We stopped the draft? When? Admittedly, it's been awhile since I turned 18, but I distinctly recall having to fill out a selective service card.
We don't currently draft the military via selective service. That's not at all the same as stopping it entirely.
Other than that, can't disagree with pretty much anything you said. Frankly, I wouldn't be surprised if we lose more military personel during peacetime training due to mistakes than during wartime. But I certainly don't have the numbers to back that up.
Unfortunately neither Windows or Linux is capable of waking up from a deep sleep via the computer's clock
Sure they are. Numerous BIOS's have a "power on at time" option.
Of course, that's an option set in the BIOS, by a human. I'm presuming you want a system that would power itself on and off whenever it wanted to, which isn't available to my knowledge.
It's also rather silly... the power consumption involved in having a PC on 24/7 (no monitor) is relatively small and you'll save wear and tear on the disks too. If you have cause for your PC to be on 12 hours a day, may as well leave it on 24.
I bought (used, from a SYSOP) one of the earlier 9600 bps HST modems (which he had just replaced with the brand new 14.4 HSTs) and eventually upgraded to a dual 14.4 HST/V.32 modem. The first one cost me something like $300, which I paid with my own cash earned doing a real job (I think I was 14).
I did, however, have to convince my father that it would actually work on a normal phone line. It was a nice upgrade from the 2400 baud modem, which was in turn a nice upgrade from the 300 baud modem.
The downside? Trying to find a serial port card with a 16550 that worked on an XT. And I couldn't write directly to floppy anymore because at 9600 bps downstream the latency on a floppy was too long and would cause transmission errors. (Again, this is on an XT w/ 360K floppies, not a 1.44M floppy which could handle the transfer rate and the interrupts).
And this surprises you? I'm amazed it surprises anyone.
You have two parts of a humongous corporation at odds with one another. Sony Music may want draconian anti-copying systems put into place (note the word may there... people who think that all of the members of an organization agree with every position of that organization are, frankly, idiots). Sony Electronics wants no such thing -- it restrains their sale of goods, it adds to cost (not only the cost of whatever components, but the need to manufacture it differently for different markets, plus the increased marketing and support costs for the different markets), and discourages the consumer from purchasing a restricted device.
The two organizations probably have little or no contact with one another, and all the parent corporation really wants is increased profits from all its children. Corporations are amoral, and it's unlikely that Sony as a whole would make a definitive statement on this kind of thing this early on.
Unbreakable was good, but it simply couldn't live up to the hype generated as "from the director/writer of The Sixth Sense". I'm hoping that Shyamalan doesn't pull the same trick in Signs -- a last second twist.
It was a good movie though, and the twist did do some gut wrenching. But perhaps a bit overwrought.
As for Keaton - nobody thought he'd be a good Wayne/Batman... but he played it right, much as Toby Wright did. You don't have to play the superhero -- that's merely a foil for the really interesting story, which is that of a man (or woman) who has immense power. And what that does to the person.
All of the later Batman movies failed because the actors tried to play the hero instead of the average guy. The script writers also focused too much on the hero instead of how the person is handling the issues.
Really this is a good bit of the "comic book curse" that afflicts non-written attempts to do superheros right.
Repeated surveys have shown preference toward flat rate billing. Look at the telephone service in the US - local is completely flat rate. Long distance is rapidly moving to flat rate. Cell phones are even starting to become flat rate.
Yes, electrical, gas, and water are billed by consumption. They're also the utilities people complain about the most regarding cost -- because while you have a general idea how much power you use month to month, there are the occasional months that just blow you out of the water.
Your TV service IS flat rate unless you get PPV movies. Glad to see that your cable company is progressive (most are not), but you still know that the cable bill is going to be a set amount every month.
Gas? Food? Yeah, it's pay as you go. But gas is virtually always the same barring long trips. I know that I have to refuel my car every week and a half or so, and while the price is going to vary some, it's not going to take drastic swings. Same thing goes with food.
What you have to realize is that people have been paying flat rate for telecomms (consumer grade) for over a decade now. The phone service has been flat rate for even longer. If you try and change that and make it "pay as you go" then you're going to get a lot of flak and resistance toward that -- particularly since I've yet to see a plan that would actually LOWER the cost for the average user.
Yes, I fully agree that people whinging "it's my bandwidth" are off the deep end and have no tie to reality. But the cable and DSL companies aren't a whole lot better. They wanted people to buy the service and not use it.
Pretty much agree with you on the AnonWiFi bit though... there are "tactics" that could counter the script kiddies, but it's always a war of strike and counterstrike, with the providers a step behind.
