First, the way you have phrased the question makes it sound like you expect FOSS organizations have lower standards for awarding scholarships (just becasue you can write some C code and might, someday, release a program under the GPL) than other scholarship awarding organizations. This is not a very impressive start. If you have the chops to get admitted to MIT (or other top tier schools) you should be able to effectively compete for scholarships by traditional means.
All that said, with a little bit of expertise in programming, you should easily be able to get a part time job that will suplement whatever tuition your parents are willing to fork out (a part-time junior programming job should easily net you $10,000-$20,000 per year, which is enough for full tuition, room and board at a good state school). In fact, if you can land a job based on your current programming skills, you might be able to get your employer to help out with (or pay for, outright) your college tuition.
All the folks that are saying that an MIT degree isn't worth any more than a state university degree are leading you astray, but not too far. Even a degree for a third rate state university will make a huge difference when you go to get a job. Lots of jobs won't even consider you without some degree. The difference that the MIT degree will make will be when (and if) you are competing for the best positions with the best companies. In fact, if you graduate from some place like MIT, you probably won't need to even interview for your first job after graduation: companies will be scouting for you. Still, you'll do quite well if you get a degree from one of the big state schools, and will cost you a hell of a lot less than MIT.
It's like proving something exists buy using something that doesn't exist.
Actually, you can prove that one thing exists by using another that doesn't exist, it's called proof by contradiction and it's used all the time. The basic method is that you assume the existance of the oposite of what you are trying to prove and then show that the assumption leads to some logical contradiction.
Now, the guy in TFA is not making a proof by contradiction, but you certainly can prove the existance of a real thing using a thing that doesn't (and can't) exist.
From the article's description, it sounds like they sped up the chronometer until the numbers were changing faster than the human persistence of vision is able to differentiate distinct images. All they showed with the experiment is that the fundamental biochemistry of the retina doesn't change under stress (that the rods and cones can't recover between sensations any more rapidly under stress than they do under unstressed circumstances). I would expect that, if our sense of time were to mean anything real (as opposed to just being a subjective illusion based on memory) then the speed of response to stimulii would decrease under stress or excitement, which should be easy to measure (did you ever take a braking response test in Driver's Ed. class? That's the kind of test they should have been doing). It is possible (even probable) that your overall response time decreases when you are stressed, because of all the extra neuro-transmitters in your system, which would certainly make things seem to happen subjectively more slowly.
This was a decent discussion on BoingBoing not too long ago.
"Is it true that it's 'not infringement once fair use kicks in' ?
Fair use is a defense to infringement where you admit infringement but say it was justified, isn't it? You affirm the boundaries of copyright but justify crossing them, rather than arguing that the boundaries should be moved. This is why it's argued on a case-by-case basis.
This article suggests some good reasons to move the boundaries, I think."
Not sure if that's right (IANAL), but it sure sounds like it to this lay-person.
While IANAL, I have taken business law, so I needed to learn some of this stuff:
Fair Use is an affirmative defense to an accusation of copyright infringement. I'm not certain, but I think an affirmative defense is more than simply an admission of guilt with an excuse. I think that an affirmative defense implies that, though the facts of the case may support the accuation ("I did make a copy of that copyrighted work"), you are asserting, as a matter of law, that you didn't violate the statute in question ("but my copy is allowed under the doctrine of fair use"). You are, in effect, claiming that no actual crime occurred, because you actions don't fall under the specific language of the statute (or are exempted by other specific language).
In any trial there are two broad groups of things at issue: issues of fact (what things actuall happened) and issues of law (how to interpret the things that happened). Fair Use is an issue of law, not of fact. When an accusation is made against you in a court of law, you may defend yourself in several ways: you can deny the facts of the case ("I never did the thing that I am accused of doing.", "I never made any copies of the copyrighted work.") and you may deny the illegality of your actions ("I did the deed, but it is allowed under the law for this reason.", "I did make a copy of the copyrighted work, but it is allowed under fair use for this reason."). You can even defend yourself on both the facts and the law ("I never did the deed, but If I had it wouldn't have been illegal under the law for this reason.").
