with software that does not adaquately thread their UI. This is because that's all that's really possible with single-CPU hardware.
The problem I have with your synopsis of Mac penetration are:
1) Legacy Standards... VHS beat Beta because of proliferation, NOT because of quality. People are use to windows, getting them to switch to ANYTHING else is a major obsticle.. Now Loghorn might be the key to screwing everything up for MS.. It's like IBM going to MCI architecture when everyone else went EISA or VLB. IF MS forces people to switch to a totally different UI experience, then something like Apple which will have a proven user experience by that time will have a chance.. But this assumes loghorn is a radical change.
2) User types.. There are 4 user types that I can think of. Gamers, developers, business, and media. Developers are specific to their platform of choice, so making one OS nicer than another is not likely to sway them in any direction (plus they're a minority (and Java support sucks on Macs)). Gamers have DirectX or consoles. The Mac doesn't have a prolific user base.. And you suggested yourself that people will need to dual boot.. Games will definitely need native execution, since MS isn't about to let other platforms utilize DirectX. More and more people are switching to consoles now that they've finally woken up and found how to use the state-of-the-art in graphics.. Nintendo's and PSP's were a joke to me when I'd read about the latest voodoo, then later Geforce cards.. Now the consoles are getting seriously competative, and they're easier to deal with as a gaming system (not that I'll ever use a console). Media doesn't have a strict advantage on any one platform. You can play/rip/encode a DVD just about anywhere.. This is really a choice of the more popular softwer vendors what platforms they want to support. Mac and Windows definintely have an edge over Linux in multi-media integration into the OS, but I don't consider this a good thing, especially where DRM is concerned... Frankly, I believe MS has the edge on DRM, since if you are a media company, which standard are you going to support, 3% market share with potential for growth, or the yellow book? Finally businesses.. They need perfect support for the latest office documents out there.. They need conferencing do-dads.. They need stability.. They need exchange / lotus-notes what-ever. MS works hard to lock businesses into their product lines. Apple isn't 100% compatible today, and I suspect MS will innovate ways of keeping them always in that "well, it 'mostly works'" category.
So there are no demographics which make me say.. Yes! there's where they're going to find their niche and expand from there.
That's why IBM is way behind Intel when it comes to device size.
I don't know, I'm just asking, but this doesn't make sense.. Why would IBM be having trouble shrinking the die due to SOI being the problem IF AMD is using the exact same technology and getting 90nm + lower power? What does AMD know that IBM doesn't?
My guess is that it's a function of finance (sales volume), not technology. It probably costs AMD boat-loads to do the SOI, but they have enough volume to justify it. How many chips of a given architecture does IBM make? Is it cheaper for them to put more transistors and better cooling on a Power chip than to shrink the die and up the clock? Probably. The only thing IBM makes that needs better price/perf are the PPCs.. So they obviously get the shaft because of too low a market volume.
Slashdot Mirror
DLP's generally have a sucky side-viewing angle. If a color-wheel is in use, it can sound like a jet-engine. Moving parts can be volitile. Many people can see a "rainbow effect" due to the strobed color consolidation of a color-wheel.
That being said, DLP's are the only "high end" option for people on a budget. Unless you're willing to go w/ Westinghouse/scepter/maxtent LCD's. E.g. the zenith of our day.
And the term organic has a chemical definition. So by merely putting a chain or two of the appropriate molecule into silicon, then for marketing purposes, I could label my computer:
A clever going-places mid-level management kind of guy.
It would be technically correct, but wholely loathsome and misleading.
You don't write 'I'll be back in two hours' memos in ancient egypt and stick them to the fridge, if the intended recipient never had shown any interest in archeology.
Except that the woman is not trying to convey information via her body language. Women recognize that both Men & Women can recognize "Anger", "happiness", etc. Thus when a woman recognizes subtle discomfort, interest, lack-thereof her experience gives her a subconcious reinforcement that the man recognizes this as well.
Women interact in their own circles and have the benifit of common biological cycles and common experiences. So when they are amongst themselves, there is a symbiosis; buliding apon each other through communication and support networks (and flat out a common agenda). When interacting w/ men they don't achieve this same experience.. They are incredibly frustrated because we do not react the way they intuitively expect (and by intuitivly, I mean empirically; there is no mysticism going on here). It's as if you were riding a bike and expect minute weight shifts to cause the bike to operate in a certain way, but when you ride someone else's bike, the body motion had completely unexpected results... Because this is internally learned communication, you are completely disoriented because you can't retrain your body properly nor do you fully underestand WHAT isn't working.
Notice the concept of a tom-boy.. Let's suppose for simplicities sake that male-ness and female-ness (e.g. the propensity for transgender attributes) does not apply. That we are (as the above has suggested) highly affected by our empirical understanding of the world. A girl that does not fraternize in homogeneous female circles will not have such a dramatic communication problem with men. Their expectations are consistently met, and there is little disconnect w/ men. The problem, however, is that they are not disconnected with female circles. Generally rejected as they no longer know how to communicate properly nor express acceptible behavior.
This doesn't even take into account the petty content of all-male or all-female circles (cars, toys, sports, science, sci-fi for boys, and boys, celebrities, fashion, etc for girls). The lack of common interest only exascerbates the disparity. But is no moreso dramatic than between geeks and jocks.
But the common attraction between men & women at least brings the two separate worlds together unlike geeks and non-geeks. (or even between different sects of geek). Each gender looks down on the other gender because they can recognize obvious defficiencies. "He can't communicate", or "She can't reason". "He is insensitive", or "She is overly emotional".