No, he's a user. What's happened is that over the past 3.5 years he's gone from treating the computer as a toy to be played with to a tool to be used. Which happens to be how the vast majority of people view a computer. It's no different than a VCR, or a car, or a lawnmower - it should work, it should do what's expected, and it shouldn't require them to spend more time fixing it than using it.
A rather large portion of the Linux community just doesn't get this. It's totally contrary to the way they think about computers. They enjoy fiddling with the little bits to make it work better, or even at all.
I used to love fiddling around with the little bits as well. I ran OS/2 and Linux back in college and for awhile afterwards. But I wouldn't run it on my home PC now because I don't want to spend time making my PC work -- I want to spend my time working on my PC. Yeah, so that "work" is web surfing, or playing games, or balancing my checkbook, or whatever. It's still a helluva lot easier under Windows than Linux.
For a server? Hell yes, go Linux or another *nix. And I'd much rather code in Unix than in Windows (and, thankfully, I do - every day). Assuming, of course, I don't have any bugs in my code. Spare me from Unix debuggers (we run AIX currently... both dbx and VisualAge suck with templates). But that, in and of itself, can be an incentive to code things right (akin to getting electrical shocks every time you do something wrong... not a great way to go about things, but surprisingly effective).
Odds are 3 years from now he'll still be using Windows. Why? Because it does what he needs with a minimal amount of work on his part. The drivers will be there, the games will/b> run, and by and large all of the apps will work as expected, in a similar fashion, and not have critical things like fonts not show up.
Why do people immediately want to involve the government? Would you quit trying to spend other people's money?
The cable companies should, rightfully, either shut off service or charge by the byte. Frankly, charging by the byte is ludicrous for the residential sector -- virtually everything is moving to flat rate. Consumers like flat rate because it allows you to budget far more easily.
Want a "public" wireless network? Start a company, decide how much it'll cost, and bill subscribers appropriately. No, it's not this pipe dream of a free-for-all wireless network where you can plug in anonymously and do whatever you want. Maybe it'll be viable in 10-15 years, but right now it's not.
So instead we leave them where they are, exposing the nearby residents to that level of radiation, massively increasing the costs of guarding that many separate nuclear waste sites, and so forth?
Gods. Put some critical thinking into this for once.
Hrm... I haven't seen a Sparrow in Atlanta. But I now have an enviable 4 mile commute, none of it on the freeways.
Bicycling in Atlanta, however, is a death wish. I used to bike a lot as a kid, but back then the area wasn't as heavily populated and traffic wasn't nearly so bad. Atlanta has virtually no bike paths along major roads. Heck, we don't even have sidewalks in a lot of areas. That 4 mile commute would get me killed on a bike. Quickly. And the only alternative route I have would be something like 20-30 miles in length. With equally heavy (or heavier) traffic.
I have to ask, why look into alternative fuel solutions
Are you serious? Or merely trolling?
There are both ecological and geopolitical reasons to look into alternative fuel vehicles. Even if you don't buy into global warming (I'm reasonably neutral on it), it's indisputable that gasoline engines generate a good bit of pollution that isn't good for anyone. And there are some pretty damn good economic and political reasons to reduce dependance on Middle Eastern oil exports -- while the majority of US foreign oil does not come from this area, enough of it does that OPEC significantly affects both our economy and our foreign policy. And a rather good percentage of the member states deeply dislike the Western world.
Of course, there's the whole issue of petroleum being a non-renewable resource. It will run out eventually. When? Hell if I know. Hell if anyone knows. Back in the 70s it was supposed to run out in 20 years. More oil fields were discovered, better extraction and refining equipment was developed, and now it looks like at least another 50 years are in the ground.
Never underestimate the power of the free market.
Never forget that the invisible hand does not weigh all costs. Pollution and other long term effects are not adequately compensated for in the free market -- by the time they become problematic the groups that have profited will not be affected by the costs thereof.
Besides which, this IS the free market doing its job -- attempting to find alternatives to the older, increasingly expensive system.
If you were charged for the massive amount of pollution your car is belching out (1 quart of oil/tank? That's an issue) then maybe you'd realize all the hidden costs too.
As I recall, most car manufacturers will quickly void your engine warrantee if you use fuel with too high an ethanol content. I don't have my manual available at this moment, but I distinctly recall such a line in my new lawnmower manual (obviously a lawnmower engine is not equivalent to a car engine).
As for the ethical bit -- didn't know Canada was ruled by a dictator, but I do know they're where the US gets the majority of its foreign oil from.
For all you GNU/Linux junkies, ATI has been much more forthcoming in information for developing XF86 drivers than NVidia(proprietary binary only).
