The SSSCA is a bill produced and supported out of valid fear; Fear of the volitility of GDP, and more recently, fear of illicit communications.
The former fear has been historically supported by American policy; to legally segregate markets so as to make local economies profitable. This is most evident in the former banking industry. Such ledgislation can be viewed as a "socially constructing" tax, which sits alongside tarrifs, tobacco taxes and income verse sales taxes. In all cases the ledgislatures balance the need against the cost.
That is to have a balance that is both "of the will of the people" and more importantly moral. Taxing a fixed sum from all citizens is arguably immoral (as could be seen in the dark ages). Backing a known corrupt monopoly is obviously immoral (which is recognized by semi-recent American ledgislatures through the inactment of anti-trust law).
The tricky part is to identify the will of the people in such a way that avoids mob-rule. Obviously anything that affects a majority of people negatively is subject to mob-rule. Exclusively taxing the rich, for example pushes too far into the realm of immorality and thus over-rules the opinion of the masses.
The SSSCA indirectly argues that intellectual property is soverign and should be protected by the government to whatever extent is possible (just as human life is to be protected). Obviously it is difficult to distribute IP and garuntee control, since once out of the protective sphere of it's owner, it can be manipulated, duplicated and redistributed (potentially for profit).
The DMCA legally protects digital copyright, but does nothing to prevent indirect circumvention (say via converting digital to analog and back again (unprotected)). The SSSCA closes this loop-hole by proposing to "regulate" the digital world such that no hardware or software can exist that "pollutes" the IP soverignty.
In the above, I have argued the pro point of view. Below I will expose the weaknesses and moral imperatives, and offer compromises.
First, the role of the government is virtually sacrad; To enforce a regional monopoly on "force". Any abuse of such a monopoly is obviously immoral. Allowing an individual to wrestle control of that monopoly for their own purposes is immoral. Upholding loopholes that allow individuals to "legally" attain control of that monopoly is immoral. An example of such legal perversion is law that can indefinately inprison citizens for non-immoral behavior. Here force is blindly applied to detain individuals that violate written law, thus anyone that calls apon a poorly written law which victimizes a moral citizen is abusing a loop-hole in the system. Such a system MUST not persist, but instead be mended. What's more it is the duty of ledgislatures to preempt such loopholes before they finalize laws.
While the primary goal of a government involves applying force to constrict a citizens use of force (or similarly to withstand foreign aggression), a physical ramification of this is social engineering via regulation. Due to a sence of "fairness" and or public safety, enforcers (police / military) have been instructed to apply the written regulations of conduct in addition to their force-monopoly. But the bounds to the American limit of efficability of such regulation is the US Constitution. However even this "minimalist" view of moral judgement is fallible (as can be seen with the introduction of failed prohibition or even slavery laws).
Thus in order to preempt moral imperatives that take supreme-court ruling and eventual lengthy legislation to amend the constitution, it is most arguably best to apply moral scurtiny to the laws before they ever achieve legal status. A ledislature MUST, therefore respect the sacred nature of his office, lest [s]he be shunned in history.
Regulating banking maintains local economies at the expense of consumers. The cost is moderate and monetary and thus moral. Regulating taxes likewise is monetarily based but runs the risk of starvation of subsets of citizens; the worst case involves death and thus achieves high scrutiny and modern governments tend to fair well in this camp. Public safety regulation prevents producers from trading safety for profit (making an industry more fair by requiring all producers utilize the more expensive safety measures). In the short term, the negatives are higher cost-of-goods to a consumer as well as the abuse of force in telling citizens how to conduct themselves. In the longer run, economies of scale diminish the extra cost of goods, while the "forced conduct" prevents indiscriminant loss of life. Hense the government is "saving us from ourselves", and more importantly reducing an automobile's role as a weapon. Next to public safety are physical property laws; those which "regulate our conduct" such that we do not deprive one another of our physical property. Since food and shelter are life-sustaining forms of property, generalizing governmental enforcement of property ownership is valuable; e.g. a government is justified in protecting the ownership of a CD just as much as protecting the milk that a baby carries.
At the bottom tear of "moral" laws are the corporate laws. These are "rules of conduct" laws which have nothing to do with maintaing a government, nor saving lives, but instead deal with the secondary effects of economies. An economy is efficient and can flurish when various rules are applied. This is where tax-games can be played to encourage conduct or physical regulation (such as hours of stock-trading, or margin-limits) are applied to avoid economic disasters. In here too is the concept of patent law and copyright, which introduces the concept of intellectual property. The social cost of a patent is to restrict our actions such that we don't act in a manner that mimics one expressly copyrighted or patented. While the intended effects are purely economical (and thus not morally justifiable), the unintended effects are to allow individuals to call apon the force of government to control citizens. Whether the citizen was indeed attempting to profit on the IP-holder is not relavent to the cost. A mother that sings happy birthday to her child is in violation of IP and thus is legally vulnerable to incarseration (if the IP-holder was sufficiently immoral). This is a dire perversion and loophole that allows a single individual to wrestle control of the government's monopoly.
One possible remidy would be to apply the condition of profitability to all copyrights and patentents. The IP-holder attains legal rights to all revenue based on that IP. This does not stiffle do-gooders who utilize the IP for charity or public-good will. Arguably, even the GPL would be in violation of such IP, since it allows for the distribution charging (which can be a disguised form of profiteering). Such a form of modifiation of existing legislation arguably needs to avoid loop-holes (mostly involving proofiteering, since the original intention was robust GDP).
The DMCA assumes copyright is a fundamental right of IP (as are life and liberty. This is, in my opinion, a perversion of the "right of happiness"). It assumes that copyright is so important that not only can you _prevent_ redistribution profiteering, but you can dictate how and when the IP can be accessed, which is a gross extension of the concept of governmentally enforced social control. It is more than simply restricting when you make your IP available to subsets of the public (which is perfectly acceptible, as a performer could choose when and where they perform), but to say that you can only hear it a certain way, to talk about it a certain way (perverted slander laws that defend corporations, even when they are blame-worthy), or to witness the IP via a monopolized medium. The DVD, the up-and-comming HDTV anti-time-delayed-viewing, the encrypted pdf format and others are examples of such IP who's viewing capability is monopolized. That companies are capable of
restricting viewership is not, in and of itself, immoral. A theater, for example, can monopolize on refreshments due to it's locality; they can deny admission if need be. But to even suggest that when you watch a home-movie, you can only eat a certain name-brand refreshment is totalitarian plain-and simple. But this is exactly what the DMCA does. It produces an artificial locality, and an artificial scarcity (via fear of governmental retaliation, wielded by the very companies that profit from your patronage). This is a direct conflict of interest; a man may provide a quasi-desirable service while wielding a governmental whip which says that you have to partake from it. So long as there are alternatives, there is the prospect of "choosing not to partake" in the bait that locks you into the negative aspects of a service, but there is a natural effect which brings like businesses into the same immoral tactics due to lack of profitability.
It is perfectly valid for manufacturers of DVD to force commercials and regional encoding, and restrict viewers to set-top boxes or highly trusted windows machines. But when ALL major IP owners in the media field shift to DVD, a consumer will have the choice of boycotting ALL cinematic/ auditory IP short of live concerts. VHS and casset will be phased out and your living room will be an extension of the controlled theater. Even this isn't a "really" a problem because we still have choice. But what the DMCA allows is the removal of other forms of media. It is illegal to convert a DVD to VHS or even to record the sound-track to cassette, since this would circumvent the encryption. Granted the intent of DMCA is to prevent the perfect replication of data which can then potentially be advantaged by profiteers (or more recently, in the production of a public "library of music" which has the potential for reducing the demand for highly priced official distribution points). But the DMCA is a loaded weapon which can be fired at existing acceptible practices (such as making a cassette for use in the car). The hidden answer is that corporations never wanted us to have the ability to freely utilized purchased licenses of IP; they simply lacked the ability to enforce it.
