also, is there any way to block certain IPs (manually that is) from sending yuo mail?
Yes, if you are running the mailserver. With more recent versions of sendmail, there is a "access" file, which should be your first line defense in the prevention of relay-rape. With this file you also block by individual IP address, IP range (class A, B, or C) or by domain name. On my mail server I filter out about 98% of spam directed to me with this method. This can cause problems, because it is pretty easy to block legitamate mail this way- unless the person trying to contact you has an alternate channel to get to you, you'll never even know they are trying to contact you.
I've worked off and on in UWB research since 1991- in fact I used to work with Doug Cummings, one of the people mentioned in the Article (at the University of Texas Applied Research Laboratories). I've been hearing UWB hype for ten years now- but really, there is no magic to it, and it has some very real limitations.
UWB as a communication method depends on the time position of signals, which can be severely affected by the motion of the transmitter or reciever, especially if it is accelerating or decellerating. Traditional modulation techniques can and will be affected by UWB, though in many cases, it may just raise the noise floor.
The biggest problem with UWB right now is that it is a political football. The established users of the spectrum want to protect their "territory" from all threats, real and imagined. From what I've seen the reasoning is very much "It's different, so it must be BAD." In truth, UWB is another form of modulation. Just like FM has advantages (and disadvantages) in comparison to AM, so will we view UWB in the years to come.
Some day it will be used- it won't revolutionize communications, but it may give us a little more efficient use of spectrum- Like Turbo codes and the like, they give us an incremental increase in what we can do. Claude Shannon's "Limit" still holds and puts bounds on what we can do.
I would consider it a componant of the radiation the device emits...
But the device isn't emitting it- the person is the source of the static discharge. The static charge is absoulutely *not* generated by the operation of the device- If it was, then I agree that it would have to be taken into account.
For static discharge, the FCC class rating of the device is a non-issue. It is a concern safety-wise, but the FCC rating is about emissions, not about safety. There are other, separate, certifications about safety.
I've taken part in EMI testing for class B certification- static discharge is not a part of it.
Class B has nothing to do with static electricity- it is just about intentional and unintentional radiation derived from the normal operation of the electronics.
I'm not even sure a (single) static discharge would show up on an EMI scan (which is done for class B certification) It would probably show up as a slight increase in the baseline noise level, but not push it beyond the class B limits.
Bullshit. I worked as a security guard at Ciba-Geigy(sp?) when I was in school. They had been doing research on pre-emergent insecticide/herbicides for years(>20), because of the promise that it would use a fraction of the chemicals and be more potent. Obviously not immediately promising, and yet the company persist. I hear this argument on/. constantly. Companies only work on stuff that will show on next quarters stock report.
I'll turn it around at you and say that most academics spend time chasing rainbows that have no application (obvious or otherwise), and that only occasionally does someone develope something that is useful. (Yes, this statement is also full of shit, but no more so than yours.)
All I said (essentially) is that a company has an obligation to its stockholders, and that obligation is to give them a return on their investment. Government is (supposed to be) a non-profit institution.
I'm a big supporter of government funded research because it has more leeway to take greater risks. I worked as a government researcher for 9 years. One of the products I worked on had potential for an actual product in maybe 20 years, and at the end of those 20 years, it still may not be a very commerically viable product, but still very useful from a humanitarian standpoint. We wanted to get support from companies, but they wouldn't touch it unless they could see a profit in 5 years. And those companies that did join us were constantly threatening to bail at every setback.
A few big companies can support large scale, long term research, but today's market doesn't encourage that- investors are pulling their money out of those types of companies (seen how good Lucent stock has done lately?) and putting them into companies with the potential of large short term growth.
It sure seems to me that GWB didn't really make a decision- what people wanted was a definitive yes or no, but he ends up giving a "you can't blame me," try to please the most people, answer. I don't see it as a decision, it is more of a postponement of the answer, and a deferral to his new "council" on the matter.
I thought people were fed up with all the politics going on- and it sure seems like more of the same- refusal to take a real stand, because, horrors of horrors, he may lose some of his constintuency.
