From that point of view, graduate school is getting a PhD. Masters and professional students serve two purposes: fund raising and an outlet for failed/burned out Ph.D. students.
Maybe it is different in engineering, but the vast majority of my professors have separate masters and PhDs. When I was investigating graduate school the expectation seemed to be that I would get a master's, then a apply to a doctorate program. Is that really so unusual?
6 to 9 months left on master's degree, looking for a project or thesis advisor, will commit unspeakable acts to obtain one with english proficiency. Would commit slightly unplesant ones to talk to professors who would just say what they wanted.
Just wanted to back this up, maybe it's different outside of engineering, but in engineering, even engineering acadamia, the actual engineer has basically no rights to anything they make. This applies equally to students and professors at the colleges I have been to.
If you're a research engineer then your compensation is that you get to publish a paper which is either the goal in itself, or you hope will lead to someone taking you more seriously (possibly paying you more, but not always) on another project. On the other hand, at least where I did my undergrad, it was expect that if you didn't research (you were primarily interested in teaching, for example) you instead worked on projects for the school, and remained as anonymous as any normal engineer.
I've got a few problems with the advertising driven business model. The first two are personal, the rest are more general concers I have with the advertising econemy.
The first is that most general advertising assumes I'm too stupid to know what I want, or to find information on products to meet those desires even if I do know what I want. Telemarketing is the most heinous example of this kind of marketing. I have received telemarketing calls asking if I wanted to change my long distance carrier on numbers with no long distance service. I get calls asking if I'd like a deal on home siding and I live in an apartment building. How is this not the worst possible waste?
The second is that in advertising products don't even pretend compete on merits. It is pretty insulting to be surrounded by base plays on my emotions all the time.
More generally, advertising seems like a bit of a pyramid scheme. Many things are financed almost entirely by advertisements and the money for those ads has to come from the consumer somewhere.
Most advertising casts a very wide net and a view results in a sale only a very small fraction of the time, and rarely in a traceable manner. This means the efficacy of any particular ad is always uncertain. Most advertising dollars are therefore wasted for many reasonable definitions of "waste."
Given that I don't see personally targeted ads on the sides of busses and subway cars, before movies, during television shows, or even on the internet happening any time soon doesn't htis mean there's an incentive to shape advertising sponsored materials towards those most likely to respond to the most generic advertising?
Doesn't this mean that I as a consumer have almost no way to directly sponsor a normally advertising supported endevor that I enjoy or find useful? If I don't respond to ads I'm useless to them.
Doesn't this mean that every time I buy something I'm indirectly sponsoring whatever that company spends it advertising money on? There is no garuntee that the product I buy and where the producers advertise will be even remotely related. Even assuming that it was always well known where a companie's advertising dollars were spent, I must now include that knowledge in my product selection, moving us still further from products competing on appropriate merits. Since it isn't always apparent what will my purchase will really be supporting, I also have to consider the time wasted in finding that information as well.
Most insultingly the fact that so much is funded nearly exclusivly through advertising means that advertisers feel that they're buying the eyeballs of consumers directly, and the content producers who are dependant on advertising income are inclined to agree with them. This gives rise to things like trying to stop consumers from fast-forwarding through commercials, but also to product placement as a more subtle but equally insideous plan to make advertising, which almost surely will be for something most of the people viewing the ad won't care about, harder to ignore.
So basically as far as I can see the advertising econemy is largely about further weakening the consumer's direct influence on the market and ability to make informed, intelligent decisions even if he or she wants to.
Well by consumer software I basically meant software whose sponsors arn't among its primary users, consumer, off the shelf software (COTS) is an acronym applied equally to a a consumer buying a video game and a business going out and buying a prefab RDBMS. That's immaterial though because my real point was that FOSS might (would) change the way software projects are selected.
The effect on small businesses with unusual needs might be bad, but it won't be any worse than it is now. If a small business has needs that can't be met by conventional consumer, off the shelf software packages and they can't afford to pay software engineers of their own, they're still screwed.
Many small to mediums sized businesses actually do have limited programming resources at their disposal. this is often either a consultant who works for and bills them on demand, or a single, cave-troll style programmer.
Which brings up another area of concern with 'scratch an itch' software of any kind, particularly if you're expecting it to be used by consumers under FOSS, the software process. All the things that go into software development that are not programming tasks. Requirements engineering, software architecture, verification and validation, documentation and training materials, those kinds of things. I'm not blind, I'm well aware that many commercial software projects lack good software process and that some F/OSS projects have pretty good process. "Scratch an itch" software doesn't tend to be among the projects with good process on either side, and when you intend for consumers to use it, that's a problem.
Support contracts would appear to be a big part of this.
From a consumer's perspective that is the problem. As a software engineer I work for whoever can pay me, so in that regard FOSS changes nothing. Given that, the next question is "Who can afford to pay me?" and that anwser also remains unchanged, businesses. Most consumers just can't afford to pay a development team to make or modify software to meet their needs, and now using a FOSS model there is no inscentive for a business to make consumer software since the software is only worth what the single highest bidder is willing to pay for it, and consumers usually motivate businesses through many people paying a lower price.
Perhaps there's a simple resolution to this that I'm not seeing, but until that solution is widely known, FOSS really isn't in the consumer's direct best interest. If you've got a solution, please let me know.
To expect safety after defeating barriers and interlocks is stupid for microwave ovens and toasters, let alone high energy robotic systems. To expect robots to be safe outside of their defined operating parameters is like expecting a car to be made of sponge so no matter how much you ignore the speed limit, you can't kill anyone.
I say the same sort of thing about software all the time, good luck getting people to understand that.
Even this service has quite a few flaws, here's what comes to mind while signing up for my 'free trial' an indie music catalog I can't search without giving them my credit card number, that's a huge red flag. Subscription service not a use based service, that's another. HTML newsletters, have another! Proprietary software, that shockingly requires an installer program on OS X yet another!
Look, I realize that in theory if you cancel before your trial period is up you don't get charged, and that you don't *have* to sign up for the newsletters or install their prorpiretary download manager, but in many ways this isn't any better than predatory CD clubs. The business model is based on the idea that some fraction of people won't use all their monthly downloads, effectivly wasting money, that most people will elect to install the software or sign up for at least one of the newsletters, thus permitting at least some spam. I'm really just not that enthused by business models based on customer stupidity, be it this or DRM.
First, I think you're mashing up several markets in ways that they aren't really related. You talk about business software liscencing as though they buy a liscence for every seat, for the majority of significant products that just isn't true and the pricing model looks a lot more like what you describe already. Businesses pay for a maximum number of concurrent users, and the liscences they have are deployed and recalled as needed from a central liscence server.