The real long-term question is as internet bandwith expands and the quality of compression software approaches that of the original masters, how can a media-based industry survive (and thereby uphold GDP which benifits all citizens indirectly). My answer is that a corporation can and should "squeeze every dollar" it can from an industry, so long as the dirty work is not enforced by the government; such an activity is called extortion, and such an organization is called the mafia. Times change and it is up to business-heads to determine new manners of supply and demand to survive. It is not moral to enforce laws that restrict citizens rights so that aging task-masters can retain their old positions. This century has seen the horrors of worker-layoffs.. It's time that the "task-masters" experience their share of layoffs. If NSync can't retain their multi-million-dollar industry, then let them suffer with more concerts and new forms of high-bandwidth (multi-channel) audio-CDs priced lower to catch the interest of more of the waning support-crowd.
The main point is that back-door operations, and incogneto sharing of IP will always exist, and times / public interest will most definately change. This is merely a last-ditch effort to safe stock-face in the short-run and punish the most visible; the naieve citizens, or the high-profile educators that teach on the topic of encryption, or the security experts who crack systems to improve apon their security. Those that are actually helping society will be hurt the most. The industry will still slump and illicit profiteering will still proliferate (since one additional law isn't going to make them any more legally vulnerable).
Lastly, the SSSCA can be shown to be built apon a series poor-ledgislation, which most definately holds to the spirit of it's predecessors, but in a most perverse way misses the whole point. The regulation that places requirements on specifications within hardware limits the diversity of that hardware, first-and foremost. The ancenstral point of the SSSCA is to foster growth, and innovation, but at step one, it grinds innovation to a screetching halt. How can an "embedded device" in your car be required to provide protected encrypted streaming audio? Surely the computers within the cars of the future will have the ability to do some tasks currently performed on laptops or PCs. But at what point will they be held liable? The answer is whenever it is most convinient for the corporate task-masters. Manufacturers will live in fear of backlash so long as they exist on that fine-line. Many will simply forego advancement, and innovation will be halted.
The government has made several failed attempts to regulate based on flawed information. Regional governments have required gas addatives that were later learned to be harmful to the environment. The best way to avoid such forced lemming rank-and-filing of citizens (doomed to fall over a cliff) is to identify the problem and require "A" solution, but then allow innovation to fullfill that requirement. To suggest that "government knows" best, and that "this is for our own good" is as blind as Hitler. (If the analogy is offensive, then ask yourself why?)
The analogy of a required built-in security equipment to a catalitic converter works when you consider that both help prevent environmental pollution. In this case, the digital pollution of economicly viable media. Both cars and computers are perfectly competative goods (within their genre) which fulfill very similar purposes and are essentially black-boxes to their users. But note that a car can be electric, and thus forego the need for a catalitic converter out of "common sence". But there is no "common sence" when it comes to computer science research, since it is a fronteer who's tip has barely been identified. This would be like requiring that "all vehicals" contain electrically shielded engine cranks 50 years after the electric starter had already been invented. There exists hardware and software alike that is incompatible with such low-level drivers that provide controlled access to media-devices. This level of information is completely incomprehensible to most ledgislatures (since it's inside the box), but it as the heart of digital science. To provide such a vague law allows individuals to wield the power of the government to stiffle their competition in this cut-throat science. The ability to provide an increasingly out-dated media in a governmentally controlled fashion comes at the cost of intellectuals who will no longer have the ability to freely experiment in the digital world. As competition for innovation comes to a half, so will the desire for innovation, and all that will be left are the greedy hands of software and hardware monopolies which will regularly fly in the face of ledgislatures as an embarrasement of capitalism. Capitalism only works [as a moral entity] in a governmentally minimalistic society.
IP laws such as copyright, trademarks and patents have values which outweigh most of the costs, so long as the loop-holes described above are eventually mended. The DMCA is completely redudant with laws that physically address profiteering, and only opens a vagueness which allows corporate abuse of governmental force. The SSSCA, if enacted will backfire on both GDP and legislatures and will eventually be repealed. It is best to preempt a human tragedy before letting court appeal delays take it's toll.
Also that means that if they _suspect_ you of selling/using drugs, they can freeze your finances completely. It gives _way_ too much control to somebody else, based on politics, purchasing habits, etc... It makes my skin crawl.
How's that different than credit / checking today? There will always be some sort of "tradeable" currency. Even if it's computers / cars, etc. They can't force you to use a certain type of transaction when you're really trying to be discrete.
What I see is the same as the gold-standard, which was "credit redeemable in cash" stamped on the cards. Coins will probably be phased out (if inflation hits again and no soda costs less than a dollar, which you already see in high profile areas with vending machines), but dollar bills should be available for more than your life-time.
In the worst case, the mafia has always been good at money laundring. I'm sure there'll be pawn-shops that act as fronts (for good ole' fashioned trade).
Then valuable metal became scarce, so we came to use pieces of paper that represented metal stored in a fort somewhere.
I'm not sure that this was the case. It wasn't that the valued goods became scarce, so much as it was impossible to wield $100,000 around in your pocket. Within a given country, the government provided security as for the value of coin, and in the worst case, that coin had some value if melted down. Later as coutries trusted this sort of exchange fiat money (or completely worthless) was used.
This required the trust that you could get something valuable back if you wanted (say to trade internationally).
In the US at least, it was eventually determined that the economy need to grow and shring, and that fixing equity on stocked goods was innefficient. If we had inflation, for example, we would have liked to have introduced new cash into society to compensate since the price of gold (the US's former standard) didn't directly vary with the rate of inflation.
Things were still safe because you could regulate the printing/coining of fiat monies. But then checking became very popular. Now you had the concept of float. One bank would honor a check (and allow accumulation of credit/cash) before the debited bank could deduct.
Later we have the concept of equity-based loans. I percieve that your good is valued such that I'll lend you most of the money for it. You take that money and spend it (via checks), but more goods and take loans out on them..
All in all, checkable money develops a velocity (the rate at which the same virtual or physical dollar is spent per year) such that our net assets are multiple times the physical printed fiat dollars total value.
In a booming economy, that multiplier increases. The problem is that that rate of boom has to be maintained or there will be a dramatic credit crunch. A recession after a boom is devistating because trillions of dollars can up and dissapear (after all checks are registered).
This would have happened even with a gold standard due to virtual assets and value.
This is something that some postmodernist thinkers saw coming a long time ago. It has to do with the continual separation from reality.
The issue has always been one of efficiency. Yes we're more at risk now that a single number can render our bank-account empty. But we have a much greater ability to refill that bank-account than we did when someone with TNT could "blow the safe" and bring you back to square one. You can be insured, bring out new mortages so you don't starve, and most importantly be paid a heck of a lot more than days of old due to incredible industry efficiencies.
Back in the early days, everyone lauded Java as a means to add multimedia content to the web.
That may have been its most obvious selling point, but that was far from it's only use. It was a cross-platform method of client/server programming. You could develop sophisticated applications with incredible performance characteristics (compared to simple DHTML) that you could distribute to a heterogenious intranet (Solaris/Linux/win32).
Eventually DHTML got more sophisticated, and generating graphics for HTML was an order of magnitude simpler than for Java. Thanks to more reliable javascript, you can also produce many of the dynamic operations that previously required Java. (procedural drop-downs, tabbed pages, refreshing, etc).
I wonder if you can compile the Linux kernel with it (minus assembly of course). That might be interesting, particularly for P4. Linux could get an instant speed boost.
Possibly, but you'd have to ship the compiler with the dist since you would be hard pressed to link dissimilar code together dynamically. I'm no expert on this, but I do remember woes with using a sun compiler on Solarix x86 with the gcc compiler. This failed miserably with Apache / mod_perl and with simple perl + CPAN libraries. We had to go with the Solaris comipler all the way, which was a royal pain, let me tell you.
Unless someone has some info to the contrary, you'd have to forgoe most any precompiled linux binaries, which will definately get into your hair, as I've definately found.
And such a radical switch in compiler might expose flaws in the code. Definately a worthwhile excercise if nothing else.
I'm not completely sure, but doesn't gcc extend C with various types of proprietary compiler attributes? I believe it's possible that the configuration stage can nullify them, and it's been a while since I've looked through Linux source, but I do remember those attributes hanging around.
