More often than not, it was simply an excuse for them to search us, and sweep for drugs while they were at it. I always wanted to hide a huge bag of dog treats in my locker to see what would happen to the drug-sniffing dogs...
In High School, I kept a few baggies in my locker. One was a baggie filled with white powder labelled something along the lines of "wake me up". The other was filled with dried green crushed plant leaves labelled "party mix." They were sugar and a blend of basil and oregano. I left them in there in case of drug searches. Nothing ever came of it. I have long wondered who found the baggies and what they thought. Was it an administrator who thought for a moment that he had finally made a big bust he could brag about? Was it a janitor that thought for a moment that he was going to have a hell of a night? Sadly, I'll never know.
I strongly encourage anybody in high school to try it. You may well get sent to jail for it, but you will be able to laught at your stupid school when it happens.
Looked at purely in economic terms, railways are a commercial nonsense and always have been. As far as I'm aware, there is no heavy rail system anywhere in the world that competes successfully with other modes of transport on price alone, nor even that makes a profit without government subsidy to drive it.
Well, it's a fair point that rail is expensive. It's also a fair point that completely private rail tends to have trouble competing against other methods of transport, thus the most successful rail systems are government subsidised and don't make a profit on their own. All these points are perfectly fair and reasonable.
So, where are the other completely unsubsidised methods of transportation that private rail can compete with? Government maintains the roads. Government has a hand in the airports. Government has a hand in the bus lines. Given the amount of money spent on road construction, road repair, police to catch speeders, judges to yell at speeders, driver's education in schools, tax breaks to companies that make gasoline, etc., on and on, is it really fair to think that rail should only be considered valid competition to the roads if the givernment doesn't help it out at all?
Imagine how expensive your commute would be if you had to either pave your own way, or go by a completely private toll road. Suddenly, even the over priced private trains seem to be like a really inexpensive alternative. In general, I don't support government assistance for business. I disagree with tax breaks for oil companies as an example. OTOH, I think it may be perfectly reasonable to fund rail. Hell, let the government build the rail and let all residents ride for free. See how much traffic that clears up, and how much less you have to spend on roads. You might damn near break even to give away train passes.
I just got One or Zero installed last night. You don't say exactly what sort of issues you are tracking, but you may find that it works for you. It isn't the best written thing in the world, but some of the other stuff I looked at seems worse. One bug I discovered (with help from another guy -- can't take credit for the discovery) is that it fails horribly if you install it in a directory named admin. I'm going to have to submit a bug report and a post to thedailywtf.com about that.
Well, you must be clinically insane. GIMP outstrips any commercial equivalent?? It doesn't even come close to Photoshop. It's not even in the same category. I think you might be the troll here. The fact is that while there are plenty of applications for Linux, most of them are far less functional than the best commercial software. It would be nice if this wasn't true, but it is. For example, show me where you can get a Linux video editing application that even comes close to commercial counterparts like Final Cut Pro or Avid.
Some people will point to video editing apps in an effort to argue with you, but I'll concede the point about Final Cut Pro. As for GIMP vs. Photoshop... I agree that they aren't exactly in the same category. I won't agree that either is better unless you specify which category you are interested in. For CMYK work, Photoshop obviously wins. Hands down, no argument. Yes, a lot of people really do need to work natively in CMYK. OTOH, not everybody needs to. And, given that the GIMP does have features that Photoshop doesn't, for some things, GIMP wins hands down. Yes, really.
Likewise, many compilers exist which produce code that beats gcc on at least one benchmark on one specific platform. gcc can probably be beaten by more compilers than any other compiler in existence, in fact. If you happen to have code that maps well to a particular benchmark on a particular platform, then your customers will thank you to go with one of those winning compilers. OTOH, the only reason that gcc is probably beaten by more compilers than any other is because it supports more platforms to get beat on than any other compiler that I have used. There isn't any compiler that can beat gcc on every platform that gcc works with. gcc also allows any of the supported platforms to be either host or target. Because of that, it's really hard to say that gcc is the best compiler, but it's also impossible to say that CCC or XCC or ICC are "better." After all, I never could figure out how to make the Intel C Compiler run under SH-BSD while targetting ARM-Linux. (Okay, I never specifically tried that exact configuration with any compiler, but you get my point...)
Similarly, Blender has an odd interface and a ton of features. It does some things stunningly well. A few things in Blender are fantastic at any price. It also does a lot of stuff terribly, IMO. I prefer Lightwave for most of my 3D work. But, Lightwave isn't better than Blender. It just maps better to my needs and experience.
... and I guess I still qualify as a noob. I really should have known better than to try it. But I'm curious by nature. I'm just glad I didn't have any work open at the time. So, enjoy the chuckles. Out of curiosity, what exactly DID happen? The best I can figure out is I was logged out immediately and sent to a terminal console. I did it again and it let me log back in with the graphical interface but I'm not exactly sure what this "feature" is supposed to be for. Can anyone clear that up for me?
You killed the X server. When you decide to play with your X configuration to get a new resolurion, or you have installed a new card or something, sometimes you screw up terribly. (At least I do.) When this happens, you get some horrible noise on your monitor, or a black screen or something, and can't click the nice button do do things properly. So, you kill the X server.
Linux allows you have have vTTY's - (press (ctrl)-alt-Fn to switch between them) so you could just switch to a new vTTY to kill X if you wanted to. Some other UNIXy OS's don't (or at least didn't) support multiple vTTY's on the same workstation, so there was no easy way to switch to another terminal to kill X, and you really needed the magic kill keys. (Assuming you couldn't log into the machine remotely to kill X...)