It is unlikely that many businesses would invest heavily in the other aspects of the model you describe exactly because "your desktop follows you wherever you go," meaning your files are stored on a server controlled by a third party and that they can easily be spread around by users without them even thinking about it. Since the business is likely to desire a high degree of control over what employees do with the business's files, as well as what applications their employees have access to, and since both businesses and users have an interest in preventing the comingling of business and personal files, (especially if they are charged for storage and bandwidth which they almost certainly will be) it is quite likely that even if this became the only way of getting MS software MS would still be double-dipping because the user would pay for their personal account and their employer would pay for a separate work account for them.
Further, if your files are stored on a Microsoft server as you imply, and there is no danger of saving them locally, as you say, (though this is doubtful) then what interest would Microsoft have in letting you have your files at all? Think about it, your monthly MS liscence fee could basically be a ransom on your data.
I find it highly unlikely that MS would liscence their 'software as a service' to Google in such a way that Google could provide it free, or more cheaply than Microsoft. Why not charge Google at least as much per user as you'd charge the users themselves?
So, discounting problems like "Where will the bandwith to support this model come from?" the basic hurdles here are, how will bandwith and storage be paid for? how would this giant pot-of-gold where user's files are stored be protected? how would the user be protected against malicious parties on route from those servers to (and including) the terminal? What if that kiosk you mentioned saved all the user data that passed through, or just the user's credentials? What kind of access control mechanisms would be put in place to allow businesses the kinds of control and separation of their files that they might need? and what garuntee do users and businesses have that MS won't just hold their data hostage?
My point was, even if what you say were true, (it's not exactly) then other things would have to be sacrificed to produce it (like time) and very few stakeholders are willing to pay that price. You could say the same thing about textbooks, nothing except the author prevents them from being flawless either, but many of them still have volumes of eratta.
Legacy code support is a pretty good counter example to what you say. Many programs don't work if Windows is running in an unprivelaged mode. If they were originally developed before Windows even had such a mode, or before it worked the way it does now (and it is changing yet again in Vista) how the author have known about these requirements before they existed? Many aspects of Windows undergo such changes fairly regularly even within releases (Win2K SP2 was a great example of this).
If I write a Java application that works fine on the JRE but doesn't work on on the third party alternatives, which one of us has the bug? If I write a web page that doesn't display the same way on all browsers, who has the bug? Now who is more likely to be asked to fix it?
Your statement that all problems that can be fixed can be avoided requires that there be some possible problem-free ideal, this simply isn't so. Some quality attributes exist in opposition to each other and you simply can't optimize for them all equally well. This is pretty clear when, for example, finding a home. Given a budget, desired location, and a desired size not all combinations of values are equally viable. Some are obviously absurd, you're almost surely not going to find a 3 bedroom house for $10 a month, others are attainable only if you sacrifice something else. You might be able to get a studio for $300 a month, but it is more likely in say, Milwaukee than in Manhattan. Size, price, and location all obviously affect consumers, even if the factors that shape the curve of viable values are not their concern. On the other hand this cuts both ways, if you're a consumer and you want a $10 a month 3 bedroom appartment in Manhattan, it isn't your real estate agent's problem, because it just isn't viable.
Software, and even the bugs in software, are no different. Obvsiously bugs are bad, but in the eyes of many so is not supporting a wide range of hardware or not quickly incorporating the newest technologies or costing more money, and those are all examples of priorities that compete against software assurances.
Security, flexibility, and ease of use also often compete against each other. One example is the practice of automatically opening ports in firewalls at the request of applications internal to the network. Doing so makes these programs easier to set up, but less secure. If your top priority is security then this is bad, if your top priority is ease of use, then making consumers fiddle with their firewalls is equally bad. You can maintain ease of use and some security if you have a centrally controlled white list of trustworthy applications stored in the firewall, but this is less flexible.
If you're arguing that the consumer software industry has their priorities out of order than that's fine, but the really important question is then "What should they be?" and as a software engineer it isn't my place to anwser that question (at least not alone). I can produce formally specified model-checked programs slowly and carefully or caffieen motivated sleep deprivation inspired spagetti code quickly, cheaply, and with no garuntees, or anything in between. If you're the customer, you get to choose. If you're not happy with the results of your choice, it isn't my problem unless you continue to pay me to fix your mistake. If you think I can write the former as quickly as the later, or the later as safely as the former, you're simply wrong.
Out of curiousity, what do you mean by "more directly?"
My phyiscs is a little rusty but, even the 'purest' reaction, particle anti-particle annialation, just results in different particles carrying the energy of that reaction (often photons, but not always). You can convert photons directly into electricity, but that's not very efficient, look at how inefficinet solar panels are (about 28% efficient for the really expensive kind). Modern steam powered generators have an efficiency about twice that.
To oversimplify, and torture metaphors, when your energy bearing photon crashes into another particle, there's a certain chance it'll motivate an electron, but a much greater chance it will be absorbed by the material as an increase in temperature. There isn't much you can do about this, the nucleus will always be a bigger target than the electrons. Once that happens the question becomes how to turn a temperature difference into electricity, and as it turns out transitioning through mechanical energy in the form of water and steam is a pretty efficient approach.
I guess I'm not sure what even the basics of a more direct approach would be, but I'm not the right kind of engineer for this so feel free to help me out by clarifying or correcting my explanation above, or explaining what a more direct approach might be like.
Government issued smartcards, with a simple PKI (and revocation system) would be a perfect method for establishing identity.
Before I accept this I would like to know how you would address several limitations inherent in such a system. There is no way people could be expected to remember thier key pair, even if they did they wouldn't be performing cryptographic hashes in their heads, so they would have to carry the card with them all the time. This increases the likelyhood that the card would be lost or stolen and the difficulty in getting by without it. Right now even if I forget my credit card it is still possible to do business as long as I remember the number, but under a PKI smart card system? "Who are you?" "Hell if I know, I left my name in my other pants!"
If the card was lost or stolen, or when they were first issued, what would be required as proof of identity to get a new one? Whatever is needed to obtain the card is nearly as valueable as having one, and usually easier to forge.
If a card is stolen, what other barriers prevents the thief from making use of the identity? For a revocation system to work with the highest degree of certainty, all applicable revocation lists need to be checked for every transaction, Doesn't this negate one of the primary advantages of public key crptyography and require that all authentication decisions be made 'online?'
The econemy in FFXI really isn't very player controlled. Lets look at some of the hard limits on player control of the econemy in FFXI. Many goods and services can only be obtained from NPCs, many goods can only be 'produced' once per character, many cannot be traded and NPC buy prices, quest rewards, and mob drops are almost all inflexible.