Still, I'm sure it's possible, and I'd be curious to learn of anyone's success..
On the massochistic side, has anyone compiled Linux with a MicroSoft comipler?:)
The bottleneck is likely to be the connection to the monitor. Unless this is an all digital monitor (which I guess would be possible), then you still have to pass an analog signal which is limited by the quality of the cable (bandwidth) and the performance of the video card.
even if it's digital, there's still the issue of feeding all that info. You can't just have one pin per pixel, so you'll need to scan by some means. I thought that there were several competing techniques to accomplish this with LCD's though I'm not too familiar with what is mostly used.
Point being that there's no garuntee that it has any faster refresh rates than traditional CRTs, nor do we know if there will be as much fading / trailing as in some older laptops.
On power consumption:
Talking out my ass here, but it seems to reason that even though the distance between cathode and anode is smaller (requiring less voltage), there is a wider area from which to draw electrons so the current might even be higher. Also, at first glance, it seems that there are more segments that require power due to a larger number of beams. There's probably some inefficiency there too.
On quality:
Does anybody know how well this would hold out against spot imperfections? One problem I see with LCD is that you can have spot imperfections that accumulate over time. With CRT you get general distortion. I tend to prefer the distortion because it has smooth transitions that the human eye adjusts to (I buy used monitors with warps and spots at a great discount). But a dot stands out and is an irritation (high contrast). It seems to me that the flat-CRT will be just as suseptible to local-imperfections as LCD due to localized damage to anodes.
Conclusions:
Assuming the cost is greater than CRT, then the only thing you get from these monitors is space savings.. If you don't absolutely need it, it probably won't be worth it. I like my used $260 21" Hitachi just fine.
I can't really see BSD / Linux, etc being used by the general non-CS public in an educational setting. Sure it might be good for educational / managerial utilities (databases, calanders, to-do, etc). Even using it as a simple web-browser in the class-room / library (on slower machines). The problem is that it still doesn't cut the mustard for office apps. I'd venture to say that 99% of students don't have a Linux-running-Office supporting OS environment at home, and administering this would be a nightmare. Most people are going to use MS Word these days, and that's what you'll need sitting somewhere on school-grounds.
It might be possible to reduce the number of Windows machines through the infiltration of Linux, but then you'd have to have UNIX administrators. Sure the acting IS staff (typically a single guy) could learn it, but can a given school justify sending these poor-overworked people to UNIX training when they're only going to provide marginal utility for the school? I discourage schools from out-sourcing such administration to student groups for several reasons.
tcsh:
set prompt = "`echo $HOST |cut -c1-2`:\\%~>"
# "co:\full-path> " for computers starting with co
bash:
PS1="[\u@\h \W]\$"
-Michael
Re:But did Kubrick write the meta-science?
on
Review: A.I.
·
· Score: 2
She's very uncurious about why
there are no other members of the family around, for instance. As I've thought about it,
though, the reason for this is that David's Mommy is having a dream.
I'm not really going to try and explain such an open-ended cop-out, except to say that most recounted experiences with "time-travel" or more simply temporal e.s.p. leave one in a sence of a dream-state.
From what I got, everything in the universe leaves it's imprint on the analog universe, much like ripples in a zero-resistance ocean, or what-ever analogy floats your boat. They speculate that the physical piece of matter (possibly the complex DNA strand, but also possibly the matter itself) is like a finger-print that can be used to searched in the cosmic ocean to reach-back for the rest of it's constituent parts. It does suggests that time-travel isn't possible (or they wouldn't really need to ecscavate, nor would it be a problem to bring her here).
I don't quite know if they're mearly finding her personality / her essence, or if they're projecting esp to or fro.
As a good master, you don't explain the details, but try and find some high-level analogy that describes the functional parameters. Details can be depressing. Something Lucas should be-retaut.
-Michael
Re:But did Kubrick write the meta-science?
on
Review: A.I.
·
· Score: 2
haha.. laughable. We're hard-wired for sexual desire, for fear of danger, for blinking when things approach the eye, or for mothering rage. So why don't we react like a programmed interrupt controller when certain conditions arise? Because the main point of our nuerons is that they're programmable. We have both actuators and inhibitors. They both battle over the initiation of events. Yes there's a hard-wiring to put into motion a response for a given perceived action. BUT, we also have the ability to surpress those hard-wired reactions.
You could go to great lengths to encode the reward / punishment / reaction system within a neural net, but the fundamental nature of neurons is that they can be super-ceeded (and the super-ceedings can become the new nature, only later to be super-ceeded by something else). A tough man learns to not blink, and to restrain his anger. Most men learn to supress their sexual urges. Most people learn to control their need to excrete.
The only way they could do it would be to have an external response mechanism that doesn't allow over-riding (which would defeat most of the point of being alive and having an adaptive neural-net).
Obviously they do this, because of the activation code; that's something that is rather important to not 'over-ride'. But notice how little of that sort of activity is actually used.
-Michael
Re:But did Kubrick write the meta-science?
on
Review: A.I.
·
· Score: 2
maintanence? He definately has air-passages and vocal devices. Even if the food didn't go to his stomach, certainly these devices would be vulnerable.
I say the Three Laws Of Robotics cannot be implemented as written. Period.
From a fictional point of view, it's not out of bounds obviously.
I can't believe nobody has introduced RoboCop into the discussion (at least the first one). Here was an attempt to integrate the human psyche with computerized control (his prime directives). The parallel is in having human emotion while having a glass ceiling. They did a half decent job in that movie exploring the complications involved. I don't believe that this movie really wanted to explore these complexities. As justification, the makers were "trying so hard to see if they could, that they never stopped to see if they should". In fact, all the emotional trauma that was caused was ultimately encouraged.
Thus, in agreement with you and in start contrast to the previous poster, asimov's laws had no place in this movie.
Get a PDA. Then the difficulty is bumped up a notch.. First they have to get access to your person.. Then they have to crack the PW-manager password. Then they have to figure out which password goes with which site (hopefully you didn't put all that info together).
Only trick is if you lose your PDA. So maybe you should have a tape-archive with the info periodically. and put that into a safe where you have the only key.
I disagree. This is definately the case with a 486 to Athlon comparison, but we're already taking into account the architectural differences (stages / pipe, etc). Part of the analysis is to monitor efficiency. This is especially true with the Pentium4 / Athlon debate since we can get 1.4GHZ Athlons.. The question is whether to purchase a 1.4GHZ Pentium 4 at significantly higher cost; to say nothing of the added cost of a 1.7GHZ setup.
The difference is more dramatic between the P5-4 and P5-3, since you max out at about 1GHZ for the P5-3, and so I'd be inclined to believe you. The Athlon, however is not yet out of steam for its current. If it can best the P5-4 in 50% of the categories (including legacy apps.. e.g. modern ones), then the value of the P5-4 is limited, even if it can produce top-notch synthetic scores.
The point is that it is not ignorant to normalize, so long as you look at the periferal factors. It's like having taking the average, but also taking the standard deviation. You do find useful information from such numbers.
On a side note, I always thought that the P-III only had 3 pipelines, one of which could execute any micro-op, and the other two of which could only execute the simpler micro-ops.
Believe you're thinking of the number it can "issue", which is separate than the number of [semi-]independent pipes. In the PPro, some instructions (like divide) would lock other pipes or stages within it's own pipe. Issuing instructions is expensive, so it's generally accepted that you issue less than the number of pipes, but as the P4/Athlon have significantly more pipes than their predecessors, they have augmented the number of issued instructinos by 1 or so.
Not sure that I'm reading you correcty. My initial impression is that SMT / CMP would hurt cache hits. If you had an app that was single threaded, then obviously SMT won't help you, but CMP would have you compete for cache space. If you had a multi-threaded app, then yes their code-cache would most likely have less thrashing, but their data stands a good chance of competing for the same space.. In single threaded operation, your cache can afford to risk having 2, 4, 16, etc memory locations overlap on a cache line, since it's not too statistically likely that you'll thrash.. But if you have SMT, then various types of applications that require large data-sets (such as text-processing web servers), enhance the chances for accessing conflicting memory regions. Even with more expensive cache architectures, the likely-hood of cache-conflicts is still higher with SMT.