These days, running a modern Linux distribution with XDM/GDM (so you login on a graphical screen instead of logging in at a text mode screen and then starting X), you can use ctrl-alt-backspace if you change some X settings and want them to start working without a reboot because GDM will relead the X server, but it will do so with the current configuration. This allows you to do things like upgrade a graphics driver without bringing down a machine that is being used for a compute/server task at the same time. Ideally, you would have a separate server machine where you don't care about the graphics driver being up to date. But, sometimes we don't get to do things as we would like.:)
Intel said it had already manufactured prototype microprocessor chips in the new 45-nanometer process that run on three major operating systems: Windows, Mac OS X and Linux.
Again, I thought it's the operating systems who run on microprocessors, not vice-versa. And I [not being a kernel developer, though] can't see any reason for an OS to stop functioning on a new processor model if the architecture is intact and no serious hardware-level bugs are introduced.
Well, yeah. That's pretty much the point. Usually, a first go a new processor has serious hardware bugs and doesn't run very well. Running existing O/S's in real silicon is a very important step in the creation of a new processor. If things are going smoothly with the design, and they have it running real code more quickly than they figured, then it speaks well to the possibility of them being able to ramp up production easily.
1. Quick rough primer
2. Source of links, some of which may end up being citeable
3. Inspiration for finely-honed Google searches for authoritative sites
4. Absolutely nothing else
Unless you are writing a research paper on the different colors of Chocobo in a specific Final Fantasy game. In this case, Wikipedia is the definitive reference.
BINGO. Encyclopedia's are reference materials, basically intended to give you a background on a particular subject -- from there you can research further. Its a giant yellow pages for researchers.
Add to that the volatility of wikipedia (e.g. you can't reference its contents, since they're always in flux), and its a poor resource of term papers.
1 - Yes, I agree that considering wikipedia as a source is stupid, just like referencing an encyclopedia is stupid. Wikipedia *specifiaclly bans* original research, so it certainly can't be called a primary source for anything except information about itself.
2 - Actually, you can cite an article's state at a specific time, which will always get you back to that same version. So, the state of flux isn't a problem to anybody who bothers to research how best to cite wikipedia. OTOH, anybody who researches how cite well enough to be able to cite wikipedia properly, probably knows that they shouldn't do it in the first place.:)
3 - The students didn't already know to reference Wikipedia's references listed at the bottom of the article instead of the article itself in order to inflate their reference count? Slackers.
It flags the submissions so you don't accidentally click on the links and support his advertising, and it strips the text so you don't have to bother reading it. I use it on most of my systems and heartily reccomend it.
I think in Australia traditionally you had technical colleges (such as TAFE) and Universities providing a clear difference in the direction of things being taught. Technical colleges producing "tradesmen" and Universities producing "thinkers". The problem has been that increasingly universities have been seen by consumers as a way of getting a job rather than as a pathway to higher learning as academia and thus there is expection by them, to be taught "practical" skills. I think a reason for this is there is a small stigma attached to technical and trade colleges as being "dumber" than their uni counterparts. I think in this way, the problem is that consumers do not really understand what the function of universities are.
Here in America, we also have Universities and Technical/Vocational schools. They are sort of on a continuum with Community Colleges in the middle, so the dividing line may not be exactly where it is in other countries. But, the technical schools have really terrible reputations. Most of them are for things like being an auto mechanic, but some try to teach things like IT. Nobody really wants to hire somebody from one of the technical schools for a job like IT. OTOH, I think that we need a sort of super-elite class of technical school.
Imagine, for example, and extremely competetive programming school. Not CS. Programming. Applicants would be expected to already have the equivalent knowlege of 3/4 of a bachelor's degree in CS when applying, with a full CS degree greatly preferred. It would have courses like "Rote memorization of every single function in the C standard library" and "Using the bells and whistles of the current version of Visual Studio" and "Dealing with the annoying quirks of the current versions of several popular SQL interface libraries."
So, anybody coming out of that sort of program would have all the skills to be an extremely productive programmer in a typical software development house. You would be expected to learn the zen and theory of CS from a university. That would be separate. I think a lot of companies want to hire programmers, so they hire guys with CS degrees. CS degrees aren't really programming degrees. They cover a lot of information which is greatly useful for being a programmer, but there are a lot of skills that are handy for developing business apps that you are expected to get on your own. OTOH, with the current technical schools, you get a little bit of practical information without ever being expected to know the theory, and you are only expected to be moderately acquianted with the practical stuff.
Even if you could apply some force to slowly accelerate a massive space ship, once you got it up to that speed wouldn't it take K^2 (squared) units of fuel to slow it down it again?
Say what? Where'd you get the idea that it takes more fuel to slow down than to speed up? Acceleration and deceleration are the same thing, change in delta-v.
I think I get what he means -- every pound of fuel you use to slow down means an extra pound of mass that you need to accelerate in the first place. (Which takes more fuel to do the initial acceleration.) So, yeah, they are the same thing, but your total required fuel does grow explosively from the bulk of all the other fuel you need to use.
To paraphrase: But Sir! If we only send 8 probes it'll take billions of years to search a mere 4% of the Milky Way galaxy!
That's why you have to make the probes self replicating.. utilizing in-situ resources to make more probes at each star they visit, the growth becomes exponential and it only takes a few thousand years to search the entire galaxy. And seeing as we're visiting all these stars anyway, how about looking for planets that don't have life on them, but have nice suitable conditions for starting life on them. Cover a virgin planet with a wide variety of Earth lifeforms and fly on.
