I myself have brewed (so far only one batch - I get hangovers VERY easily and am trying to lose 20 pounds so don't drink very often.) a batch so far.
The basics are simple, but if you get into it things can become very complex.
Advanced beer brewing makes an excellent geek hobby, as it mixes biology, chemistry, thermodynamics, and electronics. Electrical + mechanical engineering skills are excellent for controlling and building advanced brewing equipment such as RIMS brewing systems. I will likely not pass the kettle-on-stove stage for another year or so myself though, but I definately want to mix some AVR hacking into the hobby soon. I may make a lagering fridge with some Peltiers and an AVR first.
For 5-6 years, a human driver could easily outbrake any ABS system. For many years, ABS systems sucked. And they still can barely compete with a human driver with an IQ higher than room temperature.
I don't mind power steering and automatic transmissions. In general they are robust mechanical systems, not prone to code bugs. The few auto transmissions I've seen that were electronically controlled (Chrysler 4-speed automatic) are reliablity nightmares. Note that Chrysler 3-speeds are purely mechanical and are considered nearly impossible to damage...
What about oncoming traffic coming around a bend? The radar will see it even though it's something you won't hit.
Without being EXTREMELY expensive and having a very large antenna, radar doesn't have the narrow beamwidth necessary to distinguish between stuff directly in front of you and things that are stationary but in an adjacent lane.
A laser-based system might work.
Don't forget approaching the bottom of a hill. The radar would think you're about to hit some solid substance, even though it's just the road leveling out.
"The Santa Cruz Operation, Inc. P.O. Box 7745 San Francisco, CA 94120-7745 United States of America"
The current SCO is NOT the same as the former SCO. (Now the Tarantella Group.)
If you read the article, you'll see that the current SCO was formerly Caldera. Caldera bought the Unix rights from SCO, the old SCO became Tarantella (which was one of their products IIRC...), and then Caldera renamed to The SCO Group.
That source offer was made by people with no management connection to McBride...
I ordered a GeForce 4 Ti4600 (PNY Verto) from them, and the card was defective. Within 20-30 minutes of bootup, text in console mode would be corrupted every few characters, and in X (or in Windows), there would be vertical yellow lines every few pixels. Probably a cold solder joint somewhere...
So I first tried dealing with PNY support. Their response: Take it to the vendor for a return.
So I try to return it to Nova. They tell me to go talk to PNY.
It took two weeks to receive an RMA number, after which I sent back the card, demanding a refund. (I did NOT want a replacement, as at this point I wanted to end all further association with both PNY and Nova due to the horrendous customer support both of them had.)
A month later and still no credit. It took me 4-5 calls before I reached someone with a clue, who told me that they had tried the unit and it wasn't defective, and that a replacement was out of stock. I told him, no, I didn't want a replacement, I never wanted to deal with PNY again. In the end I got ripped off for a $30 restocking fee for a defective product and a month of hassles.
Stay away from Nova Computech. Don't know about TigerDirect, although they've had bad press here before. Walmart's return policies in-store are great, I don't know about online. I'm sad to say this, but Wally World is likely your best bet for one of these systems. (I'll be ordering one shortly - It's perfectly suited as a basic personal webserver.)
Of how Boeing does things right, look up the Gimli Glider.
The only reason that plane was able to land was because Boeing engineers were so damn conservative and had backups for every possible contingency, including dual-engine failure (Which the airline industry and Boeing considered a near impossibility, but included a backup for anyway.)
I don't care if the A380 might have a better "concept". Airbus engineers take shortcuts in their design of metal-composite joints that are just plain Bad Engineering, and I'm going to avoid flying on an Airbus jet if at all possible after what I've heard regarding their tail designs from someone familiar with the industry.
I refer you to the Brooklyn crash a few months after 9/11 where the tail of an Airbus plane basically fell apart...
You should be rejoicing
on
42-Volt Autos
·
· Score: 1
Ever considered why power companies transmit their power at thousands of volts instead of 120 system-wide? Even though it means transformers wasting some power at the ends?