It doesn't look like this decision will make embryonic stem cell research any easier- now they will need documentation on the particular line of the cell and so forth. Embryonic stem cell research has been really hard already for any entity that is not privately funded- the government has required an extreme amount of separation in the facilities of Universities doing this sort of research (to the point of demanding completely separate, off-campus facilities, with absolutely no sharing of equipment/staff/support, and so on... there was a NPR report on this a few weeks back).
I don't think that this decision will really advance the cause of stem cell research, more just push it further into the arms of commerical entities. Commercial work in this area is great- but companies need to push for profits and drop research in areas that are not immediately promising. The value of government investment is that it enables researchers to work on deeper, long term projects that may not have an obvious path to profit. Research does not always mean success.
I used to live in the DC area and there were plenty of really good hamfests, which always had tons of computer stuff. I live in Austin, TX now, and the local hamfest-type scene is *pathetic*.
Look for the Hamfests in Bowie, Timonium and Manassas. The Gaithersburg hamfest (F.A.R.Fest) was the biggest I remember, but since I've left the area, I think it moved to Bowie. I remember plenty of smaller ones, but the smaller ones tended to be more oriented to amateur radio, and less cool technology junk.
I run my own email server, and I admit, every once in a while, I get pretty obsessive about looking at the mail logs. For a few weeks earlier this year, I had someone from a [big national ISP] dialup pounding my server with requests that came up with 'unknown user' bounces. The usernames were common first names, and names like "marketing", just trying to get a hit. My best guess is someone was using a dictionary type attack to find valid usernames to spam. I sent email to [big national ISP] giving them the logs and the specific IP address that these were originating from. No response, attack continued. I finally denied that IP range with the sendmail 'access' file.
How can you fight this type of harvesting? I can't figure out how... having some sort of feedback when an legitimate email has a mistyped username is useful, so I don't want to accept and route to/dev/null all the 'unknown user' emails.
Microchip has a new part, the MCP2150, which is designed to fit between a microcontroller's serial port and the infra-red transducers. Here are a bunch of application notes that apply. Earlier this week my local Microchip representative got me a couple samples of the MCP2150. I'm looking to make an interface between my Palm Pilot and a datalogging microcontroller. When will I get to it? Who knows- it might just add to my stack of sample parts...
One thing I really like about Microchip is that they are pretty committed to distributing free samples and free/low cost tools for development. If you've ever priced the development hardware from Intel or Motorola, you can see why this is such a big attraction to the small time/home developer.
According to the encyclopedia Britanica: one bel=.1 decibles. In other words, for those of you who dont remember high school physics, a 2 bel hard drive is very quiet.
You've got it the wrong way around- 1 bel is 10 decibels. Remember that "deci" is the metric suffix for 10^-1, "deka" is the metric suffix for 10. These drives are 25 dB when seeking.
Monitor all the traffic with a proxy server- including your own surfing, and go over the logs together. The best way you can teach is to set a good example.
Many kids will accidentally go to someplace like whitehouse.com and get the suprise of their life- as long as they understand that is something they should not be looking at- that should be fine, lesson learned, and they know not to go there again. Eventually the kid will see something you find morally troubling- and so will you. Take the time to explain what you think is wrong about things like that. The child needs to know why she shouldn't be browsing something, more than just "bad place- stay away!"
Is thermal output in direct relation with power usage on processors?
Absolutely! The power has to go somewhere. The power can be stored, but usually that's too inefficient to be worthwhile, so we just radiate it into the environment, mostly as heat. The less heat a computer makes, the less power it is using. Yes, some of that power is going into making other forms of lower frequency EM radiation, but that's insignificant. Eventually all the energy we expend ends up as heat. The laws of thermodynamics are kind of depressing. Stolen from the web somewhere:
The simplified Laws of Thermodynamics: 1. You can't win; 2. You always lose; and 3. You can't quit.
Both Dell and Compaq make a 1U foldaway rack drawer that has a keyboard and a flat LCD screen: Dell, and Compaq. They're not exactly cheap, but with a KVM you would only need one of them for a whole room full of servers.
I'm pretty sure that other manufacturers have similar products.
The easy way to tell what is drawing the most power is to see what is getting hot. The power that you're spending has to go someplace, and >99% of it is going into heat, a tiny fraction into excess RF energy. Look at the parts with the most fans and the biggest heat sinks- those are drawing the most power. Feel the wall-warts/power supplies if they are external. If it's getting warm, there's some ineffiency there. Many times you can't avoid inefficiency (no matter what, entropy always wins) but you can do things like using a powerstrip to switch off unused things on the AC side.