People who whine about this just prove how inane and stupid the free software movement can be.
Look at this, and virtually every other thread, regarding ATI. See how many complaints there are about the poor drivers despite the superior hardware. Contrast to nVidia, whose drivers support every card made going back 3 years, have great performance, and are usually very stable.
Now tell me again how there aren't trade secrets in that driver code?
Probably because it's cheaper to build new hardware than it is to try and optimize what's out there. Seriously.
If you can spend $1M on hardware development and come out with a new chip that's 20% faster or spend $1M on software and put out drivers that are 5% faster, where is the money better spent? Besides which, you can charge for the new hardware. Charging for the new drivers is not acceptable to consumers.
Freezing the hardware for "a couple years" is not acceptable. Companies will simply cease to exist. Upgrades are part of the business model of the industry, and that modern systems are capable of doing virtually all tasks home and business users would require of them is part of the reason for the technology bust in the past couple years.
Look, it's simple. If you don't play the latest and greatest games, or don't care if you can play those games at uber-high res with all the effects turned on, then you don't need to upgrade. And yes, you can generally play the new games just fine on an older computer (my system is an Athlon 750, 512MB PC133, 32MB GF2 and runs DS and NWN just fine. Plays Q3 just fine. Will it play UT2 just fine? I doubt it... but it's 2 years old now).
As for "nobody asked hardware companies to push" -- speak for yourself. Go look at the Doom3 demo. You simply can't do that on current hardware with any semblance of speed. Yeah, you can run it on a GF3/4/ATI 8500, but you'll have to run it at a lower resolution and turn off features. Run it on a GF2? An ATI 7500? An MX anything? Maybe. It won't have anywhere near the eye candy.
Once we're to photorealistic scenes being rendered in realtime with no drops below ~60 fps on large, outdoor scenes you can say we've gone far enough. And by that time we'll probably want 3D or something else that will continue to push the bleeding edge.
Until then, there is room for improvement. And there's a lot more room on the hardware side then there is on the software side.
Most programmers are lazy. I know I am. I know the other (senior) coder on our current project is. But it's a controlled laziness.
Being lazy means that instead of doing something "manually" we'll find some way to make the computer do it for us. Sure, doing it by hand may only take 5 minutes and we'll spend 15 minutes just trying to convince the computer to do it, but that's not the point. The point is that - 1) you just learned something new, 2) you avoided doing a repetitive task repeatedly (yes, yes, I know), 3) odds are you'll do the same damn thing, or something similar in the future, and then you'll know how to do it in 3 seconds.
Which is why programmers gravitate to complex but programmable tools, like vi/vim/emacs/ultraedit, to Unix, to scripting languages, etc.
It's also why we don't understand the user more often then not. Most users just want to get the job done. Sure, there may be a more efficient way to do it if you spend some time learning the system, but they either don't view the system as their job or have been burned by it too many times to trust it. So they do things the way they know, or the way they think it should work. Which is often not the way the lazy programmer thinks.
Laziness does have downsides though. Documentation tends to be banal. Comments are almost never as good as you'd like them to be 3 months later. And while algorithmic simplicity is a noble goal, it's often the wrong one - too simple of an algorithm can be wasteful of resources (bubble sort is a whole lot simpler than quicksort) simply because the programmer didn't learn the underlying hardware or didn't think through the implications of a design. Which leads to the issue of too simple of a design meaning massive rework as the system grows.
Of course the more experienced programmers have learned these things and discovered that spending the time up front means you can be lazier in the long run.
Knuth is one of these experienced programmers. There are gems of wisdom throughout his writings.
Wish the bastard didn't want us to work to find them though.
Nice guys who treat women with respect get to be friends. Jerks who show women no respect get laid
Shrug.
I met this girl, she was nice, I was nice back (although I did make a buffoon of myself the first time I met her). We talked, got to know each other, and awhile later we got married.
At no point did I not treat her with respect. I still do. And she respects and loves me for it. And yes, there is a good bit of sex involved in our relationship, thank you very much.
I guess it comes down to - do you want to just fuck something or do you want to have a relationship with another human being? If the former, I suggest renting some porno tapes or surfing the net and jacking off. If the latter, then you better respect the other person. Otherwise you're not deserving of respect yourself.
Because it's not cost effective to produce them. If you have a facility that can produce 30 GB platters, why should you downgrade it to one fifth of that capacity? There's a limited amount of fab space to all companies. If they dedicate it to a lower density platter then they will not be able to manufacture as many high density platters as they need. And when a 40G disk is ~$50 now, how much waste are you talking about here really? The marginal cost (or profit) is negligible.