Now how about some things that you could technically do but that the game doesn't support? Such as comission something or in any other way illustrate the demand for something that's not currently being sold? There is no in-game support for buy orders.
What about large scale distribution without an army of mules? Why can we only have seven items on auction at a time per character? Why can we only have 8 outgoing packages at a time? This certainly isn't anyhere near appropriate to the production capacity of a crafter unless its sole purpose is to cripple him or her. Bazaars are a joke, fleamarkets have thier place, but often even those are more organized than walking up to random, unsorted vendors and assaulting them for a list of goods.
What about community ownership of property or money? Why can't linkshells be more than simple chat channels?
What about proper record keeping? The game goes out of its way to prevent this by not only having no logging to speak of, the prices for the things you buy arn't even displayed in the log, but by preventing alt-tabbing so that if you want to keep records you can either get another computer, use pen and paper, or cheat.
Probably the biggest limitation on the FFXI econemy is that the main advancement producing activity of the game, killing mobs to gain EXP, results in such small amounts of gil as to be mostly non-monetary, and such random and unreliable loot compared to farming that it is basically useless for dependable industry. Gil isn't a sign of advancement, it is an advancement stopping technique.
Define normalizing? Falling prices cut both ways, and you get less of a profit on the things you sell to earn gil towards that now-cheaper thing you want to buy.
This is complicated by the fact that it is basically impossible to make a profit by selling things to NPCs, so gil is streaming out of the econemy (NPCs still sell useful goods and services) faster than it is coming in.
The only question that really matters is how much playtime a player has to spend farming/crafting/etc to get a particular item, and in my case, deflation has not improved those numbers, and sometimes made them worse.
There are many, many problems with FFXI and the econemy would be one of them even if gillsellers had never existed.
The grandparent of my post, and my own post, are concerned only with charging users a fee for the development of that software, so I assume that the parent of my post also applies to this. My own post said that the GPL probably does not allow this even though it does allow for a distribution fee, and that even if it did allow for such an inflated distribution fee as to reflect the development cost there is no mechanism for the author to enforce the payment of the fee more than once. In this sense our comments are identical?
It is true that someone can charge to distribute copies of open source programs but to pretain to my post and the grandparent, please explain how a software author may ensure that they make money charging for binaries wich reflects their cost for developing their software and prefferably is based on the number of users?
If you'd ever read the GPL, you'd notice that source only needs to distributed to the people who got the binary, and the binary can be charged for.
I'm looking at the GPL right now and that does not appear to be true. Under "Terms and conditions for copying, distribution and modification" point 1 the second paragraph reads as follows:
You may charge a fee for the physical act of transferring a copy, and you may at your option offer warranty protection in exchange for a fee.
That seems to imply that you cannot charge a fee for the binary (and accompanying source code) that reflects your time to develop the software.
Even if this were not so, because the source code must acompany any binary distribution or be available to all third parties under point 3 of the same heading it would be difficult for the author to charge for more than the initial copy of the program unless those who already had it were unwilling to redistribute it themselves for some reason not related to the liscence. It would be virtually impossible for the creator to charge for more than the initial copy of any program intended for consumer use since consumers have no incentive not to copy amongst themselves.
I am not a lawyer, so I admit that my understanding may be flawed, please clarify your position.
Okay, first of all, the way you're calculating it, the proposed national retail sales tax tax rate isn't 23%, it is 30%. The 23% rate they give is what they call 'tax inclusive.' For example if the sales tax is 5% then a $1.00 item would have a cost with tax of $1.05 but the 'tax inclusive' rate is only 4.8% because $0.05 is only about 4.8% of $1.05. See this question in the fairtax.org FAQ.
Next, food and medical expenses are taxed under the reasoning that rich people spend much more on their neccecities than poor people do. See this question in the fairtax.org FAQ.
Instead of exempting items by category, every family in the country gets a check in the mail equal to the amount of taxes they will pay on spending up to the poverty level.
Lastly, the point I think the original post with regards to overhead cost (mom and pop vs Walmart) was making was that the effort involved in managing the collection of taxes is proporitonally greater for a mom-and-pop business with few transactions than for a large company like Walmart with many transactions. I'm not sure I believe that argument either, but I have other problems with this proposed national sales tax.
What you're talking about is user authentication using nonces and it has its uses, but it has a host of problems in a situation like this as well. Many people would like to do online banking or what have you on more than one computer. People have a hard enough time remembering passwords and are not likely to remember a cryptographic key. How are they going to transfer their credentials between machines?
There are ways to generate keys from passwords, but if these were used, you'd probably type your passowrd into a web page which would locally hash it up and use that as the key to process the nonce. But wait, if you type your password into a webpage, you're vulnerable to fishing again!
You could have someone use a key fob with their key in it, either based on the time or where they enter your nonce on the fob and type in what the fob shows them. But then many people would have multiple fobs, and you have to deal with distributing and maintaining them, and what hapens if they're stolen?
There are a couple of problems with the prize approach, first is the speculative nature of research and the associated upfront costs. The entire point of research is that people don't already know what the answer is or how long it is going to take to find it. Even taking on something with an apparently straightforward goal such as solar panel efficiency might need a mechanical and electrical engineer and a materials scientist just off the top of my head. Someone would have to supply tools and materials for an unknown period of time, absorbing the cost of their many inevitable mistakes, revisions, and dead-ends all without money coming in or with reliable estimates of when the payoff will be, in the hope that some eventual payoff will exceed their cost, but someone else could always beat them to it, leaving them nothing but a load of debt.
Speaking of beating them to it, this system hardly encourages cooperation. Science and engineering are not one-person jobs, as your statement on rewards would seem to suggest. There are simply too many facets to complex, real world problems to expect many individuals to be professionally competent in all of them. Even when there are such individuals, cooperation and free exchange are key to success in basic research. One of the founding principles of academic journals is that discussion between researchers can help spot problems and generate new ideas. A spirit of friendly (and even not so friendly, as in publish-or-perish) competition is one thing, but that's a far cry from the fear that if one researcher helps another the helpful person may be risking not just future success, but present well being. If you want evidence that hypercompetitive secrecy isn't good for research take a look at the soviet scientific establishment, particularly the soviet atomic bomb project.
Even if you did find some way to fairly distribute rewards among the people involved, the problem with prizes of the magnitude you're discussing, that would set the recipients up for life is that they remove the most successful people from the talent pool. If a recipient is set for life, why continue working in a competitive environment? Even if recipients continue working for the love of their field, why would anyone else consent to work with them? The resentment towards recipients could be huge, and even if it wasn't they couldn't be said to be on a level playing field with non-recipients. Two people in need might help each other at a risk to themselves to improve the odds that at least one of them will succeed, but why would they help someone who, from their perspective, has no need?