Ummm... The galaxy is something like 100,000 ly across. So, if we pretend that we were in the middle, that's 50,000 ly of travel in every direction. Assuming we could get a probe close enough to c as makes no never mind, then that's 100,000 years minimum for the probe to get to the other side of the galaxy and transmit a "Hello World" back to us. (Does it count as 'explored' if we can only declare that the probe probably made it by now, but we won't know if it made it for another 50,000 years?) I don't really conisder 100,000 to be a "few thousand." And *we* couldn't do it that fast -- only some hypothetical alients in the dead center.
Now, that assumes that the probe could accelerate to top speed really fast, so the time spend accelerating doesn't count. If the probe is stopping to build more of itself every few star systems, then it is going to have to slow down, stop, and spend time building new probes. On a 50,000 year journey, 4-ish years spent getting up to speed is quite negligible. And, since close enough to c as makes no never mind is really quite remarkably fast, it will probably take at least that long to get going (or to get stopping). Assuming a pit stop every ten ly (despite the fact that stars may be less than 1 ly apart in the densest parts of the galaxy, and our nearest neighbor is only about 4 ly), with the stopping and going deceleration time, it'll take about 14 years to cover the 10 ly. So, the outer edge of our fleet of probes will take something like 140,000 years to cross half the galaxy. Given our actual position, we will actually have like 3/4 of the galaxy to cross to get to the far edge, which would push up the time for exploration up to the best part of 200,000 years.
Figuring out how to accelerate your probes to close enough to c for this to be close to right, while also carrying enough propellant to slow down in the next system and refuel to be able to do it again is left as an excercise to the reader. Oh, and you have to lug an antenna big enough to send a signal back. (Or maybe just to the nearest probe which will relay it... But, that means no straight line path for the signal back home, which means it takes longer...)
Anyhow, the replicating probes idea is pretty neat, and I'm all for it, but we certainly won't have the whole galaxy explored in a "few" thousand years for any forseeable technology. If we can develop FTL and whatnot, all bets are off. the whole problem may turn out to be an x ly trip to the nearest black hole, and a y ms hop to anywhere in the universe.
Seriously, I had a look at a few pages, and when I eventually managed to CTRL-ALT-DELETE my browser into submission, I made damn well sure never to go back there. Are there people that actually have enough computing power to handle some of those profiles?
Yes - there are tons of people who have plenty of power for browsing myspace. Tops on the list would be those with noscript installed and not set to trust myspace.com. Next would be people who browse using lynx or links. (Of course, in Links, it always says you have new messages and invitations and everything, which is kind of annoying if you are just checking to see if you have new mail.)
What? Yes, I really have gone to myspace.com using links.
AFAIK, the Total War sequence of games has no problem tracking dead bodies, and there may be thousands of them!
Certainly Rome:Total War leaves the dead on the battlefield, even if they are simplified. Even missiles, such as arrows are tracked into the ground and only disappear after a while.
I fail to see the problem with letting the dead pile up, they're just objects like everything else.....
Total War doesn't let you build any units during combat, so it always knows that if it can draw the start of the battle, it can hang onto all the corpses. The only technological problem comes when you can build during combat, like in most RTS games. If you have your barracks pumping out grunts to send to the front lines, you could get a stalemate condition with potentially unlimited corpses to keep track of. That's always rough with finite memory. So, even with the new super powered consoles, some games will have to continue to have disappearing corpses.
Hopefully, modern RTS games will have the corpses rot beautifully and be absorbed into the ground to become resource deposits, or something, rather than just disappearing. I think it would be interesting to have a strategic element to scavenging battlefields. You see it somewhat with necromancer type units, but I mean just looting the corpses for raw materials.
No matter how sexy or cool things seemed early on, the day will come when you will wonder what the fuck you were thinking.
I thought Apple hardware was like getting married because I don't worry about my parents being able to interact with it, and I don't have to worry as much about getting a virus from the chick that everybody has had.
Not 60% of the time, but 60% more than expected if it were chance alone. So more likely 1.6 * (30*.5) = 24/30, not 18/30.
But of course the actual number isn't in the article.
Duh! What kind of moron are you? Chance would be 50%. This is 60% more than chance, which would be 110%. So, they correctly identified 33 out of 30. That's statistically significant, right? I mean, come on guys, don't you people know how scientific sounding numbers are made up?
Certainly, I think an interviewer has zero obligation to spend his time explaining to somebody what they did wrong. Certainly not for free.
That said, I think in many circumstances, it can be a good thing to explain to somebody why they didn't get the gig. If they undertake a course of self improvement, they could potentially apply for a different position in a few years and prove a really valuable asset. Before I left my last job, there was a huge amount of bitterness related to internal job applications for position transfers. People would be rejected with no idea why. It was killing morale. I don't know if they ever improved the situation, but it would have been really easy to say, "Look, Suzie Q, when we open up to public applications, most of the people applying for this type of position have qualifications X,Y, and Z in these amounts. You only have X, and only in this amount. So, it's not personal, but I think we are going to keep looking. If you really want to move into this position, we really think that only A and not B will be the best route to getting Y and Z."
Instead, with really vague requirements, people thought they were perfectly qualified, and had no idea how to get better-qualified. They also thought that it was just a matter of personal grudges.