I could sign a specially-worded contract putting me under legal obligation to learn how to breathe margarine and turn the moon into a Buick, but it doesn't mean I'll be able to in practice.
sorry, alreadydone. The Lisa II (or Lisa 2 or Mac XL) was the follow on to the original Lisa. It replaced the Lisa's dual 5.25-inch "Twiggy" floppies with a single 400KB 3.5-inch floppy and came bundled with a Macintosh emulation program (MacWorks XL: basically Mac ROM code adapted to run on the Lisa hardware). It was released around the same time as the original Mac and cost slightly more, but had a larger display, hard disk drive, expandable memory (up to 2MB) and an internal I/O expansion bus. The processor was slower (6MHz for the Lisa vs. 8MHz for the Macintosh) but there was a custom memory protection/memory management chip that was, unfortunately, never used in Macintosh mode.
It was one thing when the Singularity crazies were cooking up new half-bakery ("The Singularity is the point at which history becomes unpredictable," as if history has been so easy to predict up till now) but now they've sunk to digging up 40 year old half-bakery. Good's definition of ultra-intelligence relies on the unproven presumption that we can build normally-intelligent machines (which, if the last half-century of progress is any indicator, we can't). As soon as we build a machine that can match my cat on all intelligent activities, then I'll start fretting about the Singularity, until then it's all just hot air and navel-gazing.
100,000 write cycles is plenty, so long as you buffer the writes and limit their frequency. All you need to do is either put a big honking RAM writeback cache next to the FlashRAM, or enforce writeback caching in the OS. If you can get the write frequency down to about two writes per hour, and do good load leveling, your FlashRAM will last for 5 years, which is about as good as most consumer grade hard disks (and possibly better, since the 'expired' FlashRAM drive could still be perfectly readable). Two writes per hour may sound like a very aggressive goal, but it wouldn't be so hard if you preferred to evict clean pages from cache before dirty pages (which is not so bad, since the read latency of FlashRAM is not nearly as long as that for a spinning platter, so refilling evicted pages is not too expensive).
Now, think of it in a new way. Suppose you were a civilization that just developed space travel, much like where we are now. You have a galaxy around you with 400 billion stars, and that's a lot. It takes you 100,000 years at light speed to cross the galaxy, and that's a long time. However, you have 2 billion years to explore.
But you (or any member of your species) don't have 2 billion years to explore, you only have 50 years or so (maybe more, but probably not more than one or two orders or magnitude, so 5000 years at the upper limit). In that time you can, if your technology is really good and you have (literally) tons of energy to burn, get to the nearest starts. Most of those stars will be uninteresting red dwarfs, with little likelihood of supporting life on their planets. Even if you used unmanned "self-replicating" probes, nobody that built or launched the probes would be alive when they reach their destinations: it's awfully hard to keep a project running across tens of lifetimes.
I have no good grasp on where humans will be 2 billion years from now, but I am sure we will be pretty advanced.
My guess, based on historical performance, is that we will be thoroughly extinct by then. Two billion years, after all, is a very long time.
The best answer to the Fermi Paradox, in my opinion, is either that interstellar travel just isn't worth it (too slow and expensive for no particular gain), or there is some pheonomena, of which we are currently unaware, that prevents it (my bet is that the interstellar medium, outside of a strong magnetic field like that produced by a star, rapidly destroys solid objects).
"when consumers realize their personal lust for faster hardware makes almost zero financial sense, and hurts the environment with greater demands for power, will they start to demand cheaper, more efficient 'third-world' computers that are just as effective?" Will ridiculously cheap laptops wean consumers off ridiculously fast components?"
Maybe these same consumers will also realize that Moore's law also means that in 18 months you will be able to do the same computational work at roughly half the power cost (modulo leakage current, of course), a fact that appears to escape the razor wits at CNet.UK!
Moore's law is the only reason that we now have $5.00 calculators running off of solar cells generating a few miliwatts from ambient light, or $10.00 quartz wrist watches that run for years off a single button cell. If anything, the $100 laptop will accellerate Moore's law by increasing the volume of products produced and resultant economies of scale.
The folks at CNet.UK are a bunch of clueless wankers.
How, I wonder, does he intend to hold this mobile device if all five fingers are involved in pressing the buttons? He also doesn't seems to think that you can combine a coding pattern space with the inactive state of the buttons (when none are being pressed). I'd say he hasn't put any thought at all into this mechanism (just as he hasn't done any research for prior work: e.g. chording keyboards). He clearly hasn't build even a non-working mechanical prototype to see if any of this is usable (take a pack of cards, draw some buttons on the box and try using the resulting chording device. For extra credit, do it with just the deck of cards without the box).
Here are some of the the problems:
You need to be able to hold the device, which robs the thumb and at least one finger from being able to operate the keys.
You can't encode a symbol on the keys when none are active, so that takes away one code point.