Really- if you want to save power through your computer- turn it off! Yes, places like SETI@Home can take advantage of your "unused cycles" but don't ever forget that somebody still has to pay for it.
I haven't seen any numbers recently, but I wonder what a chart of watts/instruction would look like as we progress through the processors of history. Are we getting more efficient, less, or just holding even?
An L shaped adapter won't work for PCI- Each slot has a few unique lines- you can't share them, unless your right-angle adapter has a PCI-PCI bridge on it.
There are 1U servers from Dell (for sure) and Compaq (I think) that support 2 cards. I'm pretty sure other manufacturers have them as well, but all the mass produced 1U boxes I've seen use custom motherboards.
If it is truly a dusty environment, and you want to keep the maintainence low, seal the box. You can get Peltier junction A/C units- not very efficient or quiet (because of the fans), but they do have the advantage of allowing a sealed box. This one will remove 550 BTU/hr. You don't need to keep it cold, just remove enough heat so that the computer doesn't fry itself. There will be some leakage of air into it- no big deal, but because of it, you will need some way to get rid of the water that condenses on the cold side of the junction- your periodic maintainence will have to include some sort of dessicant, or at least making sure the drain holes are clear.
If you can't afford a lawyer to write your NDA, how will you be able to afford a lawyer to defend your NDA?
Every business has some risk associated with it. If you're so risk-averse that you never disclose it to anyone, it is very likely that nothing will ever come of your idea. Choose carefully, but ultimately you are going to have to trust who you choose, NDA or not.
About 6 is the absolute maximum number of slots you can get on a PCI bus segment. There are a limited number of "loads" that a PCI bus can support, usually about 10-12 max. The connector counts as a load and the card itself is a load. To go greater than that, you have to have a PCI-PCI bridge. Unfortunately, when you have to go across a bridge, your performance goes down. Because of the way PCI works, transfers on the bus must be interruptible. If the transfer is going across a bridge, you will gain latency when the bridge has to re-negoitate the transfer.
PCI-X has improved this, but right now, PCI-X devices (and motherboards) are pretty rare. You can't mix PCI and PCI-X on the same bus segment without the bus segment downgrading to PCI functionality.
Also, according to the PCI spec, a card can draw up to 25W, which increases the size of your power supply.
The best bet is to go with a motherboard with some integrated functionality- as others have said, compatability is an issue, but there are motherboard makers out there who use high quality parts (which are well supported in Free-OSes), Adaptec for SCSI, Intel for ethernet, and so forth, but you're going to have to pay more for it.
But even in Hawaii, which traditionally has some of the highest electricy rates in the US, this report http://www.hawaii.gov/dbedt/ert/pv_hi.html states advantages and disadvantages- notably:
1. Photovoltaic-produced electricity is presently more expensive than power supplied by utilities.
...
4. Batteries need periodic maintenance and replacement.
5. Some of the materials used in the manufacturing of PV panels are toxic.
Most of the references I found when googling the topic were positive when they were trying to sell you PVs, and neutral otherwise- look at the sources of the articles- Siemens has a great interest in you buying their PVs. Really, it gets into the classic case of shifting the pollution to a lower-rent district- not really providing a global solution.
With PVs, not every place is the same- KWh/year of sunlight at installation location, PV conversion efficiency, energy costs, area available all have an effect on the final cost. No payback time estimate is valid unless this data is included. I'll maintain my original proposition- PVs don't make sense if you're trying to save money on energy for the average home.
No. Unfortunately, photovoltaics take more energy to produce than they will produce in their lifetime. Melting down all that silicon takes a lot of Joules, and they're relatively inefficient- a very, very good PV is about 20%, typically near 10%. I remember an estimate of about 1200W/sq-m for sunlight, so you could get maybe 120W/sq-m- In a sunny area, I think you get the equivalen of about 6 hours of sunlight/day- which is about.7 KW/hr per day, per sq. meter of PV. PV's don't last forever, either, they break when things drop on them, they corrode, they get dirty.