Let's say you write a book.... You go to a publisher, and say "I will sell you the rights to my book, if you give me 50% of the profits it makes." They agree.
So, lots of software developers get paid a percentage of sales? While companies profit directly from the number of copies sold, to the best of my knowledge most software developers do not. They receive continued pay only for continued work regardless of how much their employer is coasting on the developer's past successes.
Of course there are logical reasons to pay developers this way, if you allow a developer to coast forever on a single success then you've created a system where the most successful people are constantly being eliminated from the talent pool by that success. The ever-less-talented remainder then floods the market with drivel in a constant search for the next "big score."
The same logic applies to companies, but even moreso. Why should the sale of a product give this company the right to extract indefinite obedience from the customer (apparently on pain of having that product taken away)? How does that motivate the company to continue serving the public in the form of the customer in the best way it can? If it does not then why should it be allowed?
Contracts exist for mutual benefit only, so if your obligation to me is over then my obligation to you must also be over. Try to fit into an agreement of the form "If you stop X then I'll stop Y" here are some good examples: "If you stop working for me, I'll stop paying you (a job)." "If you stop following my instructions I'll stop trying to help you when they don't work (a warranty)." "If you stop returning the things I loan you in good condition, I'll stop loaning you things."
Okay, you're right my parenthetical had is backwards, because it was late and I was tired, however you're misinterpreting the rest of my comment. I never claimed that you could tell if an acceleration 'increased or decreased' velocity in any absolute sense, I was only using the earth shuttle system as an example. From my perspective as an engineer there is no such thing as 'deccleeration' only acceleration in a different vector, it just happened that in the example I chose both parties would intuit the same signs.
The reason I mentioned accleeration at all was because it's the presence of acceleration that causes a ship traveling at relativisitc speeds relative to a planet to experience 'time dilation' while the planet appears no to. During the ship's cruising phase people on the planet would say the ship's clocks are running slow and the people on the ship would say the planet's clocks are running slow, however the ship experiences accelerations to match velocities with the planet so it can orbit/land/whatever and it is the acceleration phases that sort out the 'twin problem' type relativity effects.
What I was trying to say (but did not) was that, in a sense when 'traveling at near light speed' is used it can someitmes be interpreted as 'relative to myself before I underwent some kind of massive acceleration.' or 'relative to the frame of the planet I started on'
The really, really short anwser: You are right, velocity is always measured as the relative velocity between two objects; however, all observers who are not accelerating always think that light travels the speed of light faster than them in all directions and will never observe anything traveling toward or away from themselves faster than the speed of light as they measure it. As far as they are concerned, they are sitting still.
Acceleration breaks this assumption though, when the shuttle lifts off we don't all experience crushing G forces, the shuttle accelerated and the earth did not, and we both agree on this.
Another way of saying what the article is trying to say accounting for the assumption that the observer is sitting still is that massive objects traveling at significant fractions of the speed of light relative to the observer will tend to push the observer out of the way. I havn't scoured the actual article but that's the basic idea, anyway.
(Note this comes mostly from special relativity and the article involves general relativity but as far as I know general relativity is a subset of secial relativity and it's basic principles still hold)
Thanks, I did, in fact RTA you provided. You did anwser my request for an exmaple of a situation where paralellism beats a single, faster chip, and you're right paralell processing does help accomodate the memory bottleneck. I'll consider multi-core paralellism a significant advance when hardware and compilers have developed to the point that, at the very least, threading doesn't change the fundamental contract between a developer and their hardware about the way their code is going to at least appear to behave.
Right now, it does. Even an experienced paralell programmer is in most instances just tries to separate functionality into independant threads as much as possible and throw locks around where needed to try and get the program to behave in anything resembling a predictable way. That's far from ideal, and what's worse, automated tools to locate the points where a threaded programmer may have made a mistake, where paralellism can sneak in and break the programmer's expectations of code, making it inherently unpredictable, don't exist in a form useful for realistic systems. Current tools in common use for Java (the only widespread language with built-in paralell processing suport) top out at only a few hundred lines of code, and due to the fact that possible paths of execution increase exponentially with each additional line, they won't be getting better any time soon.
The lack of automated testing and analysis tools just underscores that no one has yet developed a simple, repeatable method of producing efficient, safe, predictable multi-threaded programs. That's not engineering, that's art, and it's unpredictability make it a black art, at that.
I look forward to the day when compiler technology takes advantage of paralellism in a way that can efficiently paralellize arbitrary code that was written in an implementation agnostic way, or at least automatically locate and flag potential flaws, but as the article also pointed out, Java is the only language with built-in concurrenty support and its not particlalry good, let alone automated.
Concurrent programming isn't mature. Widespread proliferation of multi-core chips is going to force this technology on rank-and-file developers who are going to look to the eggheads for tools and a mehodology to use them, and the eggheads are going to have to say "we don't have it."
These kinds of chips don't represent a move towards becoming multithreaded, they represent a desire to be multithreaded. What was that you said about magic again?
The short version: What I'm really arguing here is multi-core single processors are not fundamentally sperior to simply having two chips working in paralell, and that two chips or cores working in paralell are never as good as a single chip that simply goes twice as fast. Having multiple chips in a single embedded device only goes to support that argument. If you can think of a case where two identical processors working together can solve a problem faster than a single chip that simply goes twice as fast, please tell me.
Okay, first lets get our definitions straight. Lots of things have multiple chips on them, and yes, that's very important, but it isn't paralellism in the sense being discussed unless the chips work simultaneously, not sequentially, and nearly interchangably. The presence of a DSP chip and a more general chip to handle menus and so forth in a cell phone (as one entirely hypothetical example) does not denote paralell computing because they're not, fundamentally, working on the same problem, nor can they easily switch roles. In the sense being discussed here, your CPU isn't paralell processing with your hard drive controller, or even your FPU, because they're specialized parts of a whole, not replicaitons of a funcitonal unit.
Further, threading it neccecary for pralellism, but doesn't inherently mean that an application will benifit very much from it. The overhead associated with interprocessor communication means that the increase in speed you get with additonal processors is always at least slightly diluted, and due to threading management issues, you often end up waiting on other processors/cores. It may get you the anwser to pressing quesitons faster, but only at the cost of lower utilization, which pretty much equates to wasted chip space.
Finally, none of that is even what is being discussed here. The issue is paralellism-on-a-chip, or multi-core processing. You havn't provided a single example where a dual core chip is prefferable to a single core chip that simply goes twice as fast. As I said, there are overheads associated with multiprocessing even in an ideal situation (which you almost never, if ever, have) so multi-core paralellism can never even be as good as a single core chip that goes twice as fast.