You need to allow for delay between synchronized keypresses (not all fingers will depress the keys simultaneously) so this will limit your typing speed.
You need to learn the damn chords, which most people are pretty bad at (heck, they're usually pretty bad at learning to type non-chorded as well).
Let's see, as I recall the history goes something like this:
1985 Windows 1.0
1987 Windows 2.0/286
1990 Windows 3.0
1992 Windows 3.1
1993 Windows for Workgroups 3.11 and Windows NT 3.1
1994 Windows NT 3.51
1995 Windows 95 (Windows 4.0, by my count)
1996 Windows NT 4.0
1998 Windows 98 (Windows 5.0, by my count)
2000 Windows 2000 (Windows NT 5.0, by my count) and Windows Me (Windows 6.0, by my count)
2001 Windows XP (Windows NT 6.0, by my count)
2007 Windows Vista (Windows 7.0, by my count)
That means that the next version should be, at least, Windows 8 or 9. Sure, maybe I shouldn't count Win98 and WinMe as full version increments, which would put as at just about 7.0 for the next version, but with all the other funny stuff MS has done with the Windows version numbers, why should I cut them any slack? If I were really a stickler for version numbers, I'd say that Windows NT should be at 5.0 as of Vista, which means that the next version is 6. Let's not even consider what the current DOS version number should be.
In the mainstream, this can easily get spun as the iPhone is extremely insecure, and has been "broken into", causing normal people to steer very clear.
Of course, because, as we all know, nobody would buy any product that was full of security holes: That's why Microsoft is having such a difficult time establishing themselves as a credible alternative on the desktop.
The problem with colonizing other planets, even within our own solar system, starts with the simple facts of distance and energy. The distances and energies involved with colonizing the continents of the earth were, pretty much, always within human ken. This is born out by the fact that, whenever europeans 'discovered' a new land, they found people already living there. Even a lone human, travelling on foot at normal walking speed, could circumnavigate the entire planet (given suitable land or ice briges) in a little under 2 years.
By contrast, humanity has only, in the last 50 years, even come close to controlling the amount of energy necessary to cross the gulfs between planets within our solar system, much less what is needed to travel to the nearest star. Anyone who compares the task of colonizing other planets to the european colonization of the new world, or the U.S. expansion into the west, is displaying the most profound ignorance imaginable.
The energy involved is important because it directly relates to the cost of the endeavor. The cost of colonizing distant continents was always within human grasp, so it is no surprise that it was done. The cost of travelling to other planets, however, is just barely within the grasp of the wealthiest nations, and there is no good reason to expect it ever to decrease very much.
The Fermi paradox has been used to imply that there is no intelligent life, other than us, in the galaxy, but there is another, perfectly good interpretation: maybe, even though it is possible to travel between the stars, it's just not economical to do so. Maybe the galaxy is full of intelligent life: life so intelligent, in fact, that it has long since given up the romantic, but entirely impractical, notion of interstellar travel.
I don't think that it is impossible to travel between the stars. In fact, I think that it is, basically, within human grasp at this very moment. I just think it is too expensive and dangerous to be undertaken by any nation (or similarly wealthy organized group) at this time. Give it a couple hundred years -- time enough to get the whole long-term-artificial-habitat thing, the safely-manage-tera-watt-power-generation thing, and the protect-ourselves-from-the-interstellar-medium thing down -- and I think it may be an option. At the moment, the best we could hope for would be unmanned probes to nearby stars. Even then, I doubt the transit time would be less than a half century.
Mr. Elgan has some definition of winning here that, in all probability, doesn't coincide with Apple's definition. Since neither we no Mr. Elgan knows what Apple's definition is, let's consider some of the possibilities:
1) Apple is trying to increase the market share of Safari to make it a more attrictive platform for web-developers to tailor thier websites for. In this case, if we consider the personal computing market as a whole, Apple only needs to get about 5% of Windows users to use Safari to double the Safari marketshare. Doubling marketshare is pretty big win for the kind of minimal investment represented by Safari on Windows.
2) Apple is using the Safari codebase as a trial baloon for some other, larger, Windows targeted software (maybe a Windows target in XCode?). Again, only a small penetration into the Windows userbase represents success for the effort.
3) Apple is delivering a development/test environment for iPhone to a huge pool of potnential developers. In this case, we would not expect most of the developers to use Safari as their regular browser, but they would use it when developing web pages for either iPhone or Mac clients. Again, total markter penetration (in this case, the market of web developers using Windows) doesn't need to be very large to deliver a huge benefit to Apple (greatly increase the number of people developing iPhone and Mac capable web apps).