From an economic standpoint, if the power grid is available, PVs will never beat the grid. If your home is extremely isolated, then PVs start making more sense, because it can be very expensive to run power out to the middle of nowhere.
Ultrawideband is not exactly new, I worked on a UWB radar almost 10 years ago. It has promise, but it also has problems that you don't see with traditional carrier based methods- since you are talking about picosecond resolution, communication with a reciever that is moving with respect to your transmitter is hard- in a picosecond, light travels about.3 mm. Virtually any motion can cause your reciever to start looking at the wrong part of the data- since the data is all time-coded.
Many have said that you can't detect UWB, but usually, they are looking at it with a traditional carrier based spectrum analyzer- you can't see it, but it is there. It just looks like additional noise. which looks very similar to spread spectum- which is exactly what UWB is, but you're spreading the spectrum very, very wide- an impulse in the time domain is uniform (all frequencies, equal amplitude) in the frequency domain.
Time domain has been around, and pretty big in the UWB arena for a long time, but most of what they have is vapor- big promises, but they fail to deliver. There are other players in the field, Aetherwire is one of the other biggies of UWB. There has been a huge amount of money poured into UWB, but dreadfully little usable technology.
Linus brings up an interesting problem with society, although he doesn't go into length about it. That problem is the reliance of capitalism in our society for intellectual advancement.
Unfortunately, scientists still have to pay the rent, and buy food. Science has never been free. Even when people are doing it just for the love of it, someone still has to support the scientist. You must remember that the scientists of the classical age had patrons, usually aristocrats, that paid for their research, supported in much the same way that artists were. Capitalism has changed the patron from an individual to a corporation, and unfortunately, corporations have stockholders that insist on a return on their money.
We've gotten pretty far away from the patron/scientist model of scientific research, in part because there is little prestige in supporting a scientist, and also because science has gotten so complex and expensive. To do much serious scientific research, you have to have many specialists with complex equipment that is often very, very expensive (the machinists that make the stuff need to feed their families too!). Open source is similar to, but not equal to scientific inquiry. It has been proven many times over the past decade that a single person with an idea can make a big difference. You can start working on open source with a minimum amount of expense- what it takes is a desire to create. Now you can do your coding on your own, without a big corporation to back you. In 300 years, this may not be the case, but by then, we should have other technologies to hack.
Competition is supposed to push better products to the fore. If that worked, we'd have light bulbs that laster for 30 years, cars you bought for a lifetime, and software for life.
We would have all those things right now, if it were not for the people that are unwilling to spend $100 on a light-bulb, $200,000 on a car, and $500,000 for a word processor. One way or another, things have to be paid for, otherwise there will be no incentive to produce more. Yeah, capitalism sucks in a lot of ways, but so do all the other -isms. Capitalism has been very successful in inspiring people to create, because by creating they can make a better life for themselves and their progeny.
Until we can divorce the pursuit of capital from advances in science, we are doomed to have any advance kepted restained by the barriers of the a accumulation of that wealth. If at any point, an advancement is deemed to be a money killer, it will be abandoned.
This is not the fault of capitalism, it is the failing of an oligarchy- where the rich have the political power. At that point it is in the best interest of the rich to keep around the mechanisms that keep them rich. The recent attacks on technologies haven't been counter-attacked on the technology front, they have been attacked on the legislative front. Don't try to tear down capitalism, try to build a democracy that isn't financed by the rich, for the rich.
As long as you are keeping the speed down (under 20 MHz or so) wire wrap works well, but as you go higher and higher in frequency, the added inductance and stubs of wire wrapping will kill you. Many of the newer parts are getting harder to find in DIP- especially as the pin count goes up. I don't know of any (modern) common part in a DIP package > 40 pins. On the other hand, 300-400 ball BGAs are getting common- imagine trying to wire wrap that! There is no real market in making a new quad XOR gate in DIP- that technology is mature- they won't really be able to improve on it, except possibly to make it faster, but a DIP package has leads, and leads have inductance, and inductance kills high frequencies, so by default, if you are improving the speed on an IC, you need to shrink it and go to SMT. They are also trying to cram more into a chip, but that implies a greater pin count- again, which pushes you to SMT.