I've just finished a great book called "Getting to the Yes" which describes how to come to agreements w/ people more effectively. The key element in the book is that you have to recognize that people are emotional creatures. Both men & women have pride, can feel threatened, feel insignificant. There is a definitive pecking order in both male and female circles, and if you feel like you should be equal in a relationship but the pecking order says otherwise, there can be depression (for either party). So when intereacting inder-genderly, you MUST recognize these differences and encorporate them into your communication.. For men interacting with women, there are obvious strategic implications here. For women, they have to recognize that there are ramifications to their tone and actions. Women are good and adapting the focus of a communicay based on changing emotional needs. They need to recognize that this can be a good thing when they "feel" that the conversation is not going to go forward successfully. Their feeling is correct; their ability to natively read the body language gives them the tools to steer them; they need only the resolve to bring about the communicay in a direction devoid of the uncomfortable feeling. If this means changing their position or delaying the conversation (by switchign to a happier
You do know that the entire currency system these days is just based on "good faith"
Bartering has always been a sense of good-faith. People would bite coins to semi-verify their authenticity. They would be suspicious of the quality of the animals that were used to make purchases.
There have been numerous issues with counterfitting gold, and the enormous issue of shipping massive amounts of gold from country to country; piracy, insider theft, organized harbor crime, etc. Thoughout the centuries, the management of gold has been an issue.. And once a cache of gold was found in a country, the country would become artificially wealthy. Look at the rise & fall of spain.
If money grew on trees, then economies would collapse, and the core intent of efficiently allocating scarse resources (such as food, water, land, etc) would no longer have meaniful trade.
When you standardize an open commodity like gold, you have to protect it. When you standardize an artificial and hard to reproduce fiat like coin/paper, countries are suddenly being valued by their assets, not their luck of finding a very specific type of natural resource. Being lighter and more compact, it's easier to protect and transport.
There are other economic issues involved with using an intrinsicly valueable commodity for currency. The "value" of gold fluctuates. This means, no matter what the state labels the value of gold, the consumers (who wear and decorate their homes with it because it's pretty) will have their own aggregate value. When there is a disconnect between the state sanctioned value and the consumer value, arbitrage can occur. This would be something similar to if the marginal cost of producing a penny or dollar were 100 times greater than the value of it. With gold, WHEN this occurs, people take their pay, melt it down, and resell it until the supply-demand equilizes ; lowering the price to match the state's stated rate. Or until the state wises up and raises the value of gold. When gold is too low relative to it's value, people will melt down their jewlry and fashion coins.
When two countries have different values for gold (e.g. I can buy a chicken with fewer gold in one country than another), then people will trade gold until their effective prices are equilized (arbitrage). The negative is that improperly set gold prices will form a massive egress of their stockpiles of gold.
In the old days, the cost of transporting something over-seas was VERY high.. You had piracy, heavy slow ships.. You had to pay the crew well to get honest enough shippers.. All sorts of things.
Now you can put it in an anonymous box for the most part and have a pretty high degree of trust.. Moreover you can ship billions of dollars worth of assets. Arbitrage is almost instantaneous in the modern world.. We no longer require physical presense of our assets.. We purchase "orange juice futures" by the billions while the commodity actually sits in another country... Likewise with gold.
A country in our modern world would very quickly find it's "currency" dramaticlly shifted due to minor fluctuations in the stock-market. And currency is extremely important to the economic welfare of a country.
If we were still on the gold standard, then inflation would NOT be controllable by the fed (who's job it is is to manage our currency). Instead of buying/selling US treasuries, the fed would have to directly buy and sell gold. But in econimicly tight times, they would eventually deplete the gold stock-piles, and surely congress would limit how low the stockpiles could be depleted. Again, remember, they never have to take the money out of fort-knox to sell it.. A serial-numbered pilon is merely designated as owned by someone other than the US government.
The 1999 forward recession has been dramatically avoided because the fed has been pumping virtual dollars into the economy like mad.... Inflating the economy. Whether you believe this is a smart move or not, the fed would NOT have this
The alarmists aren't happy unless they're running around screaming "the sky is falling!".
In positional arguing, one side needs to over-exaggerate their position such that the "compromise" is closer to their intended goal.
Forget global warming, and other death-to-earth claims. (Nature will have the final word here.) What about industrial pollution of drinking water and the rise in asma and lung disease in cities due to direct human-affecting pollution or freon leakage..
There are so many more direct problems associated with cost-effective pollution. Maybe people care about it, maybe not. But by extending the problem to killing off innocent people that don't live/work in urban areas, it becomes a moral crisis.. The only problem is that these people go too far in their plausibility and their arguments are thereby dismissed.
costco online is about to have a $1,700 full resolution 1900x1080p 37" LCD w/ full inputs (DVI, multiple HDMI, component, VGA, composite, s-vid).
We're not talking about supporting current hardware. We're talking about supporting a content format; essentially the software. The cost of the DVD player is NOTHING.. We're talking about replacing thousands of dollars of DVD collections. Are you going to gamble this collection upgrade on one format when another one might win.
I'm assuming we're not going to see dual releases of the same titles either.. It's like XM v.s. sireus (sp?). Or HBO v.s. Showtime.. Different content which you'd have to pay double to the full range.
I'm pretty sure DVD doesn't support 1080i.
No, he means the player. HDMI supports a specific sets of resolutions; standard DVD is one of them.