Simple. I^2*R losses. The power loss through a wire is the square of the current times the resistance. If you triple the voltage (Note: the system proposed is a 36/42 system, triple that of our current 12/14 system. First being nominal battery voltage, second being charging voltage.), you have 1/3 the current to deliver the same power. This means that resistive losses in the automotive wiring is 1/9 what it was previously.
Also, with a 36V system, it may become practical to drive the valves with solenoids rather than camshafts. All of a sudden the engine's "sweet spot" of maximum efficiency becomes MUCH wider. (If implemented properly, the engine could run at max efficiency at nearly any load/RPM combo with valve timing that is fully electronically controlled.)
One thing I've seen (and happen to have, and like), is a "slow" jumpstart device. Instead of being able to leech enough power to start the engine in one instant, it is designed to trickle charge the battery for 10-20 minutes, just enough to start the engine. It works (believe it or not) through the cigarette lighter socket, and works quite well. (You just have to wait some time before trying to start the car.)
It would be trivial for someone to start making these with a step-up DC/DC converter.
The nicest thing is that partially charging a dead battery needs little to no regulation at all...
While I don't have kids yet, I was recently on vacation with the extended family, including little kids.
For various reasons, for most of vacation I was not able to use my laptop in the car (No room, also didn't want to have something that valuable being left in a parking lot in the middle of nowhere while we were hiking.)
On the way back we had a 4-5 hour drive from Grand Junction, CO to Denver. (Couldn't get a flight into GJ.)
Pulled out my laptop and some DVDs... Quietest and calmest ride of the entire trip. (Needless to say, during the week some of the rides were pure hell thanks to my cousin's hyperactive kids.)
Without modern CPUs, home video editing would not be practical (and hence the market for DV camcorders would be much smaller.)
You obviously haven't tried compressing 2 hours of video into DVD-quality MPEG-2, let alone trying to compress it into DivX to send home videos to some relatives.
Would we really need more than 800 MHz on a home computer? I have a 1.7 GHz P4 laptop, and a 1.1 GHz Athlon. Upgrading to a Barton 3000+ (2 GHz or so actual clockrate, but much more efficient per clock than my current TBird) would take my 14-hour encoding jobs down to 7 hours. A difference between taking most of the day and running while I sleep.
And reencoding 1080i HDTV recordings into a more managable size... yikes... I've had 24 hour encoding jobs before.
So my suggestion: Go buy a DV camcorder, or an HDTV tuner card. I guarantee you you'll be desperate to upgrade that poke-ass 800 MHz machine in under two weeks.
I would think that they'd almost HAVE to use these in conjunction to get any benefits.
While they may be able to get 30-35% improvement for PMOS alone, and 20-25% for NMOS (or was it the other way around?), if implemented in a chip, improved PMOS transistors without improved NMOS would result in almost no maximum speed improvements. (It would likely improve power consumption, but not as much as the speed benefit of the transistor itself.)
This is because any given gate involves both NMOS and PMOS transistors. The most basic gate type is an inverter, which consists of one NMOS and one PMOS. Improve only one and you'll improve either the turn-on or turn-off time of the gate quite a bit, but not both (You'll get some improvements to both, but it will be unbalanced). So the unimproved time will be your speed barrier. Use both technologies and you're improving the whole gate, not just half of it.
What I would love to see is if someone could come up with a gate architecture that could provide great improvements in fan-out capability with minimal penalties in gate delay. (Fanout is why memory speed hasn't kept up with CPU speed - The larger the memory gets the more loaded the address drive lines get, and high fanout = slow speed.)
I did some scientific computing work a few summers ago.
At that time, my new iPaq (Not the PDA, the business desktop system type that seems to be relatively unknown) was competitive with some much more expensive (but 2-3 years old) high-end computing hardware. My boss was impressed at how well the $1000 system he bought for his summer intern performed. And that system was only 500 MHz.
1 month into my internship, I started running simulations on that machine. Some only ran for 10 minutes, but each batch would include 20-30 runs that got progressively longer. The final runs in a batch would be over a day each - To collect the one dataset I needed to work with took over a month of total CPU time.
With a modern $600ish system I could finish those computations in *under a week*
I wasn't storage-bound. My end data from each batch was only a 200x200 matrix.