With external applicants, I think it is less important, but it doesn't usually hurt. I suppose you might consider it valuable to keep some of the stunning idiots in the industry in hopes that they will work with your competitors. But, you may eventually work with them too. And, you will have to maintain their code. Probably safer for everybody just to point out to them how clueless they are.
And, when I'm away from my day job, I do theater stuff. I was recently involved in some auditions to expand an improv troupe I am in. Not everybody got individual commentary, but the folks dismissed in the first round did at least get a general explanation. Everybody who made it past the first cut got an explanation of what impressed the director, and what he thought they could most work on - both the folks who made it and those who didn't. Personally, I wish we could have taken a little more time to offer personal advice to some of the folks in the first round. I would have liked suggesting that the hot chicks take classes that I can sit in on and watch them learn. Especially one blonde. I tried to convince the director that she should join the troupe and just not be allowed to say anything. I would have been cool with that.
Those specs don't seem to include a lot of the actual specs... Which I find odd. If this were a mobile OS-X device of my dreams, I'd know what sort of processor it has, and what sort of graphics hardware, etc. Unfortunately, it seems like they may plan on pushing it like a phone. Locked down, limited access to development tools, etc. That really could completely kill my interest in the device. Why is the fact that it runs OS X interesting to me if I can't easily program it? Am I supposed to get some pointless pride in the fact that in runs a particular kernel I can't interact with? Is it supposed to be more impressive to me than a phone running some other kernel if I can't actually see any difference?
"They can't stream content if you want video. The connection isn't the problem, the problem is NAS isn't fast enough to get content on the wire," he said. With a faster processor in its Media Server, data can now be streamed off a server, either wireless or wired.
WTF? I stream videos off my 400mhz K6 fileserver and have never had problems with CPU load. Are NAS devices seriously that slow?
No, they aren't. But, they are trying to convince everybody to do it wrong in order to sell more hardware. I think they want to have a smart server that transcodes video in real time to deliver it to the front end. Personally, I think that's generally pretty stupid. I mean, I probably downloaded the video off the internet anyway, so it was most likely compressed reasonably small for internet transport. If I try to recompress it on the fly, then I will probably just make bigger uglier artifacts in the video.
Sure, there are occasionally files that are not well compressed that live on my server and I want to play wirelessly on a front end machine. It might be reasonable to have the front end try to play something, realise it can't, and then volunteer to download locally in the background (if there is local storage), and also give the option of transcoding and playing the result. But, I find that such things are pretty rare. I've even played raw DVD files (MPEG-2) off my server w/ 802.11g without problems. The minimally compressed files that I do tend to have that actually wouldn't play over wireless are either funny scinetific simulation videos that I have downloaded, or my own work-in progress stuff that I'm not ready to compress yet. Neither case is common enough for a home user to need to worry about it in the Mom and Pop media server.
Of course... I use an Alpha Server as my home media storage box, so I guess I can't really complain about people using more machine than they need...:)
Move back? They were never on them. POWER6 != powerpc (though they are similar in more ways than not).
I think Apple is perfectly happy with the Intel move at this point. One of the reasons for the migration (if you can get past Jobs' reality distortion field of blah blah per watt or whatever) was that IBM wasn't able to keep up with demand, either with getting the speeds up, or with delivering the slow crappy ones they already had.
Well, no, Apple never used POWER6 specifically, but they did use PPC, and IBM's current marketing literature says that PowerPC is POWER. They used to say POWER was PowerPC, but anyhow, according to: http://www-03.ibm.com/chips/power/aboutpower/
"Power Architecture encompasses PowerPC®, POWER4(TM) and POWER5(TM) processors."
So, Apple won't use POWER6. Apple never used the earlier "POWER" branded chips because they never built any systems that would use them. Macs just aren't that big. Also, Apple has switched completely to X86, and hasn't bothered to really keep alive any hope of a mixed platform for the Mac. But, if Apple did move to POWER6, it would count as going back to the old architecture, rather than moving to a brand new one.
No, I have not read TFA all the way through (sorry, I have to work), but just how are the graphics capabilities "locked down"? I have a PS3 and I am waiting for a bigger drive to come in (just ordered) before installing YDL as I don;t want to do it twice. I have not yet heard about the graphics lock down though- just how bad is it?
From everything I have read, it's pretty much exactly what the article said in the summary. No accelerated graphics. You also don't get full access to the hard drive, or direct access to any of the hardware. All the hardware works except for GPU, for which you get just a dumb frame buffer. Not sure exactly what you are asking when you want to know how bad it is.
I could be wrong - but I thought these days, ARM massively outsold x86 because it's in so many embedded devices - and dozens of these are sold for each PC sold. (Of course, ARM is optimized for low electric power, where vast computing power is not needed).
A fair and valid point! Indeed, there are oodles of chips sold that aren't X86. Your car may have a dozen small computers, or more. Your cell phone, your iPod, your PDA, your Wii, your router, etc don't have X86. When you get to exotic DSP's and microcontrollers, it gets hard to decide where exactly you draw the line when determining how many processors are actually sold.
I sort of assumed the conversation was desktop specific because the question was "Why is x86 the winner" rather than "Is x86 the winner?" The desktop is certainly the most visible segment, but it's also the *only* one where x86 is really the winner! OTOH, the desktop (And I'm being fuzzy with terms and counting desktop, and some server, and workstation, etc., as the "desktop" uebercategory) tends to be an extremely lucrative segment of the market. It has much higher volume than crazy supercomputer chips, and the chips tend to cost much more than the seven cent 8 bitters that wind up in your singing greeting cards and whatnot.