Currently you can get hdmi-capable DVD players. It passes both audio and video to your hdmi TV. I believe there are even HDMI audio recievers; though I don't know if DVD players have dual HDMI outputs thus that seems kind of silly.
Not all new TV's have HDMI inputs either. Many have DVI which can have an HDMI adaptor applied (if it doesn't ship with the TV).
From my most recent research, what I've found is that the optimal cost-effective way to get 37+ inches of flat-panel is to get the clearest-picture TV that you can afford and get a DVD player with high quality conversion to one of the several HDMI signals.. A more expensive DVD player is going to be much cheaper than a high quality conversion device on the TV.
The end result is that HDMI is a preferable solution going forward. Most likely it is going to win out; we're just going to have to accept the associated DRM.
I'll leave for further discussion the techincal merits of displaying full resolution 1080i or 1080p in analog or even through DVI.
Consumers are not stupid.
Maybe not stupid, but definitely powerless. Boycotts don't exist anymore; they have to be forced on people; e.g. international embargos, regulations, militarily-enforced cartels. Facing a lack of options, most modern people (especially now-minded Americans) will take the best of what is available to them. As a result, anything short of fearing for safety, isn't sufficient to maintain a boycott.
I'm not aware of flash memory handling 150Meg / second random-seek time. And definitely not for writing.
As for 16 hour time, I always hook $80 UPS's to my machines, so if your machine is off but you're hooked up to a UPS, you'll have more than a week of life if the power goes out. Yes if your machine itself is on when the power goes out and you're not there to shut the machine off (thereby letting the UPS feed the memory drive) you'll have a problem, but even then, most power outages are less than a couple hours.
Price wise, if you're talking about a $500 machine, then the price is of course impractical.. But if you have a $5,000 server, a $2,500 work-station or a $3,000 gaming machine (including monitor + speakers + else), then the price isn't bad; just looking at it as one additional performance enhancement.
If you're a dot-com company and only have 10 gig of database data to worry about, you could set up several drives and partition your databases across them getting tremendous performance improvements (continuous random seeks). Not sure about mysql, but postgres does NOT make use of memory directly, it's memory footprint never seems to get above 9Meg per process. You can up various numbers, but they're dangerous if you're not an expert. Simply throwing a fast ram-drive is easy to deal with (with the stated above exception of power-outage). It's better than a system RAM-drive solution because you have more reliable fail-over (no need to copy to-from physical disk space on start-up shutdown/crash).
The consolidation to 939 was probably nothing to do with having a "foreward looking standard" as having fewer current generation parts to test.. Every quarter, there is a new technolog, either driven by the video card industry or the CPU industry. The common denominator is the motherboard.. So in essence, choice of a MB is more critical than the other periferals like CPU / vid-card, because it limits your choices.. Now imagine that you are a MB manufacturer and there are two major CPU makers, and EACH maker requires 6 different configurations. THEN, you have 4 different classes of chipset makers (SiS, Via, AMD/Intel(native), nVidia, and soon ATi). To provide a complete suite of choices for the customer (so nobody can say, hey, ASUS, I like you, but I needed this particular configuration).. If as an MB maker, you picked the wrong horse (ATi, v.s. nVidia for your high end chipset), and your competition chooses the opposite, then there is a generation of customers that have cozied up to another name-brand.
As MB makers produces the dozens of possible combinations, they have to test these guys, and invariably, some have problems.. That's why you have different versions of the same motherboard (though they never advertise which version you're purchasing). You generally never want a rev1, but by the time you get to a stable rev, the next generation is out with dramatic enhancements.
Thus, in a given period of time, it is in the interest of CPU makers, motherboard makers, and chipset makers alike to limit the number of possible combinations of motherboards needed. If any one component is unstable, then the user experience will be negative.
BUT, as we generally require new types of features (such as DDR2 v.s. DDR1, PCI-x v.s. AGP), you absolutely require a new MB from time to time.. But, you want this new generation of parts to have as few combinations.
Wasn't the whole point of the Kyoto protocol to pay large sums for pollution? Be it $500 million or 1 billion - that's just the short-term effect on economy, the idea is that by taxing polution NOW
... others
growth == greater consumption in future.
enhanced efficiency == less consumption per capital
enhanced efficiency ~ technological advancement
growth costs entail:
cost of status quo
cost of adding new people and virtual people (such as businesses)
cost of doing newer things (flying v.s. driving, disposing dipers instead of hand wringing cloth)
cost of research to advance technology
Thus efficiency is part of growth cost.
Moreoever, taking a step back often reduces the value of existing efficiencies (economies of scale , excess profits due to consumerism which are often allocated to research or upgrades of machinery/processes to existing technological levels).. Thus ending growth, or worse, enducing a recession counter-acts efficiency, so the reduction in consumption means less activity AND greater cost per unit of activity.
Thus, take a small European country that is currently growing and consuming a lot of resources (where clean air is considered one consumable resource). Assume it pollutes heavily, but produces a lot of goods (energy, merchandise, etc). Let's say its overall efficiency is above average is it uses the most dollar-efficient technology (the cleanest technology usually isn't the best dollar value technology; you pay for cleanliness). Since it's a small country, it isn't operating at peek efficiency, but it isn't practical for it to grow much more. In other words, the country is full of industry with good management; making appropriate trade-offs.
Now we introduce a new dynamic.. The value of clean-air today has increased. Lets say pollution is starting to leave visible (though not yet repulsive) negative effects on the immediate surroundings. Moreoever, a Kyoto-like treaty is being heavily adopted everywhere and the fear that the local government will implement it is on the horizon.