But, as the person who posed the question implied, SMT is difficut for a hobbiest to work with. One alternative is to make your own PCBs, either by etching them at home, or go to some place like APCircuits where you can get prototype PCBs made pretty cheaply. And one hint- if you are doing SMT, get the solder mask, it will make your life very much easier.
I still wire wrap when I need to- and Digi-Key has had most of what I've needed, though the little pins to mount discrete components are outrageously expensive. (In particular, the T68A bifurcated pins, with room for 3 levels of wrap, at $92.25/1000.)
First, in a nutshell, look for PCI-X. It makes PCI a switch based medium instead of a bus, will let you utilize your existing PCI devices..
PCI-X does have some very good advantages, but unfortunately if you plug a PCI device into a PCI-X bus, you've just turned that bus (segment) into a PCI bus, no PCI-X advantages. Plus there are serious limitations on the number of loads (cards) on a PCI-X bus- which makes for more bridges on the motherboard, and a performance hit when you have to cross it.
It _should_ be possible the create a PCI-Specification with let's say 133Mhz, and 64 Bits...
It already exists- its called PCI-X, coming to a computer near you, but it has its own problems, like a limited number of targets per bus (very limited, 1 or 2 I think). One problem with PCI is that it was designed as a low power bus- which is a good idea when you have 64 bits of data coming down the line. Unfortunately that makes it difficult to deal with when you're getting to higher speeds.
PCI has deeper limitations- when a transfer is initated, the target has no idea who is the master of the transaction, which makes it harder to restart a transaction that has been pre-empted. PCI-X attempts to solve this, and other issues that have come up, but maintaining backwards compatibility will only result in band-aid fixes, the fundamental flaws are still there. Sometimes you need to cut your losses and run- like the ISA/EISA to PCI jump- no compatibility there, but higher speed processors make different demands on the bus. As we've built faster and faster systems, we've found new limitations.
This would make transition easier, as old PCI cards would run in the new slots (at least this works for 64Bits Pci-Slots, which can also run 32Bit Card). If this doesn't work for Slots using a higher frequency, the chipsets could include 2 Pci-Controllers, each driving 3 or 4 slots, and each controller could fall back to 33Mhz if there is one card which doesn't support higher speeds.
This technology exists right now- if you have a PCI "bridge" in your system that supports 66MHz/64 bit, that is how it is supposed to work. Unfortunately if your data has to cross a bridge, you get a performance hit.
We can continue patching PCI, or we can learn from our experience and design a new bus.
Slashdot Mirror
Yes, if you are running the mailserver. With more recent versions of sendmail, there is a "access" file, which should be your first line defense in the prevention of relay-rape. With this file you also block by individual IP address, IP range (class A, B, or C) or by domain name. On my mail server I filter out about 98% of spam directed to me with this method. This can cause problems, because it is pretty easy to block legitamate mail this way- unless the person trying to contact you has an alternate channel to get to you, you'll never even know they are trying to contact you.
I've worked off and on in UWB research since 1991- in fact I used to work with Doug Cummings, one of the people mentioned in the Article (at the University of Texas Applied Research Laboratories). I've been hearing UWB hype for ten years now- but really, there is no magic to it, and it has some very real limitations.
UWB as a communication method depends on the time position of signals, which can be severely affected by the motion of the transmitter or reciever, especially if it is accelerating or decellerating. Traditional modulation techniques can and will be affected by UWB, though in many cases, it may just raise the noise floor.
The biggest problem with UWB right now is that it is a political football. The established users of the spectrum want to protect their "territory" from all threats, real and imagined. From what I've seen the reasoning is very much "It's different, so it must be BAD." In truth, UWB is another form of modulation. Just like FM has advantages (and disadvantages) in comparison to AM, so will we view UWB in the years to come.
Some day it will be used- it won't revolutionize communications, but it may give us a little more efficient use of spectrum- Like Turbo codes and the like, they give us an incremental increase in what we can do. Claude Shannon's "Limit" still holds and puts bounds on what we can do.
I would consider it a componant of the radiation the device emits...
But the device isn't emitting it- the person is the source of the static discharge. The static charge is absoulutely *not* generated by the operation of the device- If it was, then I agree that it would have to be taken into account.
For static discharge, the FCC class rating of the device is a non-issue. It is a concern safety-wise, but the FCC rating is about emissions, not about safety. There are other, separate, certifications about safety.