I realize there are significant technological hurdles to making faster and faster single core chips, but as you said yourself, existing embedded hardware such as your cell phone can already accomodate multiple chips anyway. So far as I know multi-core chips are simply a sort of appology. "We can't make the underlying chips faster, so here, we'll incorproate the work-around most people would use anyway upfront."
I stand by my original assertion (which I think has been lost) multi core processors do not represent any sort of fundamental advance. That is all.
Slashdot Mirror
From that point of view, graduate school is getting a PhD. Masters and professional students serve two purposes: fund raising and an outlet for failed/burned out Ph.D. students.
Maybe it is different in engineering, but the vast majority of my professors have separate masters and PhDs. When I was investigating graduate school the expectation seemed to be that I would get a master's, then a apply to a doctorate program. Is that really so unusual?
6 to 9 months left on master's degree, looking for a project or thesis advisor, will commit unspeakable acts to obtain one with english proficiency. Would commit slightly unplesant ones to talk to professors who would just say what they wanted.
Just wanted to back this up, maybe it's different outside of engineering, but in engineering, even engineering acadamia, the actual engineer has basically no rights to anything they make. This applies equally to students and professors at the colleges I have been to.
If you're a research engineer then your compensation is that you get to publish a paper which is either the goal in itself, or you hope will lead to someone taking you more seriously (possibly paying you more, but not always) on another project. On the other hand, at least where I did my undergrad, it was expect that if you didn't research (you were primarily interested in teaching, for example) you instead worked on projects for the school, and remained as anonymous as any normal engineer.
I've got a few problems with the advertising driven business model. The first two are personal, the rest are more general concers I have with the advertising econemy.
The first is that most general advertising assumes I'm too stupid to know what I want, or to find information on products to meet those desires even if I do know what I want. Telemarketing is the most heinous example of this kind of marketing. I have received telemarketing calls asking if I wanted to change my long distance carrier on numbers with no long distance service. I get calls asking if I'd like a deal on home siding and I live in an apartment building. How is this not the worst possible waste?
The second is that in advertising products don't even pretend compete on merits. It is pretty insulting to be surrounded by base plays on my emotions all the time.
More generally, advertising seems like a bit of a pyramid scheme. Many things are financed almost entirely by advertisements and the money for those ads has to come from the consumer somewhere.
Most advertising casts a very wide net and a view results in a sale only a very small fraction of the time, and rarely in a traceable manner. This means the efficacy of any particular ad is always uncertain. Most advertising dollars are therefore wasted for many reasonable definitions of "waste."
Given that I don't see personally targeted ads on the sides of busses and subway cars, before movies, during television shows, or even on the internet happening any time soon doesn't htis mean there's an incentive to shape advertising sponsored materials towards those most likely to respond to the most generic advertising?
Doesn't this mean that I as a consumer have almost no way to directly sponsor a normally advertising supported endevor that I enjoy or find useful? If I don't respond to ads I'm useless to them.
Doesn't this mean that every time I buy something I'm indirectly sponsoring whatever that company spends it advertising money on? There is no garuntee that the product I buy and where the producers advertise will be even remotely related. Even assuming that it was always well known where a companie's advertising dollars were spent, I must now include that knowledge in my product selection, moving us still further from products competing on appropriate merits. Since it isn't always apparent what will my purchase will really be supporting, I also have to consider the time wasted in finding that information as well.
Most insultingly the fact that so much is funded nearly exclusivly through advertising means that advertisers feel that they're buying the eyeballs of consumers directly, and the content producers who are dependant on advertising income are inclined to agree with them. This gives rise to things like trying to stop consumers from fast-forwarding through commercials, but also to product placement as a more subtle but equally insideous plan to make advertising, which almost surely will be for something most of the people viewing the ad won't care about, harder to ignore.
So basically as far as I can see the advertising econemy is largely about further weakening the consumer's direct influence on the market and ability to make informed, intelligent decisions even if he or she wants to.
Well by consumer software I basically meant software whose sponsors arn't among its primary users, consumer, off the shelf software (COTS) is an acronym applied equally to a a consumer buying a video game and a business going out and buying a prefab RDBMS. That's immaterial though because my real point was that FOSS might (would) change the way software projects are selected.
The effect on small businesses with unusual needs might be bad, but it won't be any worse than it is now. If a small business has needs that can't be met by conventional consumer, off the shelf software packages and they can't afford to pay software engineers of their own, they're still screwed.
Many small to mediums sized businesses actually do have limited programming resources at their disposal. this is often either a consultant who works for and bills them on demand, or a single, cave-troll style programmer.
Which brings up another area of concern with 'scratch an itch' software of any kind, particularly if you're expecting it to be used by consumers under FOSS, the software process. All the things that go into software development that are not programming tasks. Requirements engineering, software architecture, verification and validation, documentation and training materials, those kinds of things. I'm not blind, I'm well aware that many commercial software projects lack good software process and that some F/OSS projects have pretty good process. "Scratch an itch" software doesn't tend to be among the projects with good process on either side, and when you intend for consumers to use it, that's a problem.
Support contracts would appear to be a big part of this.
From a consumer's perspective that is the problem. As a software engineer I work for whoever can pay me, so in that regard FOSS changes nothing. Given that, the next question is "Who can afford to pay me?" and that anwser also remains unchanged, businesses. Most consumers just can't afford to pay a development team to make or modify software to meet their needs, and now using a FOSS model there is no inscentive for a business to make consumer software since the software is only worth what the single highest bidder is willing to pay for it, and consumers usually motivate businesses through many people paying a lower price.
Perhaps there's a simple resolution to this that I'm not seeing, but until that solution is widely known, FOSS really isn't in the consumer's direct best interest. If you've got a solution, please let me know.
To expect safety after defeating barriers and interlocks is stupid for microwave ovens and toasters, let alone high energy robotic systems. To expect robots to be safe outside of their defined operating parameters is like expecting a car to be made of sponge so no matter how much you ignore the speed limit, you can't kill anyone.
I say the same sort of thing about software all the time, good luck getting people to understand that.
Even this service has quite a few flaws, here's what comes to mind while signing up for my 'free trial' an indie music catalog I can't search without giving them my credit card number, that's a huge red flag. Subscription service not a use based service, that's another. HTML newsletters, have another! Proprietary software, that shockingly requires an installer program on OS X yet another!
Look, I realize that in theory if you cancel before your trial period is up you don't get charged, and that you don't *have* to sign up for the newsletters or install their prorpiretary download manager, but in many ways this isn't any better than predatory CD clubs. The business model is based on the idea that some fraction of people won't use all their monthly downloads, effectivly wasting money, that most people will elect to install the software or sign up for at least one of the newsletters, thus permitting at least some spam. I'm really just not that enthused by business models based on customer stupidity, be it this or DRM.