There will therefore be a shift in the optimal operating environment. However, since there aren't sufficient economies of scale, we don't see a graceful transition. It is staggard, and may leave some companies bankrupt (through letigation, existing competative markets not leaving room for increased costs etc). The general effect will be that the country will move further away from the optimal operating environment; often being cheaper to simply produce less than to upgrade equipment to cleaner (and thereby higher operating-cost) varients.
We're talking well managed companies in this country, so the prices sold to customers is based on well researched consumer-preferences, so an increase in cost can not merely be fully thrown into higher prices - instead there MUST be both an increase and price and decrease in production effecting to a decrease in profits. As stated above, the effect is a reduction in research and implementation of existing better technologies.
So now a year later, if we compare the two possible futures of the same country.. One had a normal progression in technology and GDP, the other had a reduction in GDP, technology and consumption. Kyoto would merely look at the reduction in consumption and crowned this as a victory. But the country is now operating less efficiently. Cleaner technology will be implemented non-uniformly throughout the country, but at the expense of other technological innovations AND at the expense of higher operating costs (furthering future year's decline).
What does the lack of efficiency mean? Well, for one, a continued reduction in innovation, which includes a reduction in efficiency generating techniques.. Thus the two future paths (one Kyoto influenced, one not) will ironically have a reduced potential for clean technologies a decade down the road. Ironic, because one would assume that with such emphasis on pollution, technology would foc
Just goes to show you that when our "elected representatives" look at us, the electorate, all they see are pockets to be picked. Whose idea was it to concentrate all that power in the hands of the very few, anyways?
I dispise politicians just like I dispise lawyers; not the people, but in a manner similar to the movie "Dogma", where you know the friend sitting next to you will eventually become your mortal enemy.
That being said. Lawyers and politicians are essentially performing a single duty, making visible gains for their client. Much like a corporate president or CEO, the person performing the duty is irrelevant, only the results they are able to achieve. Who-so-ever can achieve the greatest results will be put into the position. Obviously, those that are willing to take risks (e.g. avoid morality in so far as they don't get caught or more importantly bad-publicity), have an advantage over those that don't.. The optimal politician is one that is imoral, backstabbing, coniving, lieing, yet appears to be wholesome to the public.
Most people that get in to politics are either idealists, or lobyists at heart. There is very little difference between an idealist and a lobyist.. They have a vision or a desire that they want to achieve at the cost of lesser important things. An abortion lobyist honestly believes their cause is more important than things such as fiscal viability. A bigotted lobyist honestly believes that suppressing deviant life-sytes is more important in the long run to their country than which country we're going to war with. As controvercial as a lobyists views are, they are the central point of their negotiations. What is an idealist if not someone that has a set of values that they hope to promote at the expense of things not part of their values?
I like McCain because he seems to value meta-issues.. things you don't ledgislate on.. The particulars such as abortion, war, etc are topics to be individually debated using the framework of good moral conduct.. The outcome, therefore is less important that the restoration of legitimate debate (which has been absent in this country).
Now with the positive and negatives of politics expressed, taxes lie at the heart. Politics unfortunately is the art of trade in peoples lives. It is inherently evil in my opinion for it commoditizes us in ways [evil] corporate America never could. The "I'll scratch your back if you scratch my back" is VERY important in a senatorial body. The cost of giving away something one's constituents don't want to give up (higher taxes, no new roads this year, nuclear waste in our back yards), but isn't likely to get one unelected, for something one personally needs or their constituency predominantly needs (such as war, no war, abortion, no abortion, hurricane-relief, etc). You strategize finding the key people that you need to make trades with to minimize your district's costs, while maximizing the probability of attaining both your and your disttricts goals (hopefully the two are compatible).
This tit-for tat usually involves the following logical argument:
Case 1:
district 1: nominal services
district 2: nominal services
tax in each district: $100
Case 2:
district 1: high local services
local tax: $50
fed tax: $100
district 2: low local services
fed tax: $100
district 1 looks bad for residents.
Case 3:
district 1: high service bundle1
district 2: high service bundle2
fed tax: $200
Now both districts are relatively equally off, so residents can not complain that it was THEIR representative that is failing them.. Both have services to show to their locals. Everyone is sharing in the cost burden (though some districts will get more utility than others'). The politicians win all around.
So case 3 almost alawys is favored.. So you're essentially LOOKING to add pork to ledgislation, because society is "better off" all around.. It's like the MS tax,, by having EVERYONE pay an extra $20 for the
I don't know the situation in Sweeden, but it occurs to me the following:
There are certain fixed expenses for any country no matter the size. If you publicly build a road, a court-house, etc. Then there is a minimum public cost. Thus the smaller the country, the greater the burden per-capita for such infrastructure.
In the US, Social Security and infrastructure needs are relatively small portions of the tax pie (no more than 30% combined)
The other issue is that things like social-security aren't really taxes, but social engineering. Basically a government is doing 3 things: 1) wealth-redistribution, giving poorer people a better deal than wealthy on retirement 2) forced savings (which runs at the cost of reduced GDP) which alleviates welfare on retirement 3) reducing inflation by removing the public's ability to spend on gambling institutions such as speculative financial markets and real-estate. The US, for example, requires S.S. dollars to go into US Treasuries (which of course is convinient for itself).
As for gas tax, thank green-peace. Our taxed vice is smoking.