Make the disk smaller? Sure. But you realize that the standards mean it's going to be 3 1/2" wide anyway, and that a smaller diameter platter means a slower drive, right?
Some of us still use SCSI just because of the extremely low CPU overhead it requires
Uh... and what speed CPU are you running? A 200 MHz Pentium2?
Modern computers have so much extra horsepower nowadays it's absurd. Even maxed out an ATA133 drive won't consume more than 2-3% of a CPU nowadays.
burning a disc for me in the background while I play a quake 3 engine game, without any fear of buffer underruns
Any decent computer built in the past 2 years can handle that too. IDE drives don't make platters like they used to -- they've got large buffers and use techniques to ensure no buffer underruns. Yeah, they use more CPU than SCSI does. See above.
I used to be a big SCSI advocate... and I finally replaced the old SCSI-2 drives I had in one of my PCs with IDE drives. I increased the storage, decreased the noise, and improved performance of the system. The cost to replace the old drives with newer SCSI equivalents would've been absurd - nearly $1000 since it meant a new controller too. Instead I spent $60 on a CD-RW (12x/32x/48x - the cheapest SCSI CD-RW was 10/12/20 for 3x the cost), used an older IDE drive I had spare, and seriously boosted my system.
Does IDE/ATA have issues? Sure. The whole lack of command reordering, one device on the bus at a time, etc. -- but none of these are ever going to impact a home user. It's becoming questionable if they significantly impact low-end servers too. If you're putting together a database or a big ass file server, yes, go SCSI/RAID and get the best you can afford. Otherwise start understanding that modern IDE is really not the same as the old, crappy IDE that evolved out of MFM/RLL.
Smaller cable? Pshaw... Sound like Martha Stewart of the Mobo set. Big cables, Baby
Bigger cables inhibit airflow. And while, yes, there are "round" IDE cables out there, they aren't as flexible as flat cables and are more prone to breaking at the connectors. All of these issues are solved by a serial standard.
I'm still of the mindset that parallel is better than serial, particularly where high bandwidth is concerned
Which is why all the high speed busses have moved to serial interfaces, right?
Yes, parallel means that you can toss more stuff over the wall at once. The problem is, you can't do that very fast or you start running into serious timing and EMI issues. High speed serial doesn't have as many timing issues, and while you do generate EMI with serial still (duh), at least you don't have to worry about causing interference on the next wire over -- you know, the one that's supposed to be handing the other side of the interface the next bit? If you have to shield every wire in a parallel cable from every other wire then you'd definitely get your big cables. As in 3-4 inches in diameter with a bend radius slightly larger than you are tall.
The only speed increase you see going from ATA66 to ATA100 to ATA133 is burst transfers from the cache. Go read the specs on the drives - not a single one has a disk to buffer transfer rate that exceeds ~50 MB/s (ATA33 maxes at 33 MB/s, 66 at 66 MB/s - get the picture yet?).
If the disk happens to have the data you request in its cache (most drives have either 2 MB or 8 MB of cache now) then yes, you'll max out the bus. For as much as 80 milliseconds (8 MB cache on an ATA100 bus). Wow. And then you're back to slogging the data off disk - which varies from ~20 MB/s (inner tracks) to ~50 MB/s (outer tracks).
You can improve this transfer rate in a couple ways - either go for higher spindle speeds or go for higher data density. SCSI drives do the former. Density keeps going up, but it hasn't jumped radically for a few years now, and there's nothing in the pipe that's going to make it jump again.
Go to large RAM drives of some sort? Sure... but you'll be paying roughly 300x as much for the same storage space. On the upside, FlashRAM is more than 300x faster. On the downside it's not available in anywhere close to the capacities you'd want.
Slashdot Mirror
We stopped the draft? When? Admittedly, it's been awhile since I turned 18, but I distinctly recall having to fill out a selective service card.
We don't currently draft the military via selective service. That's not at all the same as stopping it entirely.
Other than that, can't disagree with pretty much anything you said. Frankly, I wouldn't be surprised if we lose more military personel during peacetime training due to mistakes than during wartime. But I certainly don't have the numbers to back that up.
Unfortunately neither Windows or Linux is capable of waking up from a deep sleep via the computer's clock
Sure they are. Numerous BIOS's have a "power on at time" option.
Of course, that's an option set in the BIOS, by a human. I'm presuming you want a system that would power itself on and off whenever it wanted to, which isn't available to my knowledge.
It's also rather silly... the power consumption involved in having a PC on 24/7 (no monitor) is relatively small and you'll save wear and tear on the disks too. If you have cause for your PC to be on 12 hours a day, may as well leave it on 24.