First, I think you're mashing up several markets in ways that they aren't really related. You talk about business software liscencing as though they buy a liscence for every seat, for the majority of significant products that just isn't true and the pricing model looks a lot more like what you describe already. Businesses pay for a maximum number of concurrent users, and the liscences they have are deployed and recalled as needed from a central liscence server.
It is unlikely that many businesses would invest heavily in the other aspects of the model you describe exactly because "your desktop follows you wherever you go," meaning your files are stored on a server controlled by a third party and that they can easily be spread around by users without them even thinking about it. Since the business is likely to desire a high degree of control over what employees do with the business's files, as well as what applications their employees have access to, and since both businesses and users have an interest in preventing the comingling of business and personal files, (especially if they are charged for storage and bandwidth which they almost certainly will be) it is quite likely that even if this became the only way of getting MS software MS would still be double-dipping because the user would pay for their personal account and their employer would pay for a separate work account for them.
Further, if your files are stored on a Microsoft server as you imply, and there is no danger of saving them locally, as you say, (though this is doubtful) then what interest would Microsoft have in letting you have your files at all? Think about it, your monthly MS liscence fee could basically be a ransom on your data.
I find it highly unlikely that MS would liscence their 'software as a service' to Google in such a way that Google could provide it free, or more cheaply than Microsoft. Why not charge Google at least as much per user as you'd charge the users themselves?
So, discounting problems like "Where will the bandwith to support this model come from?" the basic hurdles here are, how will bandwith and storage be paid for? how would this giant pot-of-gold where user's files are stored be protected? how would the user be protected against malicious parties on route from those servers to (and including) the terminal? What if that kiosk you mentioned saved all the user data that passed through, or just the user's credentials? What kind of access control mechanisms would be put in place to allow businesses the kinds of control and separation of their files that they might need? and what garuntee do users and businesses have that MS won't just hold their data hostage?
My point was, even if what you say were true, (it's not exactly) then other things would have to be sacrificed to produce it (like time) and very few stakeholders are willing to pay that price. You could say the same thing about textbooks, nothing except the author prevents them from being flawless either, but many of them still have volumes of eratta. Legacy code support is a pretty good counter example to what you say. Many programs don't work if Windows is running in an unprivelaged mode. If they were originally developed before Windows even had such a mode, or before it worked the way it does now (and it is changing yet again in Vista) how the author have known about these requirements before they existed? Many aspects of Windows undergo such changes fairly regularly even within releases (Win2K SP2 was a great example of this). If I write a Java application that works fine on the JRE but doesn't work on on the third party alternatives, which one of us has the bug? If I write a web page that doesn't display the same way on all browsers, who has the bug? Now who is more likely to be asked to fix it?
Your statement that all problems that can be fixed can be avoided requires that there be some possible problem-free ideal, this simply isn't so. Some quality attributes exist in opposition to each other and you simply can't optimize for them all equally well. This is pretty clear when, for example, finding a home. Given a budget, desired location, and a desired size not all combinations of values are equally viable. Some are obviously absurd, you're almost surely not going to find a 3 bedroom house for $10 a month, others are attainable only if you sacrifice something else. You might be able to get a studio for $300 a month, but it is more likely in say, Milwaukee than in Manhattan. Size, price, and location all obviously affect consumers, even if the factors that shape the curve of viable values are not their concern. On the other hand this cuts both ways, if you're a consumer and you want a $10 a month 3 bedroom appartment in Manhattan, it isn't your real estate agent's problem, because it just isn't viable.
Software, and even the bugs in software, are no different. Obvsiously bugs are bad, but in the eyes of many so is not supporting a wide range of hardware or not quickly incorporating the newest technologies or costing more money, and those are all examples of priorities that compete against software assurances.
Security, flexibility, and ease of use also often compete against each other. One example is the practice of automatically opening ports in firewalls at the request of applications internal to the network. Doing so makes these programs easier to set up, but less secure. If your top priority is security then this is bad, if your top priority is ease of use, then making consumers fiddle with their firewalls is equally bad. You can maintain ease of use and some security if you have a centrally controlled white list of trustworthy applications stored in the firewall, but this is less flexible.
If you're arguing that the consumer software industry has their priorities out of order than that's fine, but the really important question is then "What should they be?" and as a software engineer it isn't my place to anwser that question (at least not alone). I can produce formally specified model-checked programs slowly and carefully or caffieen motivated sleep deprivation inspired spagetti code quickly, cheaply, and with no garuntees, or anything in between. If you're the customer, you get to choose. If you're not happy with the results of your choice, it isn't my problem unless you continue to pay me to fix your mistake. If you think I can write the former as quickly as the later, or the later as safely as the former, you're simply wrong.
Out of curiousity, what do you mean by "more directly?"
My phyiscs is a little rusty but, even the 'purest' reaction, particle anti-particle annialation, just results in different particles carrying the energy of that reaction (often photons, but not always). You can convert photons directly into electricity, but that's not very efficient, look at how inefficinet solar panels are (about 28% efficient for the really expensive kind). Modern steam powered generators have an efficiency about twice that.
To oversimplify, and torture metaphors, when your energy bearing photon crashes into another particle, there's a certain chance it'll motivate an electron, but a much greater chance it will be absorbed by the material as an increase in temperature. There isn't much you can do about this, the nucleus will always be a bigger target than the electrons. Once that happens the question becomes how to turn a temperature difference into electricity, and as it turns out transitioning through mechanical energy in the form of water and steam is a pretty efficient approach.
I guess I'm not sure what even the basics of a more direct approach would be, but I'm not the right kind of engineer for this so feel free to help me out by clarifying or correcting my explanation above, or explaining what a more direct approach might be like.
Government issued smartcards, with a simple PKI (and revocation system) would be a perfect method for establishing identity.
Before I accept this I would like to know how you would address several limitations inherent in such a system. There is no way people could be expected to remember thier key pair, even if they did they wouldn't be performing cryptographic hashes in their heads, so they would have to carry the card with them all the time. This increases the likelyhood that the card would be lost or stolen and the difficulty in getting by without it. Right now even if I forget my credit card it is still possible to do business as long as I remember the number, but under a PKI smart card system? "Who are you?" "Hell if I know, I left my name in my other pants!"
If the card was lost or stolen, or when they were first issued, what would be required as proof of identity to get a new one? Whatever is needed to obtain the card is nearly as valueable as having one, and usually easier to forge.
If a card is stolen, what other barriers prevents the thief from making use of the identity? For a revocation system to work with the highest degree of certainty, all applicable revocation lists need to be checked for every transaction, Doesn't this negate one of the primary advantages of public key crptyography and require that all authentication decisions be made 'online?'