My understanding of the proposed SMT on x86 is that you simply switch to another thread when there's a memory stall. I think SPARCs have done that for a while... What I believe you're referring to is the reduction in the number of times you have to context switch and thereby flush your cache. Though it's true that having fewer distinct processes (even LWP ones) requires fewer context switches, I believe that you are not given a time-delta extension simply because you have 2 or more threads associated with a process for an SMT core. Thus, I believe the time-delta is still the same for all processes (minus HW interrupts), and the number of cache flushes per second is the same. Hence, little realized benifit.
Just for completeness, what I think you do get is fewer memory stalls within your time-delta. Additionally, if each thread is stalling, then you at least have multiple concurrent memory requests, which I believe does suite RDRAM well. You could achieve a similar situation by having multiple independant banks of SDRAM (like nVida's GeForce 3).
In summary, if anything, cache is the weak link towards multi-core / multi-threading.
yes and no.. how do you categorize a new chip? The AMD K5 / K6 were roughly inline with the P5, but they were separately designed.. It of course comes down to marketing. BUT, what you can look at is the generation of the design. Pentium introduced (for x86's anyway) relatively deep pipes and multi-instructions. The next generation was OOE. You may or may not be able to categorize the Pentium4 as a new generation based on it's double-pumped integer. I think that all the other aspects of the P4 are simply augmentations or incorporations of nifty ideas (like caching the decoded ops, which I believe AMD did a while ago).
I use to call the IA-64 the P7 just so that may lay-friends could know what I was talking about. It's VLIW / speculative execution could probably be considered a new generation. But in reality it's a completely separate product with hardly any ability to compare to the x86 line.
I think, however, that I'd recognize SMT / CMP as a next generation label.
This is an interesting take, but there are other considerations.. First, if you read the Findings of Fact against MS, you can easily be led to believe that MS's real purpose was to lock java to the MS platform by polluting it was ms-only extensions that proliferate on the net. This practice showed up too often in other areas to ignore its likelihood.
Next, why would MS want a write-once run anywhere development environment for themselves. They're not about to build their drivers and win32 API in Java, and any apps that they build on top of them is pure C++, so all it would take is a simple recompile for the different platforms.
When Java came out, I don't believe that Alpha-NT was that popular, and SGI-NT was being dropped (not certain about the timing, but it seems about right).
I agree with you about win 9x being stepping stones, but I don't think cross-platform was a big focus for NT. Yeah they have the Hardware abstraction layer, but I don't know that this wasn't more for stability and protected code than for true platform independance. Thought it was really just a carry-over from VMS.
I think Java is a fine education language for the following reasons.
It is cleaner than C++.
There is more exteraneous noise in Java than in several other OO-capable languages (C++ excluded). This extra required syntax and the REQUIREMENT to program in OO makes java as a first or even high-school language difficult.
If you take someone straight out of BASIC or Pascal and throw them into Java, you might have some obsticles - They're learning too much all at once. In High school, the added complexity probably isn't warranted (we want to spark interest in our youth), and in College, people that transition majors may need more hand-holding.
It is widely and freely avalaible.
Not really true nor a factor. For a large number of institutions (primarily High School), Windows is the development platform. It's harder to get easy-to-use free Java tools for Windows. Yes, you can find them, but it's not an obvious choice. The second part of my rebuttle suggests that an institution spends the money on the right tools (within their priorities for a given carriculum). We used Apple IIe's with Apple Pascal. It was a purchased license, but the whole system cost was marginal.
To counter this, Java requires expensive hardware (namely, no donated 386 / 486 is going to be happy in a [education] development environment), though these are probably what you'll find in poorer schools.. Which has to facilitate advanced students taking AP courses...
It is being used widely in the industry, and I think educational institutions have a responsibility to realease students with marketable skills.
Then they'd all teach Visual Basic as a first language. Since you're garunteed some sort of employment. It's almost impossible to avoid VB these days.
It's not strictly true that x86's are limited to 32bits of addressable memory. With VM and segment selectors (which I personally diplore) you can achieve a much higher number (don't remember if it was 36, 42, 46 bits or what). I'm not sure, but I was also under the impression that more than 32 address pins were available from the CPU.. Can someone confirm or refute this? (I'm too lazy to look it up).
With this, a sufficiently designed database / scientific app could handle very large memory/ disk mappings without too much difficulty.
Of course, things get a lot easier when all the data is 64 bits. It's kind of like the comparison of the old 8086 with 16bit segment selectors which could address 20bits of memory. It was done, but it wasn't pretty, plus each contiguous block was limited in size.
It's the same situation as the x86 line.. It's a language that has proven so useful that it's archeic syntax is sadly being extended to handle classes it was never meant for.
Namely, constants, multi-dimensional arrays, data-typing have no place in a RAD prototype or a 5 line shell-script. But they _are_ useful in large-scale development (in data-type validation or in optimization). The idea of an integer multi-dimensional array, for example, allows the savings of potentially megs of memory. For one-shot operations or an occasional web page, this is an acceptible cost, but performance-minded developers will want to squeeze those 10's 50's of percents of improvement without switching to another language.
So what happens? Bolted on code.. You can't write a compiler that won't work with old code, and you can't require all users to learn entirely new code.. Thus evolution takes place (humans still have an appendix, ya know). I think it gives the language more character.. For example, anybody know why we have "y/x/y/" or $$? Legacy syntaxes that weren't hurting anybody by lingering on.
So Larry is laboring on that trade-off of clearity of new features verses compatibility (both in code and mind-set). The above syntax really involves attribute setting. The syntax may seem bizzar (and we'll probably lose several developers), but the net result will be an extension that is consistent and powerful.
Perl was and is perfect for the task for which it was invented. Additionally, it does a pretty darn good job of web development and small-scale scripted-applications (especially as wrappers). It's been tweaked to the point that it barely looks like a scripting language. It has an incredible freely available module library which furthers the notion of RAD. I can't say that it won't go the same way as COBOL (a la archaic legacy maintanance issues), but it's definately had an interesting life thus far. And at least we won't have lost the source code.
A couple years ago, we put forth a huge project to create an FDA-compliant document control system. We called it DCMS (document control management system). It was a strictly internal project, but I was the main developer and one day wanted to make a GPL'd version for the public.
Among the features: search criteria for many attributes, including indexed text-lookups (by using the unix "strings" tool). There were two modes.. One was "browser-mode", who could only search for "released" versions of documentation and click on a link to download the document. The other was "editor-mode", which required a secured login. From here you could browse any version: draft, approved and or released. At which point, you could "check-out" a document, which would lock it under your name and save a copy in your work-space, which is on a samba/nfs-exported directory (Your X drive in windows). It was su'd to your name so that only you could edit it. Documents were categorized into classes, subclasses, and document-types (product/project/quality.. Then product-name, project-number. And finally, doc-type which was technical spec (TS), functional spec (FS), requirements (RQ), etc..). Each document was given an auto-incrementing document number for unique identification.. So a document name was "DCMS_FS_00079.doc". Managers controled the doc-categories and creation of new documents.. It became a full time position, but FDA requirements are pretty strict.
When you were all done, you used the web to check it back in, which created a new daft version and allowed others to check it out.
authors would submit requests for review of documentation to manager-users, who would ultimately "approve" and later "release" a given version (making it available for public viewing).
Our first version used "clear-case" as the version control (which allowed direct URL-links to files, instead of explicit extraction from RCS), and as the meta-data storage repository (since it had it's own searching engines). We made heavy use of Perl and at the time Netscape-server.
I was in the process of rev-2.0 which would use Apache, Perl, RCS and MySQL and ideally be open-sourced, but our company took a major turn which seized all projects. However, DCMS is still fully running today (under different management), and aside from the performance issues of the clear-case dependencies, works great.
If I had the time, and if I were to do it today, I'd use postgres (for roll-back support), flat files (no real point to RCS if you let the DB do everything), Apache and maybe something like zope with python (though I'm still partial to perl).
I don't see how this has anything to do with Linux v.s. MS.