Well, whoever it was they had pretty good access. The first 3 emails to my mindspring account were spam.
The spaminator service must be doing some good though - my address was obviously sold at some point, but it's not getting deluged in spam.
Newb.
:)
I bought (used, from a SYSOP) one of the earlier 9600 bps HST modems (which he had just replaced with the brand new 14.4 HSTs) and eventually upgraded to a dual 14.4 HST/V.32 modem. The first one cost me something like $300, which I paid with my own cash earned doing a real job (I think I was 14).
I did, however, have to convince my father that it would actually work on a normal phone line. It was a nice upgrade from the 2400 baud modem, which was in turn a nice upgrade from the 300 baud modem.
The downside? Trying to find a serial port card with a 16550 that worked on an XT. And I couldn't write directly to floppy anymore because at 9600 bps downstream the latency on a floppy was too long and would cause transmission errors. (Again, this is on an XT w/ 360K floppies, not a 1.44M floppy which could handle the transfer rate and the interrupts).
Wow.... that was a long time ago
And this surprises you? I'm amazed it surprises anyone.
You have two parts of a humongous corporation at odds with one another. Sony Music may want draconian anti-copying systems put into place (note the word may there... people who think that all of the members of an organization agree with every position of that organization are, frankly, idiots). Sony Electronics wants no such thing -- it restrains their sale of goods, it adds to cost (not only the cost of whatever components, but the need to manufacture it differently for different markets, plus the increased marketing and support costs for the different markets), and discourages the consumer from purchasing a restricted device.
The two organizations probably have little or no contact with one another, and all the parent corporation really wants is increased profits from all its children. Corporations are amoral, and it's unlikely that Sony as a whole would make a definitive statement on this kind of thing this early on.
There may very well be a feature length adaptation of HHGTTG sitting on a shelf somewhere. In fact, I bet there's dozens. In script form.
Once shooting begins, it's unlikely to stop. And you're going to have something to put out, even if only on direct-to-video.
A feature length adaptation already shot and sitting on a shelf? Not bloody likely.
Unbreakable was good, but it simply couldn't live up to the hype generated as "from the director/writer of The Sixth Sense". I'm hoping that Shyamalan doesn't pull the same trick in Signs -- a last second twist.
It was a good movie though, and the twist did do some gut wrenching. But perhaps a bit overwrought.
As for Keaton - nobody thought he'd be a good Wayne/Batman... but he played it right, much as Toby Wright did. You don't have to play the superhero -- that's merely a foil for the really interesting story, which is that of a man (or woman) who has immense power. And what that does to the person.
All of the later Batman movies failed because the actors tried to play the hero instead of the average guy. The script writers also focused too much on the hero instead of how the person is handling the issues.
Really this is a good bit of the "comic book curse" that afflicts non-written attempts to do superheros right.
Repeated surveys have shown preference toward flat rate billing. Look at the telephone service in the US - local is completely flat rate. Long distance is rapidly moving to flat rate. Cell phones are even starting to become flat rate.
Yes, electrical, gas, and water are billed by consumption. They're also the utilities people complain about the most regarding cost -- because while you have a general idea how much power you use month to month, there are the occasional months that just blow you out of the water.
Your TV service IS flat rate unless you get PPV movies. Glad to see that your cable company is progressive (most are not), but you still know that the cable bill is going to be a set amount every month.
Gas? Food? Yeah, it's pay as you go. But gas is virtually always the same barring long trips. I know that I have to refuel my car every week and a half or so, and while the price is going to vary some, it's not going to take drastic swings. Same thing goes with food.
What you have to realize is that people have been paying flat rate for telecomms (consumer grade) for over a decade now. The phone service has been flat rate for even longer. If you try and change that and make it "pay as you go" then you're going to get a lot of flak and resistance toward that -- particularly since I've yet to see a plan that would actually LOWER the cost for the average user.
Yes, I fully agree that people whinging "it's my bandwidth" are off the deep end and have no tie to reality. But the cable and DSL companies aren't a whole lot better. They wanted people to buy the service and not use it.
Pretty much agree with you on the AnonWiFi bit though... there are "tactics" that could counter the script kiddies, but it's always a war of strike and counterstrike, with the providers a step behind.
No, he's a user. What's happened is that over the past 3.5 years he's gone from treating the computer as a toy to be played with to a tool to be used. Which happens to be how the vast majority of people view a computer. It's no different than a VCR, or a car, or a lawnmower - it should work, it should do what's expected, and it shouldn't require them to spend more time fixing it than using it.
A rather large portion of the Linux community just doesn't get this. It's totally contrary to the way they think about computers. They enjoy fiddling with the little bits to make it work better, or even at all.