Winning, at least for Apple, doesn't need to mean kicking MS off the web browser hill, in the same way that winning in the personal computer market hasn't, aparantly, had to mean capturing more than 5%-10% of the market. Apple seems to be doing very well with their minority share of the personal computer and laptop markets (admittedly, their share occupies the higher end of the price range, which doesn't hurt). Apple doesn't seem to think about these things as zero sum, as many other people seem to.
Yes, Roland used to spam Slashdot with articles linked to his own web page, which was reason enough to hate him. Thankfully, his posts now link to the actual articles, but that doesn't really help matters all that much. Unfortunately, Roland never seems to understand the point of the articles to which he links. He always writes sesantionalist headlines and summaries that have almost nothing to do with the actual article he references. He picks out specific buzwords from the article, buzwords that excite the most crackpottish centers of the brain, then constructs a boilerplate submission based on those buzwords.
Is this really a good reason to hate Roland? probably not. It's not like anyone is forcing you to read his drivel, you can always ignore his articles. At least he isn't pushing his own crackpottery with his articles (ala Time Cube or Mentifex). Still, he gets under the skin.
The main reason that 'parallel' programming is hard is that current tools support it only as an afterthought. All mainstream tools (languages, compilers, libraries, debuggers, etc.) are built, primarily, to support single-threaded code. It is up to the individual programmer to write code that is multi-thread safe by enclosing dangerous sections of code in mutexes, erecting semephores to coordinate multiple threads, and so on. Even languages with explicit support for multi-threading (e.g. Java) only support it as the exception: all code is presumed to be inteded for a single-threaded environment unless the programmer explicitly identifies it as multi-threaded (e.g. 'synchronized').
A better solution would be to assume that all code is run in a multi-threaded environment and have the tools build the code for mutual exclusion, unless explicitly told otherwise by the programmer. The language and libraries could also provide basic utilities and data structures in thread-safe implementations by default, requiring the programmer to specifically select a non-thread-safe version when desired. Monitors (thread-safe objects) were invented decades ago, but are still not in common use in modern languages.
The fundamental problem is that people are really, really bad at reasoning about thread-safety and figuring out where mutually exclusive sections of code and race conditions are. Relying on the programmer to get such things right is a guarantee of failure. The tools (languaes, compilers, etc.) should be able to determine when methods/functions access the same common state and automatically insert mutexes/semephores to enforce mutual exclusion, or at least assume that, unless told otherwise, all methods/functions must have mutexes/semephores. You'd still have instances where a programmer thought, incorrectly, that some code could be run with a mutex, but it would be the exception rather than the rule.
Now all we need is to identify some 230-some Microsoft patents that could be construed to be infringed by specific OSS packages mentioned by MS executives, and we have ourselves a field-day invalidating MS patents. Even if only a few of the cases succeed (I'm assuming one declaratory jundgement suit per patent) it would be pretty damaging to Microsoft to 1) lose some of it's patents, 2) have to mount the separate defenses in court and 3) deal with the resulting PR storm.
The initial research phase is exactly the sort of thing the OSS community should be good at: a distributed analysis of technical documents. MS has, admittedly, a lot of patents, but they have given us a number of clues to work from. It really shouldn't be all that hard to mobilize a small army of OSS-sympathetic law students/junior lawyers/paralegals to comb through MS's patent portfolio along with the list of allegedly infringing OSS projects and develop a candidate patent list.
The lawsuits may be a bit more trouble: you would need parties with pretty deep pockets to initiate and persue each lawsuit, who also happened to be either large scale users of OSS, distributors of OSS or the actual developers of the possibly infringing OSS projects.
A side benefit of the entire effort would be to develop a list of 'bugs' in OSS which, when fixed, would immunize the OSS community against further MS FUD.
Isn't it be possible to get some sort of declaratory judgement from a court? Say you're RedHat (or any other Linux distributor), who happens to sell Linux and related services - in light of MS's statements, wouldn't you be entitled to know which patents are involved? MS's statements have a direct impact on your business.
And if MS refused to tell you then couldn't you get a declaration from a court that your product doesn't infringe? IIRC, this is similar to what RedHat is pursuing in its case against SCO (which is on hold while SCO v IBM drags on).
Maybe a small Linux distributor with no assets and not much to lose could pursue a case like this against MS.