CDDB collected a set of information (IP) and sells access to it. They're basically getting paid for their hard work (as far as I can tell). Yes, someone can go to the record store and get that info.. But someone can also drive to a city to see what the night-life is; doesn't mean it's not useful enough for someone to pay someone else to compile that info.
Linux, on the other hand was written from the ground up, after simply looking at the functionality of other OS's, including ones that were already open.
The real question is whether the CD-informtion was stolen from CDDB; since they did, after all, spend time and money collecting it (free rider problem in economics). If not, then CDDB has no recourse. (shouldn't at least)
Slashdot Mirror
The following was my submission:
The SSSCA is a bill produced and supported out of valid fear; Fear of the volitility of GDP, and more recently, fear of illicit communications.
The former fear has been historically supported by American policy; to legally segregate markets so as to make local economies profitable. This is most evident in the former banking industry. Such ledgislation can be viewed as a "socially constructing" tax, which sits alongside tarrifs, tobacco taxes and income verse sales taxes. In all cases the ledgislatures balance the need against the cost.
That is to have a balance that is both "of the will of the people" and more importantly moral. Taxing a fixed sum from all citizens is arguably immoral (as could be seen in the dark ages). Backing a known corrupt monopoly is obviously immoral (which is recognized by semi-recent American ledgislatures through the inactment of anti-trust law).
The tricky part is to identify the will of the people in such a way that avoids mob-rule. Obviously anything that affects a majority of people negatively is subject to mob-rule. Exclusively taxing the rich, for example pushes too far into the realm of immorality and thus over-rules the opinion of the masses.
The SSSCA indirectly argues that intellectual property is soverign and should be protected by the government to whatever extent is possible (just as human life is to be protected). Obviously it is difficult to distribute IP and garuntee control, since once out of the protective sphere of it's owner, it can be manipulated, duplicated and redistributed (potentially for profit).
The DMCA legally protects digital copyright, but does nothing to prevent indirect circumvention (say via converting digital to analog and back again (unprotected)). The SSSCA closes this loop-hole by proposing to "regulate" the digital world such that no hardware or software can exist that "pollutes" the IP soverignty.
In the above, I have argued the pro point of view. Below I will expose the weaknesses and moral imperatives, and offer compromises.
First, the role of the government is virtually sacrad; To enforce a regional monopoly on "force". Any abuse of such a monopoly is obviously immoral. Allowing an individual to wrestle control of that monopoly for their own purposes is immoral. Upholding loopholes that allow individuals to "legally" attain control of that monopoly is immoral. An example of such legal perversion is law that can indefinately inprison citizens for non-immoral behavior. Here force is blindly applied to detain individuals that violate written law, thus anyone that calls apon a poorly written law which victimizes a moral citizen is abusing a loop-hole in the system. Such a system MUST not persist, but instead be mended. What's more it is the duty of ledgislatures to preempt such loopholes before they finalize laws.
While the primary goal of a government involves applying force to constrict a citizens use of force (or similarly to withstand foreign aggression), a physical ramification of this is social engineering via regulation. Due to a sence of "fairness" and or public safety, enforcers (police / military) have been instructed to apply the written regulations of conduct in addition to their force-monopoly. But the bounds to the American limit of efficability of such regulation is the US Constitution. However even this "minimalist" view of moral judgement is fallible (as can be seen with the introduction of failed prohibition or even slavery laws).
Thus in order to preempt moral imperatives that take supreme-court ruling and eventual lengthy legislation to amend the constitution, it is most arguably best to apply moral scurtiny to the laws before they ever achieve legal status. A ledislature MUST, therefore respect the sacred nature of his office, lest [s]he be shunned in history.
Regulating banking maintains local economies at the expense of consumers. The cost is moderate and monetary and thus moral. Regulating taxes likewise is monetarily based but runs the risk of starvation of subsets of citizens; the worst case involves death and thus achieves high scrutiny and modern governments tend to fair well in this camp. Public safety regulation prevents producers from trading safety for profit (making an industry more fair by requiring all producers utilize the more expensive safety measures). In the short term, the negatives are higher cost-of-goods to a consumer as well as the abuse of force in telling citizens how to conduct themselves. In the longer run, economies of scale diminish the extra cost of goods, while the "forced conduct" prevents indiscriminant loss of life. Hense the government is "saving us from ourselves", and more importantly reducing an automobile's role as a weapon. Next to public safety are physical property laws; those which "regulate our conduct" such that we do not deprive one another of our physical property. Since food and shelter are life-sustaining forms of property, generalizing governmental enforcement of property ownership is valuable; e.g. a government is justified in protecting the ownership of a CD just as much as protecting the milk that a baby carries.
At the bottom tear of "moral" laws are the corporate laws. These are "rules of conduct" laws which have nothing to do with maintaing a government, nor saving lives, but instead deal with the secondary effects of economies. An economy is efficient and can flurish when various rules are applied. This is where tax-games can be played to encourage conduct or physical regulation (such as hours of stock-trading, or margin-limits) are applied to avoid economic disasters. In here too is the concept of patent law and copyright, which introduces the concept of intellectual property. The social cost of a patent is to restrict our actions such that we don't act in a manner that mimics one expressly copyrighted or patented. While the intended effects are purely economical (and thus not morally justifiable), the unintended effects are to allow individuals to call apon the force of government to control citizens. Whether the citizen was indeed attempting to profit on the IP-holder is not relavent to the cost. A mother that sings happy birthday to her child is in violation of IP and thus is legally vulnerable to incarseration (if the IP-holder was sufficiently immoral). This is a dire perversion and loophole that allows a single individual to wrestle control of the government's monopoly.
One possible remidy would be to apply the condition of profitability to all copyrights and patentents. The IP-holder attains legal rights to all revenue based on that IP. This does not stiffle do-gooders who utilize the IP for charity or public-good will. Arguably, even the GPL would be in violation of such IP, since it allows for the distribution charging (which can be a disguised form of profiteering). Such a form of modifiation of existing legislation arguably needs to avoid loop-holes (mostly involving proofiteering, since the original intention was robust GDP).
The DMCA assumes copyright is a fundamental right of IP (as are life and liberty. This is, in my opinion, a perversion of the "right of happiness"). It assumes that copyright is so important that not only can you _prevent_ redistribution profiteering, but you can dictate how and when the IP can be accessed, which is a gross extension of the concept of governmentally enforced social control. It is more than simply restricting when you make your IP available to subsets of the public (which is perfectly acceptible, as a performer could choose when and where they perform), but to say that you can only hear it a certain way, to talk about it a certain way (perverted slander laws that defend corporations, even when they are blame-worthy), or to witness the IP via a monopolized medium. The DVD, the up-and-comming HDTV anti-time-delayed-viewing, the encrypted pdf format and others are examples of such IP who's viewing capability is monopolized. That companies are capable of
restricting viewership is not, in and of itself, immoral. A theater, for example, can monopolize on refreshments due to it's locality; they can deny admission if need be. But to even suggest that when you watch a home-movie, you can only eat a certain name-brand refreshment is totalitarian plain-and simple. But this is exactly what the DMCA does. It produces an artificial locality, and an artificial scarcity (via fear of governmental retaliation, wielded by the very companies that profit from your patronage). This is a direct conflict of interest; a man may provide a quasi-desirable service while wielding a governmental whip which says that you have to partake from it. So long as there are alternatives, there is the prospect of "choosing not to partake" in the bait that locks you into the negative aspects of a service, but there is a natural effect which brings like businesses into the same immoral tactics due to lack of profitability.
It is perfectly valid for manufacturers of DVD to force commercials and regional encoding, and restrict viewers to set-top boxes or highly trusted windows machines. But when ALL major IP owners in the media field shift to DVD, a consumer will have the choice of boycotting ALL cinematic/ auditory IP short of live concerts. VHS and casset will be phased out and your living room will be an extension of the controlled theater. Even this isn't a "really" a problem because we still have choice. But what the DMCA allows is the removal of other forms of media. It is illegal to convert a DVD to VHS or even to record the sound-track to cassette, since this would circumvent the encryption. Granted the intent of DMCA is to prevent the perfect replication of data which can then potentially be advantaged by profiteers (or more recently, in the production of a public "library of music" which has the potential for reducing the demand for highly priced official distribution points). But the DMCA is a loaded weapon which can be fired at existing acceptible practices (such as making a cassette for use in the car). The hidden answer is that corporations never wanted us to have the ability to freely utilized purchased licenses of IP; they simply lacked the ability to enforce it.