I used to love fiddling around with the little bits as well. I ran OS/2 and Linux back in college and for awhile afterwards. But I wouldn't run it on my home PC now because I don't want to spend time making my PC work -- I want to spend my time working on my PC. Yeah, so that "work" is web surfing, or playing games, or balancing my checkbook, or whatever. It's still a helluva lot easier under Windows than Linux.
For a server? Hell yes, go Linux or another *nix. And I'd much rather code in Unix than in Windows (and, thankfully, I do - every day). Assuming, of course, I don't have any bugs in my code. Spare me from Unix debuggers (we run AIX currently... both dbx and VisualAge suck with templates). But that, in and of itself, can be an incentive to code things right (akin to getting electrical shocks every time you do something wrong... not a great way to go about things, but surprisingly effective).
Odds are 3 years from now he'll still be using Windows. Why? Because it does what he needs with a minimal amount of work on his part. The drivers will be there, the games will/b> run, and by and large all of the apps will work as expected, in a similar fashion, and not have critical things like fonts not show up.
Why do people immediately want to involve the government? Would you quit trying to spend other people's money?
The cable companies should, rightfully, either shut off service or charge by the byte. Frankly, charging by the byte is ludicrous for the residential sector -- virtually everything is moving to flat rate. Consumers like flat rate because it allows you to budget far more easily.
Want a "public" wireless network? Start a company, decide how much it'll cost, and bill subscribers appropriately. No, it's not this pipe dream of a free-for-all wireless network where you can plug in anonymously and do whatever you want. Maybe it'll be viable in 10-15 years, but right now it's not.
So instead we leave them where they are, exposing the nearby residents to that level of radiation, massively increasing the costs of guarding that many separate nuclear waste sites, and so forth?
Gods. Put some critical thinking into this for once.
First off, there's 8766 hrs/yr (assuming 365.25 days/year), not 8544.
As for percentages:
- 90% uptime: Down 876.6 hours, or 36.525 days. 3 days per month.
- 99% uptime: 87.66 hours, or 3.65 days.
- 99.9% uptime: 8.77 hours
- 99.99% uptime: 0.877 hours, or 52.6 minutes
- 99.999% (5 nine's!) uptime: 0.0877 hours, or 5.26 minutes
- 99.9999% uptime: 0.0088 hours, or 0.526 min, or 31.56 seconds
- 99.99999% uptime: 0.0526 min, or ~3 seconds
- 99.999999% uptime: ~1/3 second outage per year
I suspect that most people could figure out the mystical, magical mathematic relations after the 90% to 99% jump....Hrm... I haven't seen a Sparrow in Atlanta. But I now have an enviable 4 mile commute, none of it on the freeways.
Bicycling in Atlanta, however, is a death wish. I used to bike a lot as a kid, but back then the area wasn't as heavily populated and traffic wasn't nearly so bad. Atlanta has virtually no bike paths along major roads. Heck, we don't even have sidewalks in a lot of areas. That 4 mile commute would get me killed on a bike. Quickly. And the only alternative route I have would be something like 20-30 miles in length. With equally heavy (or heavier) traffic.
I have to ask, why look into alternative fuel solutions
Are you serious? Or merely trolling?
There are both ecological and geopolitical reasons to look into alternative fuel vehicles. Even if you don't buy into global warming (I'm reasonably neutral on it), it's indisputable that gasoline engines generate a good bit of pollution that isn't good for anyone. And there are some pretty damn good economic and political reasons to reduce dependance on Middle Eastern oil exports -- while the majority of US foreign oil does not come from this area, enough of it does that OPEC significantly affects both our economy and our foreign policy. And a rather good percentage of the member states deeply dislike the Western world.
Of course, there's the whole issue of petroleum being a non-renewable resource. It will run out eventually. When? Hell if I know. Hell if anyone knows. Back in the 70s it was supposed to run out in 20 years. More oil fields were discovered, better extraction and refining equipment was developed, and now it looks like at least another 50 years are in the ground.
Never underestimate the power of the free market.
Never forget that the invisible hand does not weigh all costs. Pollution and other long term effects are not adequately compensated for in the free market -- by the time they become problematic the groups that have profited will not be affected by the costs thereof.
Besides which, this IS the free market doing its job -- attempting to find alternatives to the older, increasingly expensive system.
If you were charged for the massive amount of pollution your car is belching out (1 quart of oil/tank? That's an issue) then maybe you'd realize all the hidden costs too.