If I had been thinking about this, I probably should have included some discussion of this in my previous comment, thanks for pointing it out!
It's never been explained, to my satisfaction, why the use of paper ballots (or at least paper TRAILS), had to be replaced with the computer-voting machines.
And not just replaced, but REPLACED RIGHT NOW with very little public input and negligible testing. Whenever I see such a huge rush to change something that's worked remarkably well for generations I get suspicious. When I see such a huge rush to change something that's worked for generations without any meaningful dialogue about whether it really should be done, I get even more suspicious.
In 2000, we realised that things were fucked up. Seriously, they really were. They had been for a long time, and we finally realised it. Florida shouldn't have gone down the way that it did.
So, after that mess happened, rather than a careful introspective discussion about the best way to ensure that democracy was served, some slick saleslobbyists showed up and the conversation went something like this:
"We have to do something!" "This is something." "We have to do this!"
Because the congresscritters knew that if they didn't look busy and make some noise at a press conference about how they passed legislation to save America and Freedom and that Democracy thingie, then they might get fired. They knew that the voters wouldn't read the bills. They knew the voters wouldn't deeply contemplate the bills. They also knew that if all hell did break lose, they would have another chance to have a press conference, pass another bill, and make some noise about how they saved that Democracy thingie that everybody talks about.
Non x86 architectures are certainly not inherently better clock for clock. That's a matter of specific chip designs more than anything else. The P4 was a fairly fast chip, but miserable clock for clock against a G4. An Athlon however, was much closer to a G4. (Remember kids, not all code takes advantage of SIMD like AltiVec!) And, the G4 wasn't very easy get bring to super high clock rates. The whole argument of architectural elegance no longer applies.
The RISC Revolution started at a time when decoding an ugly architecture like VAX or x86 would require a significant portion of the available chip area. The legacy modes of x86 significantly held back performance because the 8086 and 80286 compatibility areas took up space that could have been used for cache or floating point hardware, or whatever. Then, transistor budgets grew. People stopped manually placing individual transistors, and then they stopped manually fiddling with individual gates for the most part. Chips grew in transistor count to the point where basically, nobody knew what to do with all the extra space. When that happened, x86 instruction decoding became a tiny area of the chip. Removing legacy cruft from x86 really wouldn't have been a significant design win after about P6/K7.
Instead of being a design win, the fixed instruction length of the RISc architectures no longer meant improved performance through simple decoding. They meant that even simple instructions took as much space as average instructions. Really complex instructions weren't allowed, so they had to be implimented as multiple instructions. Something that was one byte on x86 was always exactly 4 bytes on MIPS. Something that was 12 bytes on x86 might be done as four instruction on MIPS, and thus take 16 bytes. So, effective instruction cache sizes and effective instruction fetch bandwidth grew on X86 compared to purer RISC architectures.
At the same time, the gap between compute performance and memory bandwidth on all architectures was widening. Instruction fetch badwidth was irrelevent in the time of the PC XT, because RAM fetches could actually be done in like a single cycle. Less that it takes to get to SRAM on-chip caches today. But, as time went on, memory accesses became more and more costly. So, if a MIPS machine was in a super tight loop that ran in L1 cache, it might be okay. But, it it was just going balls to the wall through sequential instructions, or a loop that was much larger than cache, then it didn't matter how fast it could compute the instructions if it couldn't fetch them quick enough to keep the processor fed. but, X86 absurdly ugly instruction encoding acted like a sort of compression, meaning that a loop was more likely to fit in a particularly sized cache, and that better use of instruction fetch bandwidth was made.
Also, people had software that ran on X86, so they bought 9000 bazillion chips to run it all. The money spent on those 9000 bazillion chips got invested in building better chips. If somebody had the sort of financial resources that Intel had to build a better chip, and they shipped it in that sort of volume, we might well se an extremely competetive desktop SPARC or ARM chip.
Slashdot Mirror
I strongly encourage anybody in high school to try it. You may well get sent to jail for it, but you will be able to laught at your stupid school when it happens.
So, where are the other completely unsubsidised methods of transportation that private rail can compete with? Government maintains the roads. Government has a hand in the airports. Government has a hand in the bus lines. Given the amount of money spent on road construction, road repair, police to catch speeders, judges to yell at speeders, driver's education in schools, tax breaks to companies that make gasoline, etc., on and on, is it really fair to think that rail should only be considered valid competition to the roads if the givernment doesn't help it out at all?
Imagine how expensive your commute would be if you had to either pave your own way, or go by a completely private toll road. Suddenly, even the over priced private trains seem to be like a really inexpensive alternative. In general, I don't support government assistance for business. I disagree with tax breaks for oil companies as an example. OTOH, I think it may be perfectly reasonable to fund rail. Hell, let the government build the rail and let all residents ride for free. See how much traffic that clears up, and how much less you have to spend on roads. You might damn near break even to give away train passes.
I just got One or Zero installed last night. You don't say exactly what sort of issues you are tracking, but you may find that it works for you. It isn't the best written thing in the world, but some of the other stuff I looked at seems worse. One bug I discovered (with help from another guy -- can't take credit for the discovery) is that it fails horribly if you install it in a directory named admin. I'm going to have to submit a bug report and a post to thedailywtf.com about that.