The real long-term question is as internet bandwith expands and the quality of compression software approaches that of the original masters, how can a media-based industry survive (and thereby uphold GDP which benifits all citizens indirectly). My answer is that a corporation can and should "squeeze every dollar" it can from an industry, so long as the dirty work is not enforced by the government; such an activity is called extortion, and such an organization is called the mafia. Times change and it is up to business-heads to determine new manners of supply and demand to survive. It is not moral to enforce laws that restrict citizens rights so that aging task-masters can retain their old positions. This century has seen the horrors of worker-layoffs.. It's time that the "task-masters" experience their share of layoffs. If NSync can't retain their multi-million-dollar industry, then let them suffer with more concerts and new forms of high-bandwidth (multi-channel) audio-CDs priced lower to catch the interest of more of the waning support-crowd.
The main point is that back-door operations, and incogneto sharing of IP will always exist, and times / public interest will most definately change. This is merely a last-ditch effort to safe stock-face in the short-run and punish the most visible; the naieve citizens, or the high-profile educators that teach on the topic of encryption, or the security experts who crack systems to improve apon their security. Those that are actually helping society will be hurt the most. The industry will still slump and illicit profiteering will still proliferate (since one additional law isn't going to make them any more legally vulnerable).
Lastly, the SSSCA can be shown to be built apon a series poor-ledgislation, which most definately holds to the spirit of it's predecessors, but in a most perverse way misses the whole point. The regulation that places requirements on specifications within hardware limits the diversity of that hardware, first-and foremost. The ancenstral point of the SSSCA is to foster growth, and innovation, but at step one, it grinds innovation to a screetching halt. How can an "embedded device" in your car be required to provide protected encrypted streaming audio? Surely the computers within the cars of the future will have the ability to do some tasks currently performed on laptops or PCs. But at what point will they be held liable? The answer is whenever it is most convinient for the corporate task-masters. Manufacturers will live in fear of backlash so long as they exist on that fine-line. Many will simply forego advancement, and innovation will be halted.
The government has made several failed attempts to regulate based on flawed information. Regional governments have required gas addatives that were later learned to be harmful to the environment. The best way to avoid such forced lemming rank-and-filing of citizens (doomed to fall over a cliff) is to identify the problem and require "A" solution, but then allow innovation to fullfill that requirement. To suggest that "government knows" best, and that "this is for our own good" is as blind as Hitler. (If the analogy is offensive, then ask yourself why?)
The analogy of a required built-in security equipment to a catalitic converter works when you consider that both help prevent environmental pollution. In this case, the digital pollution of economicly viable media. Both cars and computers are perfectly competative goods (within their genre) which fulfill very similar purposes and are essentially black-boxes to their users. But note that a car can be electric, and thus forego the need for a catalitic converter out of "common sence". But there is no "common sence" when it comes to computer science research, since it is a fronteer who's tip has barely been identified. This would be like requiring that "all vehicals" contain electrically shielded engine cranks 50 years after the electric starter had already been invented. There exists hardware and software alike that is incompatible with such low-level drivers that provide controlled access to media-devices. This level of information is completely incomprehensible to most ledgislatures (since it's inside the box), but it as the heart of digital science. To provide such a vague law allows individuals to wield the power of the government to stiffle their competition in this cut-throat science. The ability to provide an increasingly out-dated media in a governmentally controlled fashion comes at the cost of intellectuals who will no longer have the ability to freely experiment in the digital world. As competition for innovation comes to a half, so will the desire for innovation, and all that will be left are the greedy hands of software and hardware monopolies which will regularly fly in the face of ledgislatures as an embarrasement of capitalism. Capitalism only works [as a moral entity] in a governmentally minimalistic society.
IP laws such as copyright, trademarks and patents have values which outweigh most of the costs, so long as the loop-holes described above are eventually mended. The DMCA is completely redudant with laws that physically address profiteering, and only opens a vagueness which allows corporate abuse of governmental force. The SSSCA, if enacted will backfire on both GDP and legislatures and will eventually be repealed. It is best to preempt a human tragedy before letting court appeal delays take it's toll.
-Michael
How's that different than credit / checking today? There will always be some sort of "tradeable" currency. Even if it's computers / cars, etc. They can't force you to use a certain type of transaction when you're really trying to be discrete.
What I see is the same as the gold-standard, which was "credit redeemable in cash" stamped on the cards. Coins will probably be phased out (if inflation hits again and no soda costs less than a dollar, which you already see in high profile areas with vending machines), but dollar bills should be available for more than your life-time.
In the worst case, the mafia has always been good at money laundring. I'm sure there'll be pawn-shops that act as fronts (for good ole' fashioned trade).
-Michael
I'm not sure that this was the case. It wasn't that the valued goods became scarce, so much as it was impossible to wield $100,000 around in your pocket. Within a given country, the government provided security as for the value of coin, and in the worst case, that coin had some value if melted down. Later as coutries trusted this sort of exchange fiat money (or completely worthless) was used.
This required the trust that you could get something valuable back if you wanted (say to trade internationally).
In the US at least, it was eventually determined that the economy need to grow and shring, and that fixing equity on stocked goods was innefficient. If we had inflation, for example, we would have liked to have introduced new cash into society to compensate since the price of gold (the US's former standard) didn't directly vary with the rate of inflation.
Things were still safe because you could regulate the printing/coining of fiat monies. But then checking became very popular. Now you had the concept of float. One bank would honor a check (and allow accumulation of credit/cash) before the debited bank could deduct.
Later we have the concept of equity-based loans. I percieve that your good is valued such that I'll lend you most of the money for it. You take that money and spend it (via checks), but more goods and take loans out on them..
All in all, checkable money develops a velocity (the rate at which the same virtual or physical dollar is spent per year) such that our net assets are multiple times the physical printed fiat dollars total value.
In a booming economy, that multiplier increases. The problem is that that rate of boom has to be maintained or there will be a dramatic credit crunch. A recession after a boom is devistating because trillions of dollars can up and dissapear (after all checks are registered).
This would have happened even with a gold standard due to virtual assets and value.
The issue has always been one of efficiency. Yes we're more at risk now that a single number can render our bank-account empty. But we have a much greater ability to refill that bank-account than we did when someone with TNT could "blow the safe" and bring you back to square one. You can be insured, bring out new mortages so you don't starve, and most importantly be paid a heck of a lot more than days of old due to incredible industry efficiencies.
-Michael
That may have been its most obvious selling point, but that was far from it's only use. It was a cross-platform method of client/server programming. You could develop sophisticated applications with incredible performance characteristics (compared to simple DHTML) that you could distribute to a heterogenious intranet (Solaris/Linux/win32).
Eventually DHTML got more sophisticated, and generating graphics for HTML was an order of magnitude simpler than for Java. Thanks to more reliable javascript, you can also produce many of the dynamic operations that previously required Java. (procedural drop-downs, tabbed pages, refreshing, etc).
-Michael
Possibly, but you'd have to ship the compiler with the dist since you would be hard pressed to link dissimilar code together dynamically. I'm no expert on this, but I do remember woes with using a sun compiler on Solarix x86 with the gcc compiler. This failed miserably with Apache / mod_perl and with simple perl + CPAN libraries. We had to go with the Solaris comipler all the way, which was a royal pain, let me tell you.
Unless someone has some info to the contrary, you'd have to forgoe most any precompiled linux binaries, which will definately get into your hair, as I've definately found.
I'm not completely sure, but doesn't gcc extend C with various types of proprietary compiler attributes? I believe it's possible that the configuration stage can nullify them, and it's been a while since I've looked through Linux source, but I do remember those attributes hanging around.
Still, I'm sure it's possible, and I'd be curious to learn of anyone's success..