First, any party bringing a lawsuit under the American legal system must have standing to sue, meaning that they must have a material interest in the outcome of the suit. Since Microsoft has not specifically threatened anybody, at the moment, it would be hard to establish standing. A really good lawyer might be able to argue that a Linux vendor is harmed by the implication that Microsoft will sue that vendors customers for patent infringement, but with any actual patent suits in process, it would be a hard sell.
Second, the American legal system refuses to issue advisory judgements, and requires that a case be 'ripe' before it can be adjudicated. Since the court would be ruling on a hypothetical ("if Microsoft were to sue for patent infringement, would we be found to be infringing?") the court would (and should) simply refuse to hear the case.
Here is a good reference for standing, advisory opinions and ripeness. A little google-foo should easily turn up others.
The only bright spot in this, from a potential plaintif's point of view, is that, as a convicted monopolist, there might be a way to accuse Microsoft of restrain-of-trade, or some other violation of the Sherman Act. Unfortunately, I think that prosecution of anti-trust cases must be brought by the federal government, and that is not very likely with the current administration. Private actions can be brought for violations of the Clayton Act but I don't quite see how it could aply in this case, and only consumers injured by their dealings with the violator have standing to sue, which puts most direct competitors out of the running.
Disclaimer: IANALBIHTBL (IANAL But I Have Taken Business Law)
Slashdot Mirror
First, the way you have phrased the question makes it sound like you expect FOSS organizations have lower standards for awarding scholarships (just becasue you can write some C code and might, someday, release a program under the GPL) than other scholarship awarding organizations. This is not a very impressive start. If you have the chops to get admitted to MIT (or other top tier schools) you should be able to effectively compete for scholarships by traditional means.
All that said, with a little bit of expertise in programming, you should easily be able to get a part time job that will suplement whatever tuition your parents are willing to fork out (a part-time junior programming job should easily net you $10,000-$20,000 per year, which is enough for full tuition, room and board at a good state school). In fact, if you can land a job based on your current programming skills, you might be able to get your employer to help out with (or pay for, outright) your college tuition.
All the folks that are saying that an MIT degree isn't worth any more than a state university degree are leading you astray, but not too far. Even a degree for a third rate state university will make a huge difference when you go to get a job. Lots of jobs won't even consider you without some degree. The difference that the MIT degree will make will be when (and if) you are competing for the best positions with the best companies. In fact, if you graduate from some place like MIT, you probably won't need to even interview for your first job after graduation: companies will be scouting for you. Still, you'll do quite well if you get a degree from one of the big state schools, and will cost you a hell of a lot less than MIT.
<voice age=old disposition=grumpy politics=conservative>
They should take a shower and get a job! Damn hippies: GET OFF MY LAWN!
</voice>
Actually, you can prove that one thing exists by using another that doesn't exist, it's called proof by contradiction and it's used all the time. The basic method is that you assume the existance of the oposite of what you are trying to prove and then show that the assumption leads to some logical contradiction.
Now, the guy in TFA is not making a proof by contradiction, but you certainly can prove the existance of a real thing using a thing that doesn't (and can't) exist.
shouldn't that be: "What have I got in my pocket, Mother F*cker!?"
From the article's description, it sounds like they sped up the chronometer until the numbers were changing faster than the human persistence of vision is able to differentiate distinct images. All they showed with the experiment is that the fundamental biochemistry of the retina doesn't change under stress (that the rods and cones can't recover between sensations any more rapidly under stress than they do under unstressed circumstances). I would expect that, if our sense of time were to mean anything real (as opposed to just being a subjective illusion based on memory) then the speed of response to stimulii would decrease under stress or excitement, which should be easy to measure (did you ever take a braking response test in Driver's Ed. class? That's the kind of test they should have been doing). It is possible (even probable) that your overall response time decreases when you are stressed, because of all the extra neuro-transmitters in your system, which would certainly make things seem to happen subjectively more slowly.
While IANAL, I have taken business law, so I needed to learn some of this stuff:
Fair Use is an affirmative defense to an accusation of copyright infringement. I'm not certain, but I think an affirmative defense is more than simply an admission of guilt with an excuse. I think that an affirmative defense implies that, though the facts of the case may support the accuation ("I did make a copy of that copyrighted work"), you are asserting, as a matter of law, that you didn't violate the statute in question ("but my copy is allowed under the doctrine of fair use"). You are, in effect, claiming that no actual crime occurred, because you actions don't fall under the specific language of the statute (or are exempted by other specific language).