On the one hand, you could argue that it's not fair for smaller countries, but you could always aggregate your infrastructure costs with neighboring countries. Otherwise, you're getting value for your high taxes. Surely not every dollar's worth, but definitely moreso than if you had no taxes at all (and thus no civil infrastructure).
There is no reason why a JVM couldn't release memory that isn't needed anymore to the OS. However, since memory resources are always in a state of flux with Java, the likelihood that hitting such a high water mark again will occur is high, so it generally isn't worth doing. Java only claims extra memory from the OS IF the programmer was dumb enough to hang on to too much memory.. This is a notorious problem with Java programmers.. They think it's too expensive to construct new objects, so they cache them ALL. Even the stupidest little ones like lists of data. So each of the thousands of classes each holds on to little caches of data, which point to other caches of data, bla bla bla. The end result is that the GC winds up doing 2 or three full sweeps and doesn't recover enough memory, so it asks the OS for more..
The way it's supposed to work is that object allocation is embarrasingly fast (compared to most other architectures), so you allocate what you need and release it as soon as you don't need it. You have thousands of very fast mini-garbage collections this way, and never get above a 32Meg foot-print. (Yes, I know 32Meg is large compared to say perl). I often run full web applications that never grow to more than 32Meg.. However, when doing full caching DB's I grow to over a gig in size. But the performance is worth it in that case. It's only when some thread-locals dont' release their wears that I find memory "leaks", which aren't really leaks, since the next time that thread is re-used, that memory will be freed and thereby recovered.. In the mean time slow (2 second) full GCs need to be performed. But it's still bad programmer design, not the VM itself. A c program written in the same way would leak just as badly, and lets not forget the propensity of c++ programs to never even free objects it's not done with.
Sure, new Matrix(A.negate().multipliedBy(C)).addedTo(D) is so much clearer than -A*C + B...
The problem is that "-", "*" become ambiguous in c++. In java you could at least say:
a.crossProd(b), or a.dotProd(b).. But if you're looking at a * b in c++ for the first time, which is it?
The ambiguity problem continues into the non numeric examples as well. Thus, w/ heavy use of over-loading, c++ becomes actually harder to understand.
It's a proportionality thing.. Often, the difficulty of programming i18n is less dramatic than the problem of actually generating all the different language sets.. Often such management means an advanced GUI (maybe not advanced, but certainly more than just raw field entries for a DB-backed widget).
i18n is generally token language-set in a 1:n relationship.. Which maps nicely to table layouts, thus I don't see any need to create i18n support in the DB itself.
If you want some degree of abstraction, java provides the ResourceBundle for which you can easily write your own DB-backed loaders. You instantiate your bundle, then pass it around to whatever device needs to render the actual text.
There is lots of i18n support in the ASP/JSP environment (I assume in PHP as well). jstl, struts, webwork have nice end-to-end support for i18n (error messages accepting tokens instead of raw text, etc).
For manipulation of static sets of text, there are generally plugins for your editors which allow you to manage a suite of bundleName_locale.properties with color highlighting, missing-term warnings, etc. I personally use the properties plugin for idea in the java environment.
Anymore, you should conciously be evaluating if anything is ever being displayed to the end-user, and organizing that material into i18n bundles of some sort. Standardizing on whatever your platform natively supports is critical because you can leverage the tonns of tools that are here and are bound to come.
Yeah, some are security consultants, white hat, etc., but I don't think these are what Ranum was primarily talking about.
But, according to my argument, it is irrelevant whether they are indeed white-hat or not.. Someone hacking into your site is always bad, if nothing else, there is an exchange of ego between the hackor and hackee.
My point is that any hack that does not involve a massive financial or legal upset is a good hack, since it adjusts the priorities of everyone involved so as to avoid a future occurance. Such a "lock-down" serves, if nothing else, as an example to the company's peers... The IT guy that has been FIGHTING to get past the beuracracy to lock down the local network now has evidence to support his claims... He is likely to get better funding, and will be able to coorse employees by removing admin rights, etc.. Something that is VERY hard to do if not already in place. This is especially true in Universities where people feel like open-access is a requirement of their job.
Finally to re-iterate, my thesis is that a slow migration of hacks is not only good, but essential for survival. Not having an ever increasily sophisticated competitor will leave you unable to cope with serious challenges ahead. It is only when the competition advances faster than you that there is a problem; rapid changes against your favor. This is how species die out, and this is how industries collapse.
Though I haven't read Wealth of Nations yet, I don't suspect that Adam smith ever said that greed was moral, but the existance of greed needs to be recognized and factored in to decision making.. Likewise, I make no claim that hacking is moral.. I do not necessarily encourage new would-be-hackers out there, but what I do say is that competition [to an otherwise altruistic and trusting society] exists as a fact of life.. Prosecute the financial obstructors, but learn to embrace the environment of natural distrust; this is the animal kingdom as it has survived for millions of years (without serious interruption).
To say hackers are evil is like saying germs, viruses, and carnivores in general are evil. By merely acting out Adam Smith's society being benifited best by each acting in his own best interests (adapted by John Nash to include societal interests for best outcome), we are keeping in step with mere nature.. A dog will forage for food, defend it's food, and kill it's food, so that it can stay alive. A rabbit will defend against other rabbits if need be (though they'll generally run away from anything else).
A patron is looking for a good deal, and will expend effort to maximize their deal, so sloopy wording on a sign on your store-front are invites to a natural onslaught of fiscal frustration. By natural, I mean there is no evil intent in people trying to keep you for your word in maintaining a good bargain (that you didn't intend).