As I recall, most car manufacturers will quickly void your engine warrantee if you use fuel with too high an ethanol content. I don't have my manual available at this moment, but I distinctly recall such a line in my new lawnmower manual (obviously a lawnmower engine is not equivalent to a car engine).
As for the ethical bit -- didn't know Canada was ruled by a dictator, but I do know they're where the US gets the majority of its foreign oil from.
Funny... seems that everyone but nVidia sucks at writing 3D driver code, since Matrox has similar issues.
Oh, but I suppose that nVidia should just give away whatever tricks they've built into the drivers to eek more performance out of lesser hardware.
Riiiiiight.
For all you GNU/Linux junkies, ATI has been much more forthcoming in information for developing XF86 drivers than NVidia(proprietary binary only).
People who whine about this just prove how inane and stupid the free software movement can be.
Look at this, and virtually every other thread, regarding ATI. See how many complaints there are about the poor drivers despite the superior hardware. Contrast to nVidia, whose drivers support every card made going back 3 years, have great performance, and are usually very stable.
Now tell me again how there aren't trade secrets in that driver code?
Probably because it's cheaper to build new hardware than it is to try and optimize what's out there. Seriously.
If you can spend $1M on hardware development and come out with a new chip that's 20% faster or spend $1M on software and put out drivers that are 5% faster, where is the money better spent? Besides which, you can charge for the new hardware. Charging for the new drivers is not acceptable to consumers.
Freezing the hardware for "a couple years" is not acceptable. Companies will simply cease to exist. Upgrades are part of the business model of the industry, and that modern systems are capable of doing virtually all tasks home and business users would require of them is part of the reason for the technology bust in the past couple years.
Look, it's simple. If you don't play the latest and greatest games, or don't care if you can play those games at uber-high res with all the effects turned on, then you don't need to upgrade. And yes, you can generally play the new games just fine on an older computer (my system is an Athlon 750, 512MB PC133, 32MB GF2 and runs DS and NWN just fine. Plays Q3 just fine. Will it play UT2 just fine? I doubt it... but it's 2 years old now).
As for "nobody asked hardware companies to push" -- speak for yourself. Go look at the Doom3 demo. You simply can't do that on current hardware with any semblance of speed. Yeah, you can run it on a GF3/4/ATI 8500, but you'll have to run it at a lower resolution and turn off features. Run it on a GF2? An ATI 7500? An MX anything? Maybe. It won't have anywhere near the eye candy.
Once we're to photorealistic scenes being rendered in realtime with no drops below ~60 fps on large, outdoor scenes you can say we've gone far enough. And by that time we'll probably want 3D or something else that will continue to push the bleeding edge.
Until then, there is room for improvement. And there's a lot more room on the hardware side then there is on the software side.
Most programmers are lazy. I know I am. I know the other (senior) coder on our current project is. But it's a controlled laziness.
Being lazy means that instead of doing something "manually" we'll find some way to make the computer do it for us. Sure, doing it by hand may only take 5 minutes and we'll spend 15 minutes just trying to convince the computer to do it, but that's not the point. The point is that - 1) you just learned something new, 2) you avoided doing a repetitive task repeatedly (yes, yes, I know), 3) odds are you'll do the same damn thing, or something similar in the future, and then you'll know how to do it in 3 seconds.
Which is why programmers gravitate to complex but programmable tools, like vi/vim/emacs/ultraedit, to Unix, to scripting languages, etc.
It's also why we don't understand the user more often then not. Most users just want to get the job done. Sure, there may be a more efficient way to do it if you spend some time learning the system, but they either don't view the system as their job or have been burned by it too many times to trust it. So they do things the way they know, or the way they think it should work. Which is often not the way the lazy programmer thinks.
Laziness does have downsides though. Documentation tends to be banal. Comments are almost never as good as you'd like them to be 3 months later. And while algorithmic simplicity is a noble goal, it's often the wrong one - too simple of an algorithm can be wasteful of resources (bubble sort is a whole lot simpler than quicksort) simply because the programmer didn't learn the underlying hardware or didn't think through the implications of a design. Which leads to the issue of too simple of a design meaning massive rework as the system grows.
Of course the more experienced programmers have learned these things and discovered that spending the time up front means you can be lazier in the long run.
Knuth is one of these experienced programmers. There are gems of wisdom throughout his writings.
Wish the bastard didn't want us to work to find them though.
Nice guys who treat women with respect get to be friends. Jerks who show women no respect get laid
Shrug.
I met this girl, she was nice, I was nice back (although I did make a buffoon of myself the first time I met her). We talked, got to know each other, and awhile later we got married.
At no point did I not treat her with respect. I still do. And she respects and loves me for it. And yes, there is a good bit of sex involved in our relationship, thank you very much.