Anyhoo... http://www.oneorzero.com/
Likewise, many compilers exist which produce code that beats gcc on at least one benchmark on one specific platform. gcc can probably be beaten by more compilers than any other compiler in existence, in fact. If you happen to have code that maps well to a particular benchmark on a particular platform, then your customers will thank you to go with one of those winning compilers. OTOH, the only reason that gcc is probably beaten by more compilers than any other is because it supports more platforms to get beat on than any other compiler that I have used. There isn't any compiler that can beat gcc on every platform that gcc works with. gcc also allows any of the supported platforms to be either host or target. Because of that, it's really hard to say that gcc is the best compiler, but it's also impossible to say that CCC or XCC or ICC are "better." After all, I never could figure out how to make the Intel C Compiler run under SH-BSD while targetting ARM-Linux. (Okay, I never specifically tried that exact configuration with any compiler, but you get my point...)
Similarly, Blender has an odd interface and a ton of features. It does some things stunningly well. A few things in Blender are fantastic at any price. It also does a lot of stuff terribly, IMO. I prefer Lightwave for most of my 3D work. But, Lightwave isn't better than Blender. It just maps better to my needs and experience.
Linux allows you have have vTTY's - (press (ctrl)-alt-Fn to switch between them) so you could just switch to a new vTTY to kill X if you wanted to. Some other UNIXy OS's don't (or at least didn't) support multiple vTTY's on the same workstation, so there was no easy way to switch to another terminal to kill X, and you really needed the magic kill keys. (Assuming you couldn't log into the machine remotely to kill X...)
These days, running a modern Linux distribution with XDM/GDM (so you login on a graphical screen instead of logging in at a text mode screen and then starting X), you can use ctrl-alt-backspace if you change some X settings and want them to start working without a reboot because GDM will relead the X server, but it will do so with the current configuration. This allows you to do things like upgrade a graphics driver without bringing down a machine that is being used for a compute/server task at the same time. Ideally, you would have a separate server machine where you don't care about the graphics driver being up to date. But, sometimes we don't get to do things as we would like.
2 - Actually, you can cite an article's state at a specific time, which will always get you back to that same version. So, the state of flux isn't a problem to anybody who bothers to research how best to cite wikipedia. OTOH, anybody who researches how cite well enough to be able to cite wikipedia properly, probably knows that they shouldn't do it in the first place.
3 - The students didn't already know to reference Wikipedia's references listed at the bottom of the article instead of the article itself in order to inflate their reference count? Slackers.
It flags the submissions so you don't accidentally click on the links and support his advertising, and it strips the text so you don't have to bother reading it. I use it on most of my systems and heartily reccomend it.
Here in America, we also have Universities and Technical/Vocational schools. They are sort of on a continuum with Community Colleges in the middle, so the dividing line may not be exactly where it is in other countries. But, the technical schools have really terrible reputations. Most of them are for things like being an auto mechanic, but some try to teach things like IT. Nobody really wants to hire somebody from one of the technical schools for a job like IT. OTOH, I think that we need a sort of super-elite class of technical school.
Imagine, for example, and extremely competetive programming school. Not CS. Programming. Applicants would be expected to already have the equivalent knowlege of 3/4 of a bachelor's degree in CS when applying, with a full CS degree greatly preferred. It would have courses like "Rote memorization of every single function in the C standard library" and "Using the bells and whistles of the current version of Visual Studio" and "Dealing with the annoying quirks of the current versions of several popular SQL interface libraries."
So, anybody coming out of that sort of program would have all the skills to be an extremely productive programmer in a typical software development house. You would be expected to learn the zen and theory of CS from a university. That would be separate. I think a lot of companies want to hire programmers, so they hire guys with CS degrees. CS degrees aren't really programming degrees. They cover a lot of information which is greatly useful for being a programmer, but there are a lot of skills that are handy for developing business apps that you are expected to get on your own. OTOH, with the current technical schools, you get a little bit of practical information without ever being expected to know the theory, and you are only expected to be moderately acquianted with the practical stuff.
I think I get what he means -- every pound of fuel you use to slow down means an extra pound of mass that you need to accelerate in the first place. (Which takes more fuel to do the initial acceleration.) So, yeah, they are the same thing, but your total required fuel does grow explosively from the bulk of all the other fuel you need to use.
Ummm... The galaxy is something like 100,000 ly across. So, if we pretend that we were in the middle, that's 50,000 ly of travel in every direction. Assuming we could get a probe close enough to c as makes no never mind, then that's 100,000 years minimum for the probe to get to the other side of the galaxy and transmit a "Hello World" back to us. (Does it count as 'explored' if we can only declare that the probe probably made it by now, but we won't know if it made it for another 50,000 years?) I don't really conisder 100,000 to be a "few thousand." And *we* couldn't do it that fast -- only some hypothetical alients in the dead center.
Now, that assumes that the probe could accelerate to top speed really fast, so the time spend accelerating doesn't count. If the probe is stopping to build more of itself every few star systems, then it is going to have to slow down, stop, and spend time building new probes. On a 50,000 year journey, 4-ish years spent getting up to speed is quite negligible. And, since close enough to c as makes no never mind is really quite remarkably fast, it will probably take at least that long to get going (or to get stopping). Assuming a pit stop every ten ly (despite the fact that stars may be less than 1 ly apart in the densest parts of the galaxy, and our nearest neighbor is only about 4 ly), with the stopping and going deceleration time, it'll take about 14 years to cover the 10 ly. So, the outer edge of our fleet of probes will take something like 140,000 years to cross half the galaxy. Given our actual position, we will actually have like 3/4 of the galaxy to cross to get to the far edge, which would push up the time for exploration up to the best part of 200,000 years.