Yup. 1 day or more to generate a 200x200 matrix.:)
And do some video editing (esp. compressing it to MPEG-2 or DivX)
You'll change your tune.
With some of the more advanced video compression algorithms (DivX for example - Yes it has legit uses, great for distributing home videos to relatives.), a 10% increase in CPU speed can mean an hour or two off of your compression time.
With exceptions, often 12 year olds are the least likely to succeed (or even play) an MMOG.
Because of the social interactions and complexity of most MMOGs, those who have good communications skills and are intelligent are most likely to succeed.
Some MMOGs have higher ages than others. DAoC, which has a large amount of hack-n-slash, seems to have a greater following in younger circles than EVE Online, which seems to have a much higher average age, partly due to the fact that the game is insanely complex and to really succeed requires a basic grasp of economic principles. (The truly rich people are those who can find good trade routes... Which is very difficult and requires a lot of thinking and calculations.)
CDMA2000 1xRTT (Sprint Vision, Verizon Express Network) is not 3G. It's considered 2.5G.
3G standards include UMTS (only 1-2 test networks in the US, and so far massive failures due to handset problems, especially in the battery life/heat arena, in Europe and Japan.), and CDMA2000 1xEV-DO (Which is available in Korea and also was rolled out by KDDI in Japan I believe.)
Where 802.11 falls flat on its face
on
Is 3G Irrelevant?
·
· Score: 2, Insightful
802.11 has some support for handoffs between APs, but not on a large scale, and not if it's occuring often.
It's also not designed to handle rapidly moving stations. Once you start going faster than walking speed, multipath fading and doppler shifts make things fun. Dealing with users traveling at 55 MPH is one of the biggest challenges of cellular network designers, and it just gets to be more and more fun as the bandwidth increases.
This isn't targeted at stationary users. It's targeted at people on the go. (Not drivers, mind you, but carpool passengers and people on buses and trains, etc.)
It might also prove useful for police work - Police departments have a use for high-speed mobile data. (I believe that the biggest customer of the Ricochet remnants in Denver is one of the local police departments. I know that there's a metropolitan wireless network somewhere that is used heavily by the local police.) It could (in theory) provide an off-the-shelf communications solution for a low-cost civilian Predator equivalent, which law enforcement entities would kill for. All of the advantages of a police helicopter without the exorbitant expenses...
Depends on the exact rules on that spectrum licensing.
Right now the providers are all begging the European government to ditch UMTS, which has proved to be a nightmare. If the gov. lets them, they can retask the spectrum to 4G technologies such as Flarion's OFDM-based network architecture.
FYI, Flarion has some excellent whitepapers on their site describing their tech, and the idea of OFDM in general.
If all goes well I'll be a Flarion employee in a month or so. (Getting laid off, applying for a position at Flarion which is 20 minutes away from here, and coming into the application with great references, as my current company and Flarion are both spinoffs from Lucent's wireless division in Whippany.) So I've done quite a bit of research into the company and their tech.:)
FYI, European digital TV broadcasts use OFDM modulation. The iBiquity IBOC radio broadcasting standard (Yet Another Lucent Spinoff) uses OFDM. IBOC was recently approved by the FCC as the standard for digital audio broadcasting in the US, although unfortunately for iBiquity, the economic downturn has caused nearly all broadcasters to cancel upgrade plans for the time being.
a) As the article mentioned, some readers use multiple read heads to achieve higher read speeds with lower spin rates. So higher than 48x is indeed possible.
b) The latest IDE standard can push 133 MB/sec. Even taking into account overhead, an ATA100 7200 RPM hard drive can easily do 40 MB/sec in the real world. (That's what my drive does.)
Since 1x CD is 150 KB/sec, 48x is only 5-6 MB/sec, a small fraction of what even ATA33 is capable of, let alone ATA133.
Slashdot Mirror
I myself have brewed (so far only one batch - I get hangovers VERY easily and am trying to lose 20 pounds so don't drink very often.) a batch so far.
The basics are simple, but if you get into it things can become very complex.