The econemy in FFXI really isn't very player controlled. Lets look at some of the hard limits on player control of the econemy in FFXI. Many goods and services can only be obtained from NPCs, many goods can only be 'produced' once per character, many cannot be traded and NPC buy prices, quest rewards, and mob drops are almost all inflexible.
Now how about some things that you could technically do but that the game doesn't support? Such as comission something or in any other way illustrate the demand for something that's not currently being sold? There is no in-game support for buy orders.
What about large scale distribution without an army of mules? Why can we only have seven items on auction at a time per character? Why can we only have 8 outgoing packages at a time? This certainly isn't anyhere near appropriate to the production capacity of a crafter unless its sole purpose is to cripple him or her. Bazaars are a joke, fleamarkets have thier place, but often even those are more organized than walking up to random, unsorted vendors and assaulting them for a list of goods.
What about community ownership of property or money? Why can't linkshells be more than simple chat channels?
What about proper record keeping? The game goes out of its way to prevent this by not only having no logging to speak of, the prices for the things you buy arn't even displayed in the log, but by preventing alt-tabbing so that if you want to keep records you can either get another computer, use pen and paper, or cheat.
Probably the biggest limitation on the FFXI econemy is that the main advancement producing activity of the game, killing mobs to gain EXP, results in such small amounts of gil as to be mostly non-monetary, and such random and unreliable loot compared to farming that it is basically useless for dependable industry. Gil isn't a sign of advancement, it is an advancement stopping technique.
Define normalizing? Falling prices cut both ways, and you get less of a profit on the things you sell to earn gil towards that now-cheaper thing you want to buy.
This is complicated by the fact that it is basically impossible to make a profit by selling things to NPCs, so gil is streaming out of the econemy (NPCs still sell useful goods and services) faster than it is coming in.
The only question that really matters is how much playtime a player has to spend farming/crafting/etc to get a particular item, and in my case, deflation has not improved those numbers, and sometimes made them worse.
There are many, many problems with FFXI and the econemy would be one of them even if gillsellers had never existed.
The grandparent of my post, and my own post, are concerned only with charging users a fee for the development of that software, so I assume that the parent of my post also applies to this. My own post said that the GPL probably does not allow this even though it does allow for a distribution fee, and that even if it did allow for such an inflated distribution fee as to reflect the development cost there is no mechanism for the author to enforce the payment of the fee more than once. In this sense our comments are identical? It is true that someone can charge to distribute copies of open source programs but to pretain to my post and the grandparent, please explain how a software author may ensure that they make money charging for binaries wich reflects their cost for developing their software and prefferably is based on the number of users?
I'm looking at the GPL right now and that does not appear to be true. Under "Terms and conditions for copying, distribution and modification" point 1 the second paragraph reads as follows: That seems to imply that you cannot charge a fee for the binary (and accompanying source code) that reflects your time to develop the software.
Even if this were not so, because the source code must acompany any binary distribution or be available to all third parties under point 3 of the same heading it would be difficult for the author to charge for more than the initial copy of the program unless those who already had it were unwilling to redistribute it themselves for some reason not related to the liscence. It would be virtually impossible for the creator to charge for more than the initial copy of any program intended for consumer use since consumers have no incentive not to copy amongst themselves.
I am not a lawyer, so I admit that my understanding may be flawed, please clarify your position.
Okay, first of all, the way you're calculating it, the proposed national retail sales tax tax rate isn't 23%, it is 30%. The 23% rate they give is what they call 'tax inclusive.' For example if the sales tax is 5% then a $1.00 item would have a cost with tax of $1.05 but the 'tax inclusive' rate is only 4.8% because $0.05 is only about 4.8% of $1.05. See this question in the fairtax.org FAQ.
Next, food and medical expenses are taxed under the reasoning that rich people spend much more on their neccecities than poor people do. See this question in the fairtax.org FAQ.
Instead of exempting items by category, every family in the country gets a check in the mail equal to the amount of taxes they will pay on spending up to the poverty level.
Lastly, the point I think the original post with regards to overhead cost (mom and pop vs Walmart) was making was that the effort involved in managing the collection of taxes is proporitonally greater for a mom-and-pop business with few transactions than for a large company like Walmart with many transactions. I'm not sure I believe that argument either, but I have other problems with this proposed national sales tax.
What you're talking about is user authentication using nonces and it has its uses, but it has a host of problems in a situation like this as well. Many people would like to do online banking or what have you on more than one computer. People have a hard enough time remembering passwords and are not likely to remember a cryptographic key. How are they going to transfer their credentials between machines?
There are ways to generate keys from passwords, but if these were used, you'd probably type your passowrd into a web page which would locally hash it up and use that as the key to process the nonce. But wait, if you type your password into a webpage, you're vulnerable to fishing again!
You could have someone use a key fob with their key in it, either based on the time or where they enter your nonce on the fob and type in what the fob shows them. But then many people would have multiple fobs, and you have to deal with distributing and maintaining them, and what hapens if they're stolen?
There are a couple of problems with the prize approach, first is the speculative nature of research and the associated upfront costs. The entire point of research is that people don't already know what the answer is or how long it is going to take to find it. Even taking on something with an apparently straightforward goal such as solar panel efficiency might need a mechanical and electrical engineer and a materials scientist just off the top of my head. Someone would have to supply tools and materials for an unknown period of time, absorbing the cost of their many inevitable mistakes, revisions, and dead-ends all without money coming in or with reliable estimates of when the payoff will be, in the hope that some eventual payoff will exceed their cost, but someone else could always beat them to it, leaving them nothing but a load of debt.
Speaking of beating them to it, this system hardly encourages cooperation. Science and engineering are not one-person jobs, as your statement on rewards would seem to suggest. There are simply too many facets to complex, real world problems to expect many individuals to be professionally competent in all of them. Even when there are such individuals, cooperation and free exchange are key to success in basic research. One of the founding principles of academic journals is that discussion between researchers can help spot problems and generate new ideas. A spirit of friendly (and even not so friendly, as in publish-or-perish) competition is one thing, but that's a far cry from the fear that if one researcher helps another the helpful person may be risking not just future success, but present well being. If you want evidence that hypercompetitive secrecy isn't good for research take a look at the soviet scientific establishment, particularly the soviet atomic bomb project.
Even if you did find some way to fairly distribute rewards among the people involved, the problem with prizes of the magnitude you're discussing, that would set the recipients up for life is that they remove the most successful people from the talent pool. If a recipient is set for life, why continue working in a competitive environment? Even if recipients continue working for the love of their field, why would anyone else consent to work with them? The resentment towards recipients could be huge, and even if it wasn't they couldn't be said to be on a level playing field with non-recipients. Two people in need might help each other at a risk to themselves to improve the odds that at least one of them will succeed, but why would they help someone who, from their perspective, has no need?