Figuring out how to accelerate your probes to close enough to c for this to be close to right, while also carrying enough propellant to slow down in the next system and refuel to be able to do it again is left as an excercise to the reader. Oh, and you have to lug an antenna big enough to send a signal back. (Or maybe just to the nearest probe which will relay it... But, that means no straight line path for the signal back home, which means it takes longer...)
Anyhow, the replicating probes idea is pretty neat, and I'm all for it, but we certainly won't have the whole galaxy explored in a "few" thousand years for any forseeable technology. If we can develop FTL and whatnot, all bets are off. the whole problem may turn out to be an x ly trip to the nearest black hole, and a y ms hop to anywhere in the universe.
Yes - there are tons of people who have plenty of power for browsing myspace. Tops on the list would be those with noscript installed and not set to trust myspace.com. Next would be people who browse using lynx or links. (Of course, in Links, it always says you have new messages and invitations and everything, which is kind of annoying if you are just checking to see if you have new mail.)
What? Yes, I really have gone to myspace.com using links.
Total War doesn't let you build any units during combat, so it always knows that if it can draw the start of the battle, it can hang onto all the corpses. The only technological problem comes when you can build during combat, like in most RTS games. If you have your barracks pumping out grunts to send to the front lines, you could get a stalemate condition with potentially unlimited corpses to keep track of. That's always rough with finite memory. So, even with the new super powered consoles, some games will have to continue to have disappearing corpses.
Hopefully, modern RTS games will have the corpses rot beautifully and be absorbed into the ground to become resource deposits, or something, rather than just disappearing. I think it would be interesting to have a strategic element to scavenging battlefields. You see it somewhat with necromancer type units, but I mean just looting the corpses for raw materials.
I thought Apple hardware was like getting married because I don't worry about my parents being able to interact with it, and I don't have to worry as much about getting a virus from the chick that everybody has had.
Duh! What kind of moron are you? Chance would be 50%. This is 60% more than chance, which would be 110%. So, they correctly identified 33 out of 30. That's statistically significant, right? I mean, come on guys, don't you people know how scientific sounding numbers are made up?
Certainly, I think an interviewer has zero obligation to spend his time explaining to somebody what they did wrong. Certainly not for free.
That said, I think in many circumstances, it can be a good thing to explain to somebody why they didn't get the gig. If they undertake a course of self improvement, they could potentially apply for a different position in a few years and prove a really valuable asset. Before I left my last job, there was a huge amount of bitterness related to internal job applications for position transfers. People would be rejected with no idea why. It was killing morale. I don't know if they ever improved the situation, but it would have been really easy to say,
"Look, Suzie Q, when we open up to public applications, most of the people applying for this type of position have qualifications X,Y, and Z in these amounts. You only have X, and only in this amount. So, it's not personal, but I think we are going to keep looking. If you really want to move into this position, we really think that only A and not B will be the best route to getting Y and Z."
Instead, with really vague requirements, people thought they were perfectly qualified, and had no idea how to get better-qualified. They also thought that it was just a matter of personal grudges.
With external applicants, I think it is less important, but it doesn't usually hurt. I suppose you might consider it valuable to keep some of the stunning idiots in the industry in hopes that they will work with your competitors. But, you may eventually work with them too. And, you will have to maintain their code. Probably safer for everybody just to point out to them how clueless they are.
And, when I'm away from my day job, I do theater stuff. I was recently involved in some auditions to expand an improv troupe I am in. Not everybody got individual commentary, but the folks dismissed in the first round did at least get a general explanation. Everybody who made it past the first cut got an explanation of what impressed the director, and what he thought they could most work on - both the folks who made it and those who didn't. Personally, I wish we could have taken a little more time to offer personal advice to some of the folks in the first round. I would have liked suggesting that the hot chicks take classes that I can sit in on and watch them learn. Especially one blonde. I tried to convince the director that she should join the troupe and just not be allowed to say anything. I would have been cool with that.
This is always a good related subject to mention. I'm still readily awaiting being able to order one of the FPGA cards to show my support!
Those specs don't seem to include a lot of the actual specs... Which I find odd. If this were a mobile OS-X device of my dreams, I'd know what sort of processor it has, and what sort of graphics hardware, etc. Unfortunately, it seems like they may plan on pushing it like a phone. Locked down, limited access to development tools, etc. That really could completely kill my interest in the device. Why is the fact that it runs OS X interesting to me if I can't easily program it? Am I supposed to get some pointless pride in the fact that in runs a particular kernel I can't interact with? Is it supposed to be more impressive to me than a phone running some other kernel if I can't actually see any difference?
No, they aren't. But, they are trying to convince everybody to do it wrong in order to sell more hardware. I think they want to have a smart server that transcodes video in real time to deliver it to the front end. Personally, I think that's generally pretty stupid. I mean, I probably downloaded the video off the internet anyway, so it was most likely compressed reasonably small for internet transport. If I try to recompress it on the fly, then I will probably just make bigger uglier artifacts in the video.
Sure, there are occasionally files that are not well compressed that live on my server and I want to play wirelessly on a front end machine. It might be reasonable to have the front end try to play something, realise it can't, and then volunteer to download locally in the background (if there is local storage), and also give the option of transcoding and playing the result. But, I find that such things are pretty rare. I've even played raw DVD files (MPEG-2) off my server w/ 802.11g without problems. The minimally compressed files that I do tend to have that actually wouldn't play over wireless are either funny scinetific simulation videos that I have downloaded, or my own work-in progress stuff that I'm not ready to compress yet. Neither case is common enough for a home user to need to worry about it in the Mom and Pop media server.