If there is money on the street, it is conceivable that:
a) the original owner will never find it again
b) someone else will take the money
So you justify taking the money yourself.
If you are hungry, you might be inclined to take two samples at a free food-sample kiosk. It's unfair as it goes beyond the intent of "sampling" and takes away from other's (since there is usually a set amount of sample provided for the day).
In reality, those that are sheltered from such harsh survival of the fittest environments will EVENTUALLY meet with that environment.. It is impossible (short of death) to avoid it. Thus the question is not IF we will meet our challenges, but when, and how quickly will the difficulty level rise.
For those with assets we fear to loose (time,money,posessions,intellectual property, etc), it is natural for them to be saught by others. Having a public wiki is valueable advertising real-estate (or a personal repository for globally accessible content). So grafiti, being merely a primitive form of marketing, is bound to happen. Bank accounts are an obvious point of content.. If you happened to come across money on the street, you are more than likely to take it. If your ATM machine started allowing you to withdraw cash w/o deducting from your bank account, there is a better than likely chance that you'll take advantage (anonymous theft when it is considered to not overwhelmingly harm someone else - proportionate loss/gain - is often self justified). There isn't much difference from taking from that ATM machine and taking from an online bank account that you've happened by. Yes there is a greater issue of proportionality (you might be stealing from someone poorer than you), but you might think to yourself (I'm teaching them a lesson).. What-ever the cause, an otherwise moral man may find themselves tempted.. To say nothing of the mafia.
And ultimately organized crime is the tyrannasauras of our internet age. The mafia being only one form of it (unfriendly governments being an even more serious threat). The age of mafia and internet "WAR" (literally between nation-states) is only a matter of time.
So if our "evolution" through natural selection and adverse environment does not "toughen" us enough to sustain such natural phenomena, then we will die (or at least the medium will die).
So lets look again at these "evil" hackers. Many of the hackers were self-professed white-hackers, or anonymous exposers. If you are inclined to see if a WEB-INF directory or IIS-specific file-set are visible on a public site, you can either email their sys-admin who might sue you for hacking, or simply ignore you (like MS tries to do with serious security alerts so long as the general public is oblivious), or you can make it a priority for them... Deface their web site, delete lots of their database records.. Make it too expensive for them NOT to resolve the issue.
These are altruistic people. Slightly less altruistic are those that advertise themselves 3l33t hacker-names advertised here and there. As they have the fun and recognition-factor of it all (especially if they get CNN coverage).
Embrace th
In this interpretation, I go back in time, and kill my great grandfather when he was a baby. Well, all that happens is that I fork the universe into two possible histories.
My main problem with this theory is that in order to have these two "alternate universes", you'd have to have twice as much matter in the universe. A timeline and space-line should be continuous (my own conjecture), thus for an external observer (I know, impossible in general relativity), you can hold fixed x,y and time coordinates and trace through a z coordinate and see continuity. Now the z dimension can contract, warp, fold, or even loop (which facilitates the idea of more than 3 spacial dimensions as described in string-theory), but it should still be continuous.. Even pinch-off points like black-holes represent continuous regions of space.
So take the worm-hole representing a space-time-loop back into time (namely you present hyper-energetic force onto an asteroid causeing it to warp space-time, such that the other end point brings you back to your living room 20 years ago). By passing through this hypothetical tunnel, you are passing along a continuous path that brings you back in time.]
For an "alternate" reality to be born of this new path-forward, there must be a new continous line that passes through an entire universe that use to exist, but with an alternate path forward. It would be like the infinite spiral.. If traveling through space-time in our relality, you could pass around this worm-hole, never going through it, and never experiencing the changes relevant to it.. But by going through a couple thousand cubic-miles of the region OF the hole, you are passing through a micro-fine alternate universe.. You have essentially compacted an entire alternate copy of the universe into a tiny region of space.
So the energy required to generate this alternate universe would be the fraction of the total energy contained in the universe for the time-space-period it spans. In other words, the amount of hyper-energy required to create the warp is simply insane.
This isn't to say that such regions of space do not exist.. The universe may have sub-divided at it's creation, sending enormous fractions of the total space-time energy into such small regions of space producing mostly symmetric alternate realities. What I suggest is that to newly create such a region would be implausible.
Even if this isn't how a worm-hole that warps time would exist, I still conjecture that energy must be conserved in any medium that allows energetic material to pass from one point in the time-line to another (since general relativity would seem to hold symmetric time to the spacial axises). To alter history would require either a finite amount of energy to pass back and assimilate into the known history of the universe (as the master article suggests), or is simply not possible. But I can not fathom the synthesis of an infinite amount of energy required to provide a continus tunnel allowing material from the present to produce an infinite number of time-lines.
The only other possibility I can consider is the concept of wave interference on particles.. Namely the double-slit experiment.. A single particle travels as if there were an infinite number of universes, each having a particle travel a different path.. But the particle feels the interference forces as if they were all part of the same universe.. Namely the fact that there is a bean of light going through two slits COULD mean that there are an infinite number of photons going through both slits and an interference pattern emerges.. But when we can measurably demonstrate that only a single photon is being fired, the photon still acts the same... The master article talks about the difference between measured and unmeasured states.. The interference pattern occurs during an unmeasured portion of the photon's travel. In fact, measuring along the path would cause new interference patterns, thereby disrupting the final pattern, so it is impossible to
. Accessing a global or shared resource? Lock. It takes one line of code (& one to unlock). It becomes second nature.