Let's say you write a book. ... You go to a publisher, and say "I will sell you the rights to my book, if you give me 50% of the profits it makes." They agree.
So, lots of software developers get paid a percentage of sales? While companies profit directly from the number of copies sold, to the best of my knowledge most software developers do not. They receive continued pay only for continued work regardless of how much their employer is coasting on the developer's past successes.
Of course there are logical reasons to pay developers this way, if you allow a developer to coast forever on a single success then you've created a system where the most successful people are constantly being eliminated from the talent pool by that success. The ever-less-talented remainder then floods the market with drivel in a constant search for the next "big score."
The same logic applies to companies, but even moreso. Why should the sale of a product give this company the right to extract indefinite obedience from the customer (apparently on pain of having that product taken away)? How does that motivate the company to continue serving the public in the form of the customer in the best way it can? If it does not then why should it be allowed?
Contracts exist for mutual benefit only, so if your obligation to me is over then my obligation to you must also be over. Try to fit into an agreement of the form "If you stop X then I'll stop Y" here are some good examples: "If you stop working for me, I'll stop paying you (a job)." "If you stop following my instructions I'll stop trying to help you when they don't work (a warranty)." "If you stop returning the things I loan you in good condition, I'll stop loaning you things."
Okay, you're right my parenthetical had is backwards, because it was late and I was tired, however you're misinterpreting the rest of my comment. I never claimed that you could tell if an acceleration 'increased or decreased' velocity in any absolute sense, I was only using the earth shuttle system as an example. From my perspective as an engineer there is no such thing as 'deccleeration' only acceleration in a different vector, it just happened that in the example I chose both parties would intuit the same signs. The reason I mentioned accleeration at all was because it's the presence of acceleration that causes a ship traveling at relativisitc speeds relative to a planet to experience 'time dilation' while the planet appears no to. During the ship's cruising phase people on the planet would say the ship's clocks are running slow and the people on the ship would say the planet's clocks are running slow, however the ship experiences accelerations to match velocities with the planet so it can orbit/land/whatever and it is the acceleration phases that sort out the 'twin problem' type relativity effects. What I was trying to say (but did not) was that, in a sense when 'traveling at near light speed' is used it can someitmes be interpreted as 'relative to myself before I underwent some kind of massive acceleration.' or 'relative to the frame of the planet I started on'
The really, really short anwser:
You are right, velocity is always measured as the relative velocity between two objects; however, all observers who are not accelerating always think that light travels the speed of light faster than them in all directions and will never observe anything traveling toward or away from themselves faster than the speed of light as they measure it. As far as they are concerned, they are sitting still.
Acceleration breaks this assumption though, when the shuttle lifts off we don't all experience crushing G forces, the shuttle accelerated and the earth did not, and we both agree on this.
Another way of saying what the article is trying to say accounting for the assumption that the observer is sitting still is that massive objects traveling at significant fractions of the speed of light relative to the observer will tend to push the observer out of the way. I havn't scoured the actual article but that's the basic idea, anyway.
(Note this comes mostly from special relativity and the article involves general relativity but as far as I know general relativity is a subset of secial relativity and it's basic principles still hold)
For marketing purposes no one receives 'spam' anymore, now they receive 'supper surprise funmail!' it tests much better focus groups.
Right now, it does. Even an experienced paralell programmer is in most instances just tries to separate functionality into independant threads as much as possible and throw locks around where needed to try and get the program to behave in anything resembling a predictable way. That's far from ideal, and what's worse, automated tools to locate the points where a threaded programmer may have made a mistake, where paralellism can sneak in and break the programmer's expectations of code, making it inherently unpredictable, don't exist in a form useful for realistic systems. Current tools in common use for Java (the only widespread language with built-in paralell processing suport) top out at only a few hundred lines of code, and due to the fact that possible paths of execution increase exponentially with each additional line, they won't be getting better any time soon.
The lack of automated testing and analysis tools just underscores that no one has yet developed a simple, repeatable method of producing efficient, safe, predictable multi-threaded programs. That's not engineering, that's art, and it's unpredictability make it a black art, at that.
I look forward to the day when compiler technology takes advantage of paralellism in a way that can efficiently paralellize arbitrary code that was written in an implementation agnostic way, or at least automatically locate and flag potential flaws, but as the article also pointed out, Java is the only language with built-in concurrenty support and its not particlalry good, let alone automated.
Concurrent programming isn't mature. Widespread proliferation of multi-core chips is going to force this technology on rank-and-file developers who are going to look to the eggheads for tools and a mehodology to use them, and the eggheads are going to have to say "we don't have it."
These kinds of chips don't represent a move towards becoming multithreaded, they represent a desire to be multithreaded. What was that you said about magic again?
Okay, first lets get our definitions straight. Lots of things have multiple chips on them, and yes, that's very important, but it isn't paralellism in the sense being discussed unless the chips work simultaneously, not sequentially, and nearly interchangably. The presence of a DSP chip and a more general chip to handle menus and so forth in a cell phone (as one entirely hypothetical example) does not denote paralell computing because they're not, fundamentally, working on the same problem, nor can they easily switch roles. In the sense being discussed here, your CPU isn't paralell processing with your hard drive controller, or even your FPU, because they're specialized parts of a whole, not replicaitons of a funcitonal unit.
Further, threading it neccecary for pralellism, but doesn't inherently mean that an application will benifit very much from it. The overhead associated with interprocessor communication means that the increase in speed you get with additonal processors is always at least slightly diluted, and due to threading management issues, you often end up waiting on other processors/cores. It may get you the anwser to pressing quesitons faster, but only at the cost of lower utilization, which pretty much equates to wasted chip space.
Finally, none of that is even what is being discussed here. The issue is paralellism-on-a-chip, or multi-core processing. You havn't provided a single example where a dual core chip is prefferable to a single core chip that simply goes twice as fast. As I said, there are overheads associated with multiprocessing even in an ideal situation (which you almost never, if ever, have) so multi-core paralellism can never even be as good as a single core chip that goes twice as fast.
I realize there are significant technological hurdles to making faster and faster single core chips, but as you said yourself, existing embedded hardware such as your cell phone can already accomodate multiple chips anyway. So far as I know multi-core chips are simply a sort of appology. "We can't make the underlying chips faster, so here, we'll incorproate the work-around most people would use anyway upfront."
I stand by my original assertion (which I think has been lost) multi core processors do not represent any sort of fundamental advance. That is all.