Of course... I use an Alpha Server as my home media storage box, so I guess I can't really complain about people using more machine than they need...
Well, no, Apple never used POWER6 specifically, but they did use PPC, and IBM's current marketing literature says that PowerPC is POWER. They used to say POWER was PowerPC, but anyhow, according to:
http://www-03.ibm.com/chips/power/aboutpower/
"Power Architecture encompasses PowerPC®, POWER4(TM) and POWER5(TM) processors."
So, Apple won't use POWER6. Apple never used the earlier "POWER" branded chips because they never built any systems that would use them. Macs just aren't that big. Also, Apple has switched completely to X86, and hasn't bothered to really keep alive any hope of a mixed platform for the Mac. But, if Apple did move to POWER6, it would count as going back to the old architecture, rather than moving to a brand new one.
From everything I have read, it's pretty much exactly what the article said in the summary. No accelerated graphics. You also don't get full access to the hard drive, or direct access to any of the hardware. All the hardware works except for GPU, for which you get just a dumb frame buffer. Not sure exactly what you are asking when you want to know how bad it is.
A fair and valid point! Indeed, there are oodles of chips sold that aren't X86. Your car may have a dozen small computers, or more. Your cell phone, your iPod, your PDA, your Wii, your router, etc don't have X86. When you get to exotic DSP's and microcontrollers, it gets hard to decide where exactly you draw the line when determining how many processors are actually sold.
I sort of assumed the conversation was desktop specific because the question was "Why is x86 the winner" rather than "Is x86 the winner?" The desktop is certainly the most visible segment, but it's also the *only* one where x86 is really the winner! OTOH, the desktop (And I'm being fuzzy with terms and counting desktop, and some server, and workstation, etc., as the "desktop" uebercategory) tends to be an extremely lucrative segment of the market. It has much higher volume than crazy supercomputer chips, and the chips tend to cost much more than the seven cent 8 bitters that wind up in your singing greeting cards and whatnot.
If I had been thinking about this, I probably should have included some discussion of this in my previous comment, thanks for pointing it out!
In 2000, we realised that things were fucked up. Seriously, they really were. They had been for a long time, and we finally realised it. Florida shouldn't have gone down the way that it did.
So, after that mess happened, rather than a careful introspective discussion about the best way to ensure that democracy was served, some slick saleslobbyists showed up and the conversation went something like this:
"We have to do something!"
"This is something."
"We have to do this!"
Because the congresscritters knew that if they didn't look busy and make some noise at a press conference about how they passed legislation to save America and Freedom and that Democracy thingie, then they might get fired. They knew that the voters wouldn't read the bills. They knew the voters wouldn't deeply contemplate the bills. They also knew that if all hell did break lose, they would have another chance to have a press conference, pass another bill, and make some noise about how they saved that Democracy thingie that everybody talks about.
One perspective on the question:
Non x86 architectures are certainly not inherently better clock for clock. That's a matter of specific chip designs more than anything else. The P4 was a fairly fast chip, but miserable clock for clock against a G4. An Athlon however, was much closer to a G4. (Remember kids, not all code takes advantage of SIMD like AltiVec!) And, the G4 wasn't very easy get bring to super high clock rates. The whole argument of architectural elegance no longer applies.
The RISC Revolution started at a time when decoding an ugly architecture like VAX or x86 would require a significant portion of the available chip area. The legacy modes of x86 significantly held back performance because the 8086 and 80286 compatibility areas took up space that could have been used for cache or floating point hardware, or whatever. Then, transistor budgets grew. People stopped manually placing individual transistors, and then they stopped manually fiddling with individual gates for the most part. Chips grew in transistor count to the point where basically, nobody knew what to do with all the extra space. When that happened, x86 instruction decoding became a tiny area of the chip. Removing legacy cruft from x86 really wouldn't have been a significant design win after about P6/K7.
Instead of being a design win, the fixed instruction length of the RISc architectures no longer meant improved performance through simple decoding. They meant that even simple instructions took as much space as average instructions. Really complex instructions weren't allowed, so they had to be implimented as multiple instructions. Something that was one byte on x86 was always exactly 4 bytes on MIPS. Something that was 12 bytes on x86 might be done as four instruction on MIPS, and thus take 16 bytes. So, effective instruction cache sizes and effective instruction fetch bandwidth grew on X86 compared to purer RISC architectures.
At the same time, the gap between compute performance and memory bandwidth on all architectures was widening. Instruction fetch badwidth was irrelevent in the time of the PC XT, because RAM fetches could actually be done in like a single cycle. Less that it takes to get to SRAM on-chip caches today. But, as time went on, memory accesses became more and more costly. So, if a MIPS machine was in a super tight loop that ran in L1 cache, it might be okay. But, it it was just going balls to the wall through sequential instructions, or a loop that was much larger than cache, then it didn't matter how fast it could compute the instructions if it couldn't fetch them quick enough to keep the processor fed. but, X86 absurdly ugly instruction encoding acted like a sort of compression, meaning that a loop was more likely to fit in a particularly sized cache, and that better use of instruction fetch bandwidth was made.
Also, people had software that ran on X86, so they bought 9000 bazillion chips to run it all. The money spent on those 9000 bazillion chips got invested in building better chips. If somebody had the sort of financial resources that Intel had to build a better chip, and they shipped it in that sort of volume, we might well se an extremely competetive desktop SPARC or ARM chip.