While I generally agree with you.. It isn't as simple as just locking every time you access a global resource.. You have deadlock issues. So you either need a monitor thread or a global lock which is required to acquire a subset of locks. Or if you feel particularly daring, you can have a convention to always lock in a particular order. The problem comes when you have different modules which acquire different locks, and you can't control the order in which the modules are invoked for a particular event/task. This becomes increasingly difficult as your application grows appendages..
Your best bet is to minimize the need to access such global resources, and to have locks encapsulated and atomic, and more importantly, not farming out to acquire other locks. But this isn't always easy to do. Minimizing global resource access implies making more use of local variables.. Passing instances up the call stack instead of having modules search for global resources.
In servlet land, you have a synchronized application context Map and a per-user synchronized Session Map. Then you have a non-synchronized request Map passed up the call tree. Here you are explicitly defining the global-ness of operations. Moreoever servlet designers recognize this explicit scope separation and place shared resources at the appropriate level.. Anything being placed into the Request scope need no synchronization since it's single-threaded access.. When you place something in the other scopes, you are making a conscious statement that you must be mindful of MT-safety.
Essentially what you propose is to write your code forcibly singlethreaded, because that's what that style of writing boils down to.
Not at all.. In another post, I suggested the following.
Collection input = repository.findE(...);
Collection output = MTCollectionUtils.map(input, myMapper).
The above essentially works like:
for (E e : input)
{
F f = operation on e;
output.add(f);
}
But allows the generic application of farming out to worker threads. Essentially explicit SIMD operation.
This could be the actualy hibernate re-materialization, or reporting or aggregation.
The point here is that this is a simple example of a procedure-local variables being used in an MT way.
As for concurrent access to "entities"; Objects which have uniquely identifiable properties (as opposed to value-objects like a date), systems like java provide a relatively easy way to access them.. Namely object-synchronization. This is most perfectly exemplified in Collection/Map synchornization. Maps are repeatedly used in MT environments in java, yet they are very reliable. (Rather when using Hashtable or the Collections.synchronizedMap() wrapper).
Concurrently entity modification in today's environment can easily utilize "optimistic" locking, whereby the entire transaction can be marked as repeatable.. You initiate a repeatable transaction. You load all relavent entities from a shared environment (such as hibernate or directly from the DB), then you update a version of some sort (like a modified-timestamp on each entity). Then you go to push the objects back to the share environment.. But in doing so, you check to make sure you have the latest version.. Otherwise you throw a concurrent modification exception... The transaction is rolled back as if nothing has happened.. At which point you minorly annoy the user by providing an error screen, or you intelligently catch the exception and restart the entire process.
The above obviously is most suitable when there are low probabilities of conflict, as this provides the highest concurrent execution performance. Not to mention is the least intrusive form of concurrent execution.. The only thing it requires is adherence to a transactional model.. But mysql's INNODB or postgres provide this nicely without an EJB framework.
When overlapping modification is too likely, then you can utilize explicit blocking locking methods. Often such locking occurs at the DB access point, not the application level. So assuming that you have an adaquate DB abstraction layer such as OJB or hibernate, then you're safe even here in terms of writing bug-free code.
When a mixture of performance and concurrent modification of the same items is paramount, then you'll probably have to resort to object locking (or even finer grained locking).. But I would venture to say that this is rarely needed.. As the benifits of writing the slower but bug-freer code and purchasing more expensive hardware is not to be under-estimated.
Of course there are programming models that don't exactly conform to a web servlet / mainframe (SIMD) work-flow. So your mileage may vary.
It's more than just setting up an environment as immutable.. What the parent post was talking about what procedure-local variable usage.. Statics are obviously an obsticle to MT, but so are field-variables. If a class is instantiated w/in the method and used by the method and invoked methods, then you are inherently thread-safe. For the purposes of sharing objects between method invocations, you have immutable singleton "verbs", and noun immutable "Value Objects". What's left are noun mutable "Entity" Objects. By separating design into verbs, value-objects and entities, then you know where you have to focus your synchronization efforts.. Mind you it takes WORK to maintain the value-objectness of non entities.
s.getAttr();
..) ;
The typical process for "verbs" is to first call set-attributes, init, then execute, then get-attributes. The problem is that if the verb object is passed around there is a constraint of single-threadedness.
If on the other hand, the verb is written as a singleton which returns value-objects, then there is no dangerous of having multiple threads acceess the verb. But it often requires creating multiple transient objects.. For example:
State s = myFactoryObj.setupVerb(attr-list);
s.execute();
It's often easier to perform:
myEntityObj.setAttr(..);
myEntityObj.doVerb(..);
myEntityObj.getAttr(
But as the name implies, being an entity object, it is likely to be used in multiple contexts.
By separating verbs from shared mutable entities, your code is more thread-safe, and as a bonus, more unit-wise testable.. You can trust that the entire "login process" works independently of a user object.
The book domain-driven-design is an excellent read on such architectural separations.
Such techniques are applicatable to even non OO environments. Side-effects are evil and stand in the way of many things. Making state modification explicit, you can more correctly identify synchornization points and more importantly reduce the number of them.
Not all apps parallelize well, and parallelizing apps is difficult in any case, so Si Lasers are not likely to yield a big breakthrough in performance.
Ok, imagine this. You put on one machine, a database, a java server on another machine running with 500 threads, and an apache server on still a 3'rd machine with mixed mode multi-threaded, multi-processed operation (to facilitate a hetero-genous php, perl, static-file, java proxied, https'd environment). This is a simple e-commerse web site.