Advanced beer brewing makes an excellent geek hobby, as it mixes biology, chemistry, thermodynamics, and electronics. Electrical + mechanical engineering skills are excellent for controlling and building advanced brewing equipment such as RIMS brewing systems. I will likely not pass the kettle-on-stove stage for another year or so myself though, but I definately want to mix some AVR hacking into the hobby soon. I may make a lagering fridge with some Peltiers and an AVR first.
And all the fish in the nuclear plant's coolant outlet pond. :)
"anti-lock brakes"
For 5-6 years, a human driver could easily outbrake any ABS system. For many years, ABS systems sucked. And they still can barely compete with a human driver with an IQ higher than room temperature.
I don't mind power steering and automatic transmissions. In general they are robust mechanical systems, not prone to code bugs. The few auto transmissions I've seen that were electronically controlled (Chrysler 4-speed automatic) are reliablity nightmares. Note that Chrysler 3-speeds are purely mechanical and are considered nearly impossible to damage...
What about a parked car by the curb?
What about oncoming traffic coming around a bend? The radar will see it even though it's something you won't hit.
Without being EXTREMELY expensive and having a very large antenna, radar doesn't have the narrow beamwidth necessary to distinguish between stuff directly in front of you and things that are stationary but in an adjacent lane.
A laser-based system might work.
Don't forget approaching the bottom of a hill. The radar would think you're about to hit some solid substance, even though it's just the road leveling out.
If you read the article you will see that TrollTech is indeed owned by the Canopy Group.
Why am I not surprised?
Given the TrollTechSCO/Caldera connection, KDE could be in for a world of hurt...
"The Santa Cruz Operation, Inc. P.O. Box 7745
San Francisco, CA 94120-7745
United States of America"
The current SCO is NOT the same as the former SCO. (Now the Tarantella Group.)
If you read the article, you'll see that the current SCO was formerly Caldera. Caldera bought the Unix rights from SCO, the old SCO became Tarantella (which was one of their products IIRC...), and then Caldera renamed to The SCO Group.
That source offer was made by people with no management connection to McBride...
Like the "safe deposit box" part.
Yeah, CDs will fall apart or die faster in non-optimal environments. But a safe deposit box is not one of those.
I've dealt with them before.
I ordered a GeForce 4 Ti4600 (PNY Verto) from them, and the card was defective. Within 20-30 minutes of bootup, text in console mode would be corrupted every few characters, and in X (or in Windows), there would be vertical yellow lines every few pixels. Probably a cold solder joint somewhere...
So I first tried dealing with PNY support. Their response: Take it to the vendor for a return.
So I try to return it to Nova. They tell me to go talk to PNY.
It took two weeks to receive an RMA number, after which I sent back the card, demanding a refund. (I did NOT want a replacement, as at this point I wanted to end all further association with both PNY and Nova due to the horrendous customer support both of them had.)
A month later and still no credit. It took me 4-5 calls before I reached someone with a clue, who told me that they had tried the unit and it wasn't defective, and that a replacement was out of stock. I told him, no, I didn't want a replacement, I never wanted to deal with PNY again. In the end I got ripped off for a $30 restocking fee for a defective product and a month of hassles.
Stay away from Nova Computech. Don't know about TigerDirect, although they've had bad press here before. Walmart's return policies in-store are great, I don't know about online. I'm sad to say this, but Wally World is likely your best bet for one of these systems. (I'll be ordering one shortly - It's perfectly suited as a basic personal webserver.)
Of how Boeing does things right, look up the Gimli Glider.
The only reason that plane was able to land was because Boeing engineers were so damn conservative and had backups for every possible contingency, including dual-engine failure (Which the airline industry and Boeing considered a near impossibility, but included a backup for anyway.)
I don't care if the A380 might have a better "concept". Airbus engineers take shortcuts in their design of metal-composite joints that are just plain Bad Engineering, and I'm going to avoid flying on an Airbus jet if at all possible after what I've heard regarding their tail designs from someone familiar with the industry.
I refer you to the Brooklyn crash a few months after 9/11 where the tail of an Airbus plane basically fell apart...