In any trial there are two broad groups of things at issue: issues of fact (what things actuall happened) and issues of law (how to interpret the things that happened). Fair Use is an issue of law, not of fact. When an accusation is made against you in a court of law, you may defend yourself in several ways: you can deny the facts of the case ("I never did the thing that I am accused of doing.", "I never made any copies of the copyrighted work.") and you may deny the illegality of your actions ("I did the deed, but it is allowed under the law for this reason.", "I did make a copy of the copyrighted work, but it is allowed under fair use for this reason."). You can even defend yourself on both the facts and the law ("I never did the deed, but If I had it wouldn't have been illegal under the law for this reason.").
Atzanteol replied:
Actually, I think that contract terms that are impossible or nonsensical can be easily voided: the term of art is mistake.
sorry, already done. The Lisa II (or Lisa 2 or Mac XL) was the follow on to the original Lisa. It replaced the Lisa's dual 5.25-inch "Twiggy" floppies with a single 400KB 3.5-inch floppy and came bundled with a Macintosh emulation program (MacWorks XL: basically Mac ROM code adapted to run on the Lisa hardware). It was released around the same time as the original Mac and cost slightly more, but had a larger display, hard disk drive, expandable memory (up to 2MB) and an internal I/O expansion bus. The processor was slower (6MHz for the Lisa vs. 8MHz for the Macintosh) but there was a custom memory protection/memory management chip that was, unfortunately, never used in Macintosh mode.
What you're looking for is the (dreaded) Lisa
That should be Registrar of Copyrights, not Register of Copyrights.
It was one thing when the Singularity crazies were cooking up new half-bakery ("The Singularity is the point at which history becomes unpredictable," as if history has been so easy to predict up till now) but now they've sunk to digging up 40 year old half-bakery. Good's definition of ultra-intelligence relies on the unproven presumption that we can build normally-intelligent machines (which, if the last half-century of progress is any indicator, we can't). As soon as we build a machine that can match my cat on all intelligent activities, then I'll start fretting about the Singularity, until then it's all just hot air and navel-gazing.
100,000 write cycles is plenty, so long as you buffer the writes and limit their frequency. All you need to do is either put a big honking RAM writeback cache next to the FlashRAM, or enforce writeback caching in the OS. If you can get the write frequency down to about two writes per hour, and do good load leveling, your FlashRAM will last for 5 years, which is about as good as most consumer grade hard disks (and possibly better, since the 'expired' FlashRAM drive could still be perfectly readable). Two writes per hour may sound like a very aggressive goal, but it wouldn't be so hard if you preferred to evict clean pages from cache before dirty pages (which is not so bad, since the read latency of FlashRAM is not nearly as long as that for a spinning platter, so refilling evicted pages is not too expensive).
But you (or any member of your species) don't have 2 billion years to explore, you only have 50 years or so (maybe more, but probably not more than one or two orders or magnitude, so 5000 years at the upper limit). In that time you can, if your technology is really good and you have (literally) tons of energy to burn, get to the nearest starts. Most of those stars will be uninteresting red dwarfs, with little likelihood of supporting life on their planets. Even if you used unmanned "self-replicating" probes, nobody that built or launched the probes would be alive when they reach their destinations: it's awfully hard to keep a project running across tens of lifetimes.
idesofmarch goes on:
My guess, based on historical performance, is that we will be thoroughly extinct by then. Two billion years, after all, is a very long time.
The best answer to the Fermi Paradox, in my opinion, is either that interstellar travel just isn't worth it (too slow and expensive for no particular gain), or there is some pheonomena, of which we are currently unaware, that prevents it (my bet is that the interstellar medium, outside of a strong magnetic field like that produced by a star, rapidly destroys solid objects).
"when consumers realize their personal lust for faster hardware makes almost zero financial sense, and hurts the environment with greater demands for power, will they start to demand cheaper, more efficient 'third-world' computers that are just as effective?" Will ridiculously cheap laptops wean consumers off ridiculously fast components?"
Maybe these same consumers will also realize that Moore's law also means that in 18 months you will be able to do the same computational work at roughly half the power cost (modulo leakage current, of course), a fact that appears to escape the razor wits at CNet.UK!
Moore's law is the only reason that we now have $5.00 calculators running off of solar cells generating a few miliwatts from ambient light, or $10.00 quartz wrist watches that run for years off a single button cell. If anything, the $100 laptop will accellerate Moore's law by increasing the volume of products produced and resultant economies of scale.
The folks at CNet.UK are a bunch of clueless wankers.