What is the advantage of putting this on one machine? Reduced latency. Tremendously reduced administration. Now what about fail-over? Ok, you build two identical machines and clustering techniques. Now you have two mirror machines (now tremendously easy to administer).. You give up some performance in clustering, but you have no single point of failure. Compare that to the usual standard of having a single database machine with a clustered web environment.
Having one (two) machines means reduces UPS, heat, administrative, power, etc requirements. This is what blade servers are all about, but even better.
Now it is entirely likely that multi CPU machine will be slower in through-put than two independent but chained computers (web + application + DB machines). But the latency is certainly traded in this situation. To compensate, it it likely that having a few unique instruction sets (httpd, JVM, DB-code) will live in common sufficiently larege cache enviornments, thus the competition for resources may be minimal enough to not be negative.. Moreoever, UNIX sockes are significantly faster than TCP sockets even on a local machine, to say nothing of the latency rich network connection.
Ok, so what about non mainframe style "small work, LOTS of jobs" operation? Take a work-station, where you have 20 major applications sitting mostly idle, but taking up enormous resources. The work-station process involves one or two major applications being active at a time, but constantly switching between them. If there is a single CPU, then there is a high probability that one will be performing background activity.. Especially for things like compilation, where you'll switch to another task while waiting. Even if a particular task has a max number of practical CPU's that it can fully utilize, having a "spare" idle CPU is GREAT for task switching.. There is no delay in contexting IN one application while either contexting out the old app, or simply keeping it active until it's finished it's current task. Granted some switching occurs 60 times a second on single CPUs, how much of that work can be avoided by not having to even switch out a task for a 32CPU environment. Taking two apps running concurrently on a single CPU always runs slower than running them sequentially. So such environments, even if only giving a small percentage gain on average can still provide a happier user experience.
Finally, take the end-user desktop environment. As other's have pointed out, innovation has typically lagged proliferation. 3D games didn't take off until we have 5 different standards in hardware video-acceleration. Eventually voodoo stood out from the pack and dramatically increased the bravery of game makers, since there was a legitimate market to show-case incredible features , while backing down to the lesser capable graphics standards.. Eventually nVidia gave validation to the DirectX standard, and after quality issues were practical to overcome, that market exploded; building upon itself; faster features, more supported software..
Today multi-threading on the desktop exists PURELY for user-responsiveness, not for performance, since 99% of all desktops are single-threaded. We now expect that the menu for an application should continue operating even when it's doing some intensive task. In fact, we feel upset with software that does not adaquately thread their UI. This is because that's all that's really possible with single-CPU hardware.
But if you take mainframe/server
The problem I have with your synopsis of Mac penetration are:
1) Legacy Standards... VHS beat Beta because of proliferation, NOT because of quality. People are use to windows, getting them to switch to ANYTHING else is a major obsticle.. Now Loghorn might be the key to screwing everything up for MS.. It's like IBM going to MCI architecture when everyone else went EISA or VLB. IF MS forces people to switch to a totally different UI experience, then something like Apple which will have a proven user experience by that time will have a chance.. But this assumes loghorn is a radical change.
2) User types.. There are 4 user types that I can think of. Gamers, developers, business, and media. Developers are specific to their platform of choice, so making one OS nicer than another is not likely to sway them in any direction (plus they're a minority (and Java support sucks on Macs)). Gamers have DirectX or consoles. The Mac doesn't have a prolific user base.. And you suggested yourself that people will need to dual boot.. Games will definitely need native execution, since MS isn't about to let other platforms utilize DirectX. More and more people are switching to consoles now that they've finally woken up and found how to use the state-of-the-art in graphics.. Nintendo's and PSP's were a joke to me when I'd read about the latest voodoo, then later Geforce cards.. Now the consoles are getting seriously competative, and they're easier to deal with as a gaming system (not that I'll ever use a console). Media doesn't have a strict advantage on any one platform. You can play/rip/encode a DVD just about anywhere.. This is really a choice of the more popular softwer vendors what platforms they want to support. Mac and Windows definintely have an edge over Linux in multi-media integration into the OS, but I don't consider this a good thing, especially where DRM is concerned... Frankly, I believe MS has the edge on DRM, since if you are a media company, which standard are you going to support, 3% market share with potential for growth, or the yellow book? Finally businesses.. They need perfect support for the latest office documents out there.. They need conferencing do-dads.. They need stability.. They need exchange / lotus-notes what-ever. MS works hard to lock businesses into their product lines. Apple isn't 100% compatible today, and I suspect MS will innovate ways of keeping them always in that "well, it 'mostly works'" category.
So there are no demographics which make me say.. Yes! there's where they're going to find their niche and expand from there.
That's why IBM is way behind Intel when it comes to device size.
I don't know, I'm just asking, but this doesn't make sense.. Why would IBM be having trouble shrinking the die due to SOI being the problem IF AMD is using the exact same technology and getting 90nm + lower power? What does AMD know that IBM doesn't?
My guess is that it's a function of finance (sales volume), not technology. It probably costs AMD boat-loads to do the SOI, but they have enough volume to justify it. How many chips of a given architecture does IBM make? Is it cheaper for them to put more transistors and better cooling on a Power chip than to shrink the die and up the clock? Probably. The only thing IBM makes that needs better price/perf are the PPCs.. So they obviously get the shaft because of too low a market volume.