On the massochistic side, has anyone compiled Linux with a MicroSoft comipler?
-Michael
The bottleneck is likely to be the connection to the monitor. Unless this is an all digital monitor (which I guess would be possible), then you still have to pass an analog signal which is limited by the quality of the cable (bandwidth) and the performance of the video card.
even if it's digital, there's still the issue of feeding all that info. You can't just have one pin per pixel, so you'll need to scan by some means. I thought that there were several competing techniques to accomplish this with LCD's though I'm not too familiar with what is mostly used.
Point being that there's no garuntee that it has any faster refresh rates than traditional CRTs, nor do we know if there will be as much fading / trailing as in some older laptops.
-Michael
On power consumption:
Talking out my ass here, but it seems to reason that even though the distance between cathode and anode is smaller (requiring less voltage), there is a wider area from which to draw electrons so the current might even be higher. Also, at first glance, it seems that there are more segments that require power due to a larger number of beams. There's probably some inefficiency there too.
On quality:
Does anybody know how well this would hold out against spot imperfections? One problem I see with LCD is that you can have spot imperfections that accumulate over time. With CRT you get general distortion. I tend to prefer the distortion because it has smooth transitions that the human eye adjusts to (I buy used monitors with warps and spots at a great discount). But a dot stands out and is an irritation (high contrast). It seems to me that the flat-CRT will be just as suseptible to local-imperfections as LCD due to localized damage to anodes.
Conclusions:
Assuming the cost is greater than CRT, then the only thing you get from these monitors is space savings.. If you don't absolutely need it, it probably won't be worth it. I like my used $260 21" Hitachi just fine.
I can't really see BSD / Linux, etc being used by the general non-CS public in an educational setting. Sure it might be good for educational / managerial utilities (databases, calanders, to-do, etc). Even using it as a simple web-browser in the class-room / library (on slower machines). The problem is that it still doesn't cut the mustard for office apps. I'd venture to say that 99% of students don't have a Linux-running-Office supporting OS environment at home, and administering this would be a nightmare. Most people are going to use MS Word these days, and that's what you'll need sitting somewhere on school-grounds.
It might be possible to reduce the number of Windows machines through the infiltration of Linux, but then you'd have to have UNIX administrators. Sure the acting IS staff (typically a single guy) could learn it, but can a given school justify sending these poor-overworked people to UNIX training when they're only going to provide marginal utility for the school? I discourage schools from out-sourcing such administration to student groups for several reasons.
-Michael
tcsh:
set prompt = "`echo $HOST |cut -c1-2`:\\%~>"
# "co:\full-path> " for computers starting with co
bash:
PS1="[\u@\h \W]\$"
-Michael
I'm not really going to try and explain such an open-ended cop-out, except to say that most recounted experiences with "time-travel" or more simply temporal e.s.p. leave one in a sence of a dream-state.
From what I got, everything in the universe leaves it's imprint on the analog universe, much like ripples in a zero-resistance ocean, or what-ever analogy floats your boat. They speculate that the physical piece of matter (possibly the complex DNA strand, but also possibly the matter itself) is like a finger-print that can be used to searched in the cosmic ocean to reach-back for the rest of it's constituent parts. It does suggests that time-travel isn't possible (or they wouldn't really need to ecscavate, nor would it be a problem to bring her here).
I don't quite know if they're mearly finding her personality / her essence, or if they're projecting esp to or fro.
As a good master, you don't explain the details, but try and find some high-level analogy that describes the functional parameters. Details can be depressing. Something Lucas should be-retaut.
-Michael
haha.. laughable. We're hard-wired for sexual desire, for fear of danger, for blinking when things approach the eye, or for mothering rage. So why don't we react like a programmed interrupt controller when certain conditions arise? Because the main point of our nuerons is that they're programmable. We have both actuators and inhibitors. They both battle over the initiation of events. Yes there's a hard-wiring to put into motion a response for a given perceived action. BUT, we also have the ability to surpress those hard-wired reactions.
You could go to great lengths to encode the reward / punishment / reaction system within a neural net, but the fundamental nature of neurons is that they can be super-ceeded (and the super-ceedings can become the new nature, only later to be super-ceeded by something else). A tough man learns to not blink, and to restrain his anger. Most men learn to supress their sexual urges. Most people learn to control their need to excrete.
The only way they could do it would be to have an external response mechanism that doesn't allow over-riding (which would defeat most of the point of being alive and having an adaptive neural-net).
Obviously they do this, because of the activation code; that's something that is rather important to not 'over-ride'. But notice how little of that sort of activity is actually used.
-Michael
maintanence? He definately has air-passages and vocal devices. Even if the food didn't go to his stomach, certainly these devices would be vulnerable.
-Michael
From a fictional point of view, it's not out of bounds obviously.
I can't believe nobody has introduced RoboCop into the discussion (at least the first one). Here was an attempt to integrate the human psyche with computerized control (his prime directives). The parallel is in having human emotion while having a glass ceiling. They did a half decent job in that movie exploring the complications involved. I don't believe that this movie really wanted to explore these complexities. As justification, the makers were "trying so hard to see if they could, that they never stopped to see if they should". In fact, all the emotional trauma that was caused was ultimately encouraged.
Thus, in agreement with you and in start contrast to the previous poster, asimov's laws had no place in this movie.
-Michael
Get a PDA. Then the difficulty is bumped up a notch.. First they have to get access to your person.. Then they have to crack the PW-manager password. Then they have to figure out which password goes with which site (hopefully you didn't put all that info together).
Only trick is if you lose your PDA. So maybe you should have a tape-archive with the info periodically. and put that into a safe where you have the only key.
-Michael
I disagree. This is definately the case with a 486 to Athlon comparison, but we're already taking into account the architectural differences (stages / pipe, etc). Part of the analysis is to monitor efficiency. This is especially true with the Pentium4 / Athlon debate since we can get 1.4GHZ Athlons.. The question is whether to purchase a 1.4GHZ Pentium 4 at significantly higher cost; to say nothing of the added cost of a 1.7GHZ setup.
The difference is more dramatic between the P5-4 and P5-3, since you max out at about 1GHZ for the P5-3, and so I'd be inclined to believe you. The Athlon, however is not yet out of steam for its current. If it can best the P5-4 in 50% of the categories (including legacy apps.. e.g. modern ones), then the value of the P5-4 is limited, even if it can produce top-notch synthetic scores.
The point is that it is not ignorant to normalize, so long as you look at the periferal factors. It's like having taking the average, but also taking the standard deviation. You do find useful information from such numbers.
-Michael
Believe you're thinking of the number it can "issue", which is separate than the number of [semi-]independent pipes. In the PPro, some instructions (like divide) would lock other pipes or stages within it's own pipe. Issuing instructions is expensive, so it's generally accepted that you issue less than the number of pipes, but as the P4/Athlon have significantly more pipes than their predecessors, they have augmented the number of issued instructinos by 1 or so.
-Michael
Not sure that I'm reading you correcty. My initial impression is that SMT / CMP would hurt cache hits. If you had an app that was single threaded, then obviously SMT won't help you, but CMP would have you compete for cache space. If you had a multi-threaded app, then yes their code-cache would most likely have less thrashing, but their data stands a good chance of competing for the same space.. In single threaded operation, your cache can afford to risk having 2, 4, 16, etc memory locations overlap on a cache line, since it's not too statistically likely that you'll thrash.. But if you have SMT, then various types of applications that require large data-sets (such as text-processing web servers), enhance the chances for accessing conflicting memory regions. Even with more expensive cache architectures, the likely-hood of cache-conflicts is still higher with SMT.
My understanding of the proposed SMT on x86 is that you simply switch to another thread when there's a memory stall. I think SPARCs have done that for a while... What I believe you're referring to is the reduction in the number of times you have to context switch and thereby flush your cache. Though it's true that having fewer distinct processes (even LWP ones) requires fewer context switches, I believe that you are not given a time-delta extension simply because you have 2 or more threads associated with a process for an SMT core. Thus, I believe the time-delta is still the same for all processes (minus HW interrupts), and the number of cache flushes per second is the same. Hence, little realized benifit.