I guess it comes down to - do you want to just fuck something or do you want to have a relationship with another human being? If the former, I suggest renting some porno tapes or surfing the net and jacking off. If the latter, then you better respect the other person. Otherwise you're not deserving of respect yourself.
Because it's not cost effective to produce them. If you have a facility that can produce 30 GB platters, why should you downgrade it to one fifth of that capacity? There's a limited amount of fab space to all companies. If they dedicate it to a lower density platter then they will not be able to manufacture as many high density platters as they need. And when a 40G disk is ~$50 now, how much waste are you talking about here really? The marginal cost (or profit) is negligible.
Make the disk smaller? Sure. But you realize that the standards mean it's going to be 3 1/2" wide anyway, and that a smaller diameter platter means a slower drive, right?
Some of us still use SCSI just because of the extremely low CPU overhead it requires
Uh... and what speed CPU are you running? A 200 MHz Pentium2?
Modern computers have so much extra horsepower nowadays it's absurd. Even maxed out an ATA133 drive won't consume more than 2-3% of a CPU nowadays.
burning a disc for me in the background while I play a quake 3 engine game, without any fear of buffer underruns
Any decent computer built in the past 2 years can handle that too. IDE drives don't make platters like they used to -- they've got large buffers and use techniques to ensure no buffer underruns. Yeah, they use more CPU than SCSI does. See above.
I used to be a big SCSI advocate... and I finally replaced the old SCSI-2 drives I had in one of my PCs with IDE drives. I increased the storage, decreased the noise, and improved performance of the system. The cost to replace the old drives with newer SCSI equivalents would've been absurd - nearly $1000 since it meant a new controller too. Instead I spent $60 on a CD-RW (12x/32x/48x - the cheapest SCSI CD-RW was 10/12/20 for 3x the cost), used an older IDE drive I had spare, and seriously boosted my system.
Does IDE/ATA have issues? Sure. The whole lack of command reordering, one device on the bus at a time, etc. -- but none of these are ever going to impact a home user. It's becoming questionable if they significantly impact low-end servers too. If you're putting together a database or a big ass file server, yes, go SCSI/RAID and get the best you can afford. Otherwise start understanding that modern IDE is really not the same as the old, crappy IDE that evolved out of MFM/RLL.
Smaller cable? Pshaw... Sound like Martha Stewart of the Mobo set. Big cables, Baby
Bigger cables inhibit airflow. And while, yes, there are "round" IDE cables out there, they aren't as flexible as flat cables and are more prone to breaking at the connectors. All of these issues are solved by a serial standard.
I'm still of the mindset that parallel is better than serial, particularly where high bandwidth is concerned
Which is why all the high speed busses have moved to serial interfaces, right?
Yes, parallel means that you can toss more stuff over the wall at once. The problem is, you can't do that very fast or you start running into serious timing and EMI issues. High speed serial doesn't have as many timing issues, and while you do generate EMI with serial still (duh), at least you don't have to worry about causing interference on the next wire over -- you know, the one that's supposed to be handing the other side of the interface the next bit? If you have to shield every wire in a parallel cable from every other wire then you'd definitely get your big cables. As in 3-4 inches in diameter with a bend radius slightly larger than you are tall.
rpm is not going to change anything in terms of transfer rates, the higher spindle speed only affects access times
Untrue. Yes, increasing the spindle speed will cut down average latency times because there's less wait for the next revolution of the disk.
It can also increase transfer speeds - you have more media passing under the head so you can read or write more data at the same data density.
The only speed increase you see going from ATA66 to ATA100 to ATA133 is burst transfers from the cache. Go read the specs on the drives - not a single one has a disk to buffer transfer rate that exceeds ~50 MB/s (ATA33 maxes at 33 MB/s, 66 at 66 MB/s - get the picture yet?).
If the disk happens to have the data you request in its cache (most drives have either 2 MB or 8 MB of cache now) then yes, you'll max out the bus. For as much as 80 milliseconds (8 MB cache on an ATA100 bus). Wow. And then you're back to slogging the data off disk - which varies from ~20 MB/s (inner tracks) to ~50 MB/s (outer tracks).
You can improve this transfer rate in a couple ways - either go for higher spindle speeds or go for higher data density. SCSI drives do the former. Density keeps going up, but it hasn't jumped radically for a few years now, and there's nothing in the pipe that's going to make it jump again.
Go to large RAM drives of some sort? Sure... but you'll be paying roughly 300x as much for the same storage space. On the upside, FlashRAM is more than 300x faster. On the downside it's not available in anywhere close to the capacities you'd want.