Ever considered why power companies transmit their power at thousands of volts instead of 120 system-wide? Even though it means transformers wasting some power at the ends?
Simple. I^2*R losses. The power loss through a wire is the square of the current times the resistance. If you triple the voltage (Note: the system proposed is a 36/42 system, triple that of our current 12/14 system. First being nominal battery voltage, second being charging voltage.), you have 1/3 the current to deliver the same power. This means that resistive losses in the automotive wiring is 1/9 what it was previously.
Also, with a 36V system, it may become practical to drive the valves with solenoids rather than camshafts. All of a sudden the engine's "sweet spot" of maximum efficiency becomes MUCH wider. (If implemented properly, the engine could run at max efficiency at nearly any load/RPM combo with valve timing that is fully electronically controlled.)
One thing I've seen (and happen to have, and like), is a "slow" jumpstart device. Instead of being able to leech enough power to start the engine in one instant, it is designed to trickle charge the battery for 10-20 minutes, just enough to start the engine. It works (believe it or not) through the cigarette lighter socket, and works quite well. (You just have to wait some time before trying to start the car.)
It would be trivial for someone to start making these with a step-up DC/DC converter.
The nicest thing is that partially charging a dead battery needs little to no regulation at all...
While I don't have kids yet, I was recently on vacation with the extended family, including little kids.
For various reasons, for most of vacation I was not able to use my laptop in the car (No room, also didn't want to have something that valuable being left in a parking lot in the middle of nowhere while we were hiking.)
On the way back we had a 4-5 hour drive from Grand Junction, CO to Denver. (Couldn't get a flight into GJ.)
Pulled out my laptop and some DVDs... Quietest and calmest ride of the entire trip. (Needless to say, during the week some of the rides were pure hell thanks to my cousin's hyperactive kids.)
Without modern CPUs, home video editing would not be practical (and hence the market for DV camcorders would be much smaller.)
You obviously haven't tried compressing 2 hours of video into DVD-quality MPEG-2, let alone trying to compress it into DivX to send home videos to some relatives.
Would we really need more than 800 MHz on a home computer? I have a 1.7 GHz P4 laptop, and a 1.1 GHz Athlon. Upgrading to a Barton 3000+ (2 GHz or so actual clockrate, but much more efficient per clock than my current TBird) would take my 14-hour encoding jobs down to 7 hours. A difference between taking most of the day and running while I sleep.
And reencoding 1080i HDTV recordings into a more managable size... yikes... I've had 24 hour encoding jobs before.
So my suggestion: Go buy a DV camcorder, or an HDTV tuner card. I guarantee you you'll be desperate to upgrade that poke-ass 800 MHz machine in under two weeks.
I would think that they'd almost HAVE to use these in conjunction to get any benefits.
While they may be able to get 30-35% improvement for PMOS alone, and 20-25% for NMOS (or was it the other way around?), if implemented in a chip, improved PMOS transistors without improved NMOS would result in almost no maximum speed improvements. (It would likely improve power consumption, but not as much as the speed benefit of the transistor itself.)
This is because any given gate involves both NMOS and PMOS transistors. The most basic gate type is an inverter, which consists of one NMOS and one PMOS. Improve only one and you'll improve either the turn-on or turn-off time of the gate quite a bit, but not both (You'll get some improvements to both, but it will be unbalanced). So the unimproved time will be your speed barrier. Use both technologies and you're improving the whole gate, not just half of it.
What I would love to see is if someone could come up with a gate architecture that could provide great improvements in fan-out capability with minimal penalties in gate delay. (Fanout is why memory speed hasn't kept up with CPU speed - The larger the memory gets the more loaded the address drive lines get, and high fanout = slow speed.)
I did some scientific computing work a few summers ago.
:)
At that time, my new iPaq (Not the PDA, the business desktop system type that seems to be relatively unknown) was competitive with some much more expensive (but 2-3 years old) high-end computing hardware. My boss was impressed at how well the $1000 system he bought for his summer intern performed. And that system was only 500 MHz.
1 month into my internship, I started running simulations on that machine. Some only ran for 10 minutes, but each batch would include 20-30 runs that got progressively longer. The final runs in a batch would be over a day each - To collect the one dataset I needed to work with took over a month of total CPU time.