Just for completeness, what I think you do get is fewer memory stalls within your time-delta. Additionally, if each thread is stalling, then you at least have multiple concurrent memory requests, which I believe does suite RDRAM well. You could achieve a similar situation by having multiple independant banks of SDRAM (like nVida's GeForce 3).
In summary, if anything, cache is the weak link towards multi-core / multi-threading.
yes and no.. how do you categorize a new chip? The AMD K5 / K6 were roughly inline with the P5, but they were separately designed.. It of course comes down to marketing. BUT, what you can look at is the generation of the design. Pentium introduced (for x86's anyway) relatively deep pipes and multi-instructions. The next generation was OOE. You may or may not be able to categorize the Pentium4 as a new generation based on it's double-pumped integer. I think that all the other aspects of the P4 are simply augmentations or incorporations of nifty ideas (like caching the decoded ops, which I believe AMD did a while ago).
I use to call the IA-64 the P7 just so that may lay-friends could know what I was talking about. It's VLIW / speculative execution could probably be considered a new generation. But in reality it's a completely separate product with hardly any ability to compare to the x86 line.
I think, however, that I'd recognize SMT / CMP as a next generation label.
-Michael
This is an interesting take, but there are other considerations.. First, if you read the Findings of Fact against MS, you can easily be led to believe that MS's real purpose was to lock java to the MS platform by polluting it was ms-only extensions that proliferate on the net. This practice showed up too often in other areas to ignore its likelihood.
Next, why would MS want a write-once run anywhere development environment for themselves. They're not about to build their drivers and win32 API in Java, and any apps that they build on top of them is pure C++, so all it would take is a simple recompile for the different platforms.
When Java came out, I don't believe that Alpha-NT was that popular, and SGI-NT was being dropped (not certain about the timing, but it seems about right).
I agree with you about win 9x being stepping stones, but I don't think cross-platform was a big focus for NT. Yeah they have the Hardware abstraction layer, but I don't know that this wasn't more for stability and protected code than for true platform independance. Thought it was really just a carry-over from VMS.
-Michael
There is more exteraneous noise in Java than in several other OO-capable languages (C++ excluded). This extra required syntax and the REQUIREMENT to program in OO makes java as a first or even high-school language difficult.
If you take someone straight out of BASIC or Pascal and throw them into Java, you might have some obsticles - They're learning too much all at once. In High school, the added complexity probably isn't warranted (we want to spark interest in our youth), and in College, people that transition majors may need more hand-holding.
Not really true nor a factor. For a large number of institutions (primarily High School), Windows is the development platform. It's harder to get easy-to-use free Java tools for Windows. Yes, you can find them, but it's not an obvious choice. The second part of my rebuttle suggests that an institution spends the money on the right tools (within their priorities for a given carriculum). We used Apple IIe's with Apple Pascal. It was a purchased license, but the whole system cost was marginal.
To counter this, Java requires expensive hardware (namely, no donated 386 / 486 is going to be happy in a [education] development environment), though these are probably what you'll find in poorer schools.. Which has to facilitate advanced students taking AP courses...
Then they'd all teach Visual Basic as a first language. Since you're garunteed some sort of employment. It's almost impossible to avoid VB these days.
It's not strictly true that x86's are limited to 32bits of addressable memory. With VM and segment selectors (which I personally diplore) you can achieve a much higher number (don't remember if it was 36, 42, 46 bits or what). I'm not sure, but I was also under the impression that more than 32 address pins were available from the CPU.. Can someone confirm or refute this? (I'm too lazy to look it up).
With this, a sufficiently designed database / scientific app could handle very large memory/ disk mappings without too much difficulty.
Of course, things get a lot easier when all the data is 64 bits. It's kind of like the comparison of the old 8086 with 16bit segment selectors which could address 20bits of memory. It was done, but it wasn't pretty, plus each contiguous block was limited in size.
-Michael
It's the same situation as the x86 line.. It's a language that has proven so useful that it's archeic syntax is sadly being extended to handle classes it was never meant for.
Namely, constants, multi-dimensional arrays, data-typing have no place in a RAD prototype or a 5 line shell-script. But they _are_ useful in large-scale development (in data-type validation or in optimization). The idea of an integer multi-dimensional array, for example, allows the savings of potentially megs of memory. For one-shot operations or an occasional web page, this is an acceptible cost, but performance-minded developers will want to squeeze those 10's 50's of percents of improvement without switching to another language.
So what happens? Bolted on code.. You can't write a compiler that won't work with old code, and you can't require all users to learn entirely new code.. Thus evolution takes place (humans still have an appendix, ya know). I think it gives the language more character.. For example, anybody know why we have "y/x/y/" or $$? Legacy syntaxes that weren't hurting anybody by lingering on.
So Larry is laboring on that trade-off of clearity of new features verses compatibility (both in code and mind-set). The above syntax really involves attribute setting. The syntax may seem bizzar (and we'll probably lose several developers), but the net result will be an extension that is consistent and powerful.
Perl was and is perfect for the task for which it was invented. Additionally, it does a pretty darn good job of web development and small-scale scripted-applications (especially as wrappers). It's been tweaked to the point that it barely looks like a scripting language. It has an incredible freely available module library which furthers the notion of RAD. I can't say that it won't go the same way as COBOL (a la archaic legacy maintanance issues), but it's definately had an interesting life thus far. And at least we won't have lost the source code.
A couple years ago, we put forth a huge project to create an FDA-compliant document control system. We called it DCMS (document control management system). It was a strictly internal project, but I was the main developer and one day wanted to make a GPL'd version for the public.
Among the features: search criteria for many attributes, including indexed text-lookups (by using the unix "strings" tool). There were two modes.. One was "browser-mode", who could only search for "released" versions of documentation and click on a link to download the document. The other was "editor-mode", which required a secured login. From here you could browse any version: draft, approved and or released. At which point, you could "check-out" a document, which would lock it under your name and save a copy in your work-space, which is on a samba/nfs-exported directory (Your X drive in windows). It was su'd to your name so that only you could edit it. Documents were categorized into classes, subclasses, and document-types (product/project/quality.. Then product-name, project-number. And finally, doc-type which was technical spec (TS), functional spec (FS), requirements (RQ), etc..). Each document was given an auto-incrementing document number for unique identification.. So a document name was "DCMS_FS_00079.doc". Managers controled the doc-categories and creation of new documents.. It became a full time position, but FDA requirements are pretty strict.
When you were all done, you used the web to check it back in, which created a new daft version and allowed others to check it out.
authors would submit requests for review of documentation to manager-users, who would ultimately "approve" and later "release" a given version (making it available for public viewing).
Our first version used "clear-case" as the version control (which allowed direct URL-links to files, instead of explicit extraction from RCS), and as the meta-data storage repository (since it had it's own searching engines). We made heavy use of Perl and at the time Netscape-server.
I was in the process of rev-2.0 which would use Apache, Perl, RCS and MySQL and ideally be open-sourced, but our company took a major turn which seized all projects. However, DCMS is still fully running today (under different management), and aside from the performance issues of the clear-case dependencies, works great.
If I had the time, and if I were to do it today, I'd use postgres (for roll-back support), flat files (no real point to RCS if you let the DB do everything), Apache and maybe something like zope with python (though I'm still partial to perl).
-Michael
I don't see how this has anything to do with Linux v.s. MS.
CDDB collected a set of information (IP) and sells access to it. They're basically getting paid for their hard work (as far as I can tell). Yes, someone can go to the record store and get that info.. But someone can also drive to a city to see what the night-life is; doesn't mean it's not useful enough for someone to pay someone else to compile that info.
Linux, on the other hand was written from the ground up, after simply looking at the functionality of other OS's, including ones that were already open.
The real question is whether the CD-informtion was stolen from CDDB; since they did, after all, spend time and money collecting it (free rider problem in economics). If not, then CDDB has no recourse. (shouldn't at least)
-Michael
OS/2, Solaris x86, QNX, DR-DOS, SCO-UNIX, QEM, novell
...
Most have professional interfaces, and some are even user-friendly.