Here are some of the the problems:
That means that the next version should be, at least, Windows 8 or 9. Sure, maybe I shouldn't count Win98 and WinMe as full version increments, which would put as at just about 7.0 for the next version, but with all the other funny stuff MS has done with the Windows version numbers, why should I cut them any slack? If I were really a stickler for version numbers, I'd say that Windows NT should be at 5.0 as of Vista, which means that the next version is 6. Let's not even consider what the current DOS version number should be.
If I weren't so confused I'd just be disgusted.
Smoke 'em if you got 'em, kids!
Of course, because, as we all know, nobody would buy any product that was full of security holes: That's why Microsoft is having such a difficult time establishing themselves as a credible alternative on the desktop.
The problem with colonizing other planets, even within our own solar system, starts with the simple facts of distance and energy. The distances and energies involved with colonizing the continents of the earth were, pretty much, always within human ken. This is born out by the fact that, whenever europeans 'discovered' a new land, they found people already living there. Even a lone human, travelling on foot at normal walking speed, could circumnavigate the entire planet (given suitable land or ice briges) in a little under 2 years.
By contrast, humanity has only, in the last 50 years, even come close to controlling the amount of energy necessary to cross the gulfs between planets within our solar system, much less what is needed to travel to the nearest star. Anyone who compares the task of colonizing other planets to the european colonization of the new world, or the U.S. expansion into the west, is displaying the most profound ignorance imaginable.
The energy involved is important because it directly relates to the cost of the endeavor. The cost of colonizing distant continents was always within human grasp, so it is no surprise that it was done. The cost of travelling to other planets, however, is just barely within the grasp of the wealthiest nations, and there is no good reason to expect it ever to decrease very much.
The Fermi paradox has been used to imply that there is no intelligent life, other than us, in the galaxy, but there is another, perfectly good interpretation: maybe, even though it is possible to travel between the stars, it's just not economical to do so. Maybe the galaxy is full of intelligent life: life so intelligent, in fact, that it has long since given up the romantic, but entirely impractical, notion of interstellar travel.
I don't think that it is impossible to travel between the stars. In fact, I think that it is, basically, within human grasp at this very moment. I just think it is too expensive and dangerous to be undertaken by any nation (or similarly wealthy organized group) at this time. Give it a couple hundred years -- time enough to get the whole long-term-artificial-habitat thing, the safely-manage-tera-watt-power-generation thing, and the protect-ourselves-from-the-interstellar-medium thing down -- and I think it may be an option. At the moment, the best we could hope for would be unmanned probes to nearby stars. Even then, I doubt the transit time would be less than a half century.
Mr. Elgan has some definition of winning here that, in all probability, doesn't coincide with Apple's definition. Since neither we no Mr. Elgan knows what Apple's definition is, let's consider some of the possibilities:
1) Apple is trying to increase the market share of Safari to make it a more attrictive platform for web-developers to tailor thier websites for. In this case, if we consider the personal computing market as a whole, Apple only needs to get about 5% of Windows users to use Safari to double the Safari marketshare. Doubling marketshare is pretty big win for the kind of minimal investment represented by Safari on Windows.
2) Apple is using the Safari codebase as a trial baloon for some other, larger, Windows targeted software (maybe a Windows target in XCode?). Again, only a small penetration into the Windows userbase represents success for the effort.
3) Apple is delivering a development/test environment for iPhone to a huge pool of potnential developers. In this case, we would not expect most of the developers to use Safari as their regular browser, but they would use it when developing web pages for either iPhone or Mac clients. Again, total markter penetration (in this case, the market of web developers using Windows) doesn't need to be very large to deliver a huge benefit to Apple (greatly increase the number of people developing iPhone and Mac capable web apps).
Winning, at least for Apple, doesn't need to mean kicking MS off the web browser hill, in the same way that winning in the personal computer market hasn't, aparantly, had to mean capturing more than 5%-10% of the market. Apple seems to be doing very well with their minority share of the personal computer and laptop markets (admittedly, their share occupies the higher end of the price range, which doesn't hurt). Apple doesn't seem to think about these things as zero sum, as many other people seem to.
Yes, Roland used to spam Slashdot with articles linked to his own web page, which was reason enough to hate him. Thankfully, his posts now link to the actual articles, but that doesn't really help matters all that much. Unfortunately, Roland never seems to understand the point of the articles to which he links. He always writes sesantionalist headlines and summaries that have almost nothing to do with the actual article he references. He picks out specific buzwords from the article, buzwords that excite the most crackpottish centers of the brain, then constructs a boilerplate submission based on those buzwords.