With a modern $600ish system I could finish those computations in *under a week*
I wasn't storage-bound. My end data from each batch was only a 200x200 matrix.
Yup. 1 day or more to generate a 200x200 matrix.
And do some video editing (esp. compressing it to MPEG-2 or DivX)
You'll change your tune.
With some of the more advanced video compression algorithms (DivX for example - Yes it has legit uses, great for distributing home videos to relatives.), a 10% increase in CPU speed can mean an hour or two off of your compression time.
With exceptions, often 12 year olds are the least likely to succeed (or even play) an MMOG.
Because of the social interactions and complexity of most MMOGs, those who have good communications skills and are intelligent are most likely to succeed.
Some MMOGs have higher ages than others. DAoC, which has a large amount of hack-n-slash, seems to have a greater following in younger circles than EVE Online, which seems to have a much higher average age, partly due to the fact that the game is insanely complex and to really succeed requires a basic grasp of economic principles. (The truly rich people are those who can find good trade routes... Which is very difficult and requires a lot of thinking and calculations.)
CDMA2000 1xRTT (Sprint Vision, Verizon Express Network) is not 3G. It's considered 2.5G.
3G standards include UMTS (only 1-2 test networks in the US, and so far massive failures due to handset problems, especially in the battery life/heat arena, in Europe and Japan.), and CDMA2000 1xEV-DO (Which is available in Korea and also was rolled out by KDDI in Japan I believe.)
802.11 has some support for handoffs between APs, but not on a large scale, and not if it's occuring often.
It's also not designed to handle rapidly moving stations. Once you start going faster than walking speed, multipath fading and doppler shifts make things fun. Dealing with users traveling at 55 MPH is one of the biggest challenges of cellular network designers, and it just gets to be more and more fun as the bandwidth increases.
This isn't targeted at stationary users. It's targeted at people on the go. (Not drivers, mind you, but carpool passengers and people on buses and trains, etc.)
It might also prove useful for police work - Police departments have a use for high-speed mobile data. (I believe that the biggest customer of the Ricochet remnants in Denver is one of the local police departments. I know that there's a metropolitan wireless network somewhere that is used heavily by the local police.) It could (in theory) provide an off-the-shelf communications solution for a low-cost civilian Predator equivalent, which law enforcement entities would kill for. All of the advantages of a police helicopter without the exorbitant expenses...
Depends on the exact rules on that spectrum licensing.
Right now the providers are all begging the European government to ditch UMTS, which has proved to be a nightmare. If the gov. lets them, they can retask the spectrum to 4G technologies such as Flarion's OFDM-based network architecture.
FYI, Flarion has some excellent whitepapers on their site describing their tech, and the idea of OFDM in general.
:)
If all goes well I'll be a Flarion employee in a month or so. (Getting laid off, applying for a position at Flarion which is 20 minutes away from here, and coming into the application with great references, as my current company and Flarion are both spinoffs from Lucent's wireless division in Whippany.) So I've done quite a bit of research into the company and their tech.
FYI, European digital TV broadcasts use OFDM modulation. The iBiquity IBOC radio broadcasting standard (Yet Another Lucent Spinoff) uses OFDM. IBOC was recently approved by the FCC as the standard for digital audio broadcasting in the US, although unfortunately for iBiquity, the economic downturn has caused nearly all broadcasters to cancel upgrade plans for the time being.
Was referring to the original article, not your post. Sorry for the misunderstanding.
Not UltraSparc.
Read: OLD. On the order of a 486 or maybe a classic Pentium in power.
a) As the article mentioned, some readers use multiple read heads to achieve higher read speeds with lower spin rates. So higher than 48x is indeed possible.
b) The latest IDE standard can push 133 MB/sec. Even taking into account overhead, an ATA100 7200 RPM hard drive can easily do 40 MB/sec in the real world. (That's what my drive does.)
Since 1x CD is 150 KB/sec, 48x is only 5-6 MB/sec, a small fraction of what even ATA33 is capable of, let alone ATA133.