Is this really a good reason to hate Roland? probably not. It's not like anyone is forcing you to read his drivel, you can always ignore his articles. At least he isn't pushing his own crackpottery with his articles (ala Time Cube or Mentifex). Still, he gets under the skin.
so, clearly, I cannot choose the cup in front of you.
so, clearly, I cannot choose the cup in front of me.
The main reason that 'parallel' programming is hard is that current tools support it only as an afterthought. All mainstream tools (languages, compilers, libraries, debuggers, etc.) are built, primarily, to support single-threaded code. It is up to the individual programmer to write code that is multi-thread safe by enclosing dangerous sections of code in mutexes, erecting semephores to coordinate multiple threads, and so on. Even languages with explicit support for multi-threading (e.g. Java) only support it as the exception: all code is presumed to be inteded for a single-threaded environment unless the programmer explicitly identifies it as multi-threaded (e.g. 'synchronized').
A better solution would be to assume that all code is run in a multi-threaded environment and have the tools build the code for mutual exclusion, unless explicitly told otherwise by the programmer. The language and libraries could also provide basic utilities and data structures in thread-safe implementations by default, requiring the programmer to specifically select a non-thread-safe version when desired. Monitors (thread-safe objects) were invented decades ago, but are still not in common use in modern languages.
The fundamental problem is that people are really, really bad at reasoning about thread-safety and figuring out where mutually exclusive sections of code and race conditions are. Relying on the programmer to get such things right is a guarantee of failure. The tools (languaes, compilers, etc.) should be able to determine when methods/functions access the same common state and automatically insert mutexes/semephores to enforce mutual exclusion, or at least assume that, unless told otherwise, all methods/functions must have mutexes/semephores. You'd still have instances where a programmer thought, incorrectly, that some code could be run with a mutex, but it would be the exception rather than the rule.
Oh. My. God! I love it.
Now all we need is to identify some 230-some Microsoft patents that could be construed to be infringed by specific OSS packages mentioned by MS executives, and we have ourselves a field-day invalidating MS patents. Even if only a few of the cases succeed (I'm assuming one declaratory jundgement suit per patent) it would be pretty damaging to Microsoft to 1) lose some of it's patents, 2) have to mount the separate defenses in court and 3) deal with the resulting PR storm.
The initial research phase is exactly the sort of thing the OSS community should be good at: a distributed analysis of technical documents. MS has, admittedly, a lot of patents, but they have given us a number of clues to work from. It really shouldn't be all that hard to mobilize a small army of OSS-sympathetic law students/junior lawyers/paralegals to comb through MS's patent portfolio along with the list of allegedly infringing OSS projects and develop a candidate patent list.
The lawsuits may be a bit more trouble: you would need parties with pretty deep pockets to initiate and persue each lawsuit, who also happened to be either large scale users of OSS, distributors of OSS or the actual developers of the possibly infringing OSS projects.
A side benefit of the entire effort would be to develop a list of 'bugs' in OSS which, when fixed, would immunize the OSS community against further MS FUD.
First, any party bringing a lawsuit under the American legal system must have standing to sue, meaning that they must have a material interest in the outcome of the suit. Since Microsoft has not specifically threatened anybody, at the moment, it would be hard to establish standing. A really good lawyer might be able to argue that a Linux vendor is harmed by the implication that Microsoft will sue that vendors customers for patent infringement, but with any actual patent suits in process, it would be a hard sell.
Second, the American legal system refuses to issue advisory judgements, and requires that a case be 'ripe' before it can be adjudicated. Since the court would be ruling on a hypothetical ("if Microsoft were to sue for patent infringement, would we be found to be infringing?") the court would (and should) simply refuse to hear the case.
Here is a good reference for standing, advisory opinions and ripeness. A little google-foo should easily turn up others.
The only bright spot in this, from a potential plaintif's point of view, is that, as a convicted monopolist, there might be a way to accuse Microsoft of restrain-of-trade, or some other violation of the Sherman Act. Unfortunately, I think that prosecution of anti-trust cases must be brought by the federal government, and that is not very likely with the current administration. Private actions can be brought for violations of the Clayton Act but I don't quite see how it could aply in this case, and only consumers injured by their dealings with the violator have standing to sue, which puts most direct competitors out of the running.
Disclaimer: IANALBIHTBL (IANAL But I Have Taken Business Law)