I've taken part in EMI testing for class B certification- static discharge is not a part of it.
Class B has nothing to do with static electricity- it is just about intentional and unintentional radiation derived from the normal operation of the electronics.
I'm not even sure a (single) static discharge would show up on an EMI scan (which is done for class B certification) It would probably show up as a slight increase in the baseline noise level, but not push it beyond the class B limits.
Bullshit. I worked as a security guard at Ciba-Geigy(sp?) when I was in school. They had been doing research on pre-emergent insecticide/herbicides for years(>20), because of the promise that it would use a fraction of the chemicals and be more potent. Obviously not immediately promising, and yet the company persist. I hear this argument on /. constantly. Companies only work on stuff that will show on next quarters stock report.
I'll turn it around at you and say that most academics spend time chasing rainbows that have no application (obvious or otherwise), and that only occasionally does someone develope something that is useful. (Yes, this statement is also full of shit, but no more so than yours.)
All I said (essentially) is that a company has an obligation to its stockholders, and that obligation is to give them a return on their investment. Government is (supposed to be) a non-profit institution.
I'm a big supporter of government funded research because it has more leeway to take greater risks. I worked as a government researcher for 9 years. One of the products I worked on had potential for an actual product in maybe 20 years, and at the end of those 20 years, it still may not be a very commerically viable product, but still very useful from a humanitarian standpoint. We wanted to get support from companies, but they wouldn't touch it unless they could see a profit in 5 years. And those companies that did join us were constantly threatening to bail at every setback.
A few big companies can support large scale, long term research, but today's market doesn't encourage that- investors are pulling their money out of those types of companies (seen how good Lucent stock has done lately?) and putting them into companies with the potential of large short term growth.
It sure seems to me that GWB didn't really make a decision- what people wanted was a definitive yes or no, but he ends up giving a "you can't blame me," try to please the most people, answer. I don't see it as a decision, it is more of a postponement of the answer, and a deferral to his new "council" on the matter.
I thought people were fed up with all the politics going on- and it sure seems like more of the same- refusal to take a real stand, because, horrors of horrors, he may lose some of his constintuency.
It doesn't look like this decision will make embryonic stem cell research any easier- now they will need documentation on the particular line of the cell and so forth. Embryonic stem cell research has been really hard already for any entity that is not privately funded- the government has required an extreme amount of separation in the facilities of Universities doing this sort of research (to the point of demanding completely separate, off-campus facilities, with absolutely no sharing of equipment/staff/support, and so on... there was a NPR report on this a few weeks back).
I don't think that this decision will really advance the cause of stem cell research, more just push it further into the arms of commerical entities. Commercial work in this area is great- but companies need to push for profits and drop research in areas that are not immediately promising. The value of government investment is that it enables researchers to work on deeper, long term projects that may not have an obvious path to profit. Research does not always mean success.
I used to live in the DC area and there were plenty of really good hamfests, which always had tons of computer stuff. I live in Austin, TX now, and the local hamfest-type scene is *pathetic*.
Look for the Hamfests in Bowie, Timonium and Manassas. The Gaithersburg hamfest (F.A.R.Fest) was the biggest I remember, but since I've left the area, I think it moved to Bowie. I remember plenty of smaller ones, but the smaller ones tended to be more oriented to amateur radio, and less cool technology junk.
I run my own email server, and I admit, every once in a while, I get pretty obsessive about looking at the mail logs. For a few weeks earlier this year, I had someone from a [big national ISP] dialup pounding my server with requests that came up with 'unknown user' bounces. The usernames were common first names, and names like "marketing", just trying to get a hit. My best guess is someone was using a dictionary type attack to find valid usernames to spam. I sent email to [big national ISP] giving them the logs and the specific IP address that these were originating from. No response, attack continued. I finally denied that IP range with the sendmail 'access' file.
/dev/null all the 'unknown user' emails.
How can you fight this type of harvesting? I can't figure out how... having some sort of feedback when an legitimate email has a mistyped username is useful, so I don't want to accept and route to
Microchip has a new part, the MCP2150, which is designed to fit between a microcontroller's serial port and the infra-red transducers. Here are a bunch of application notes that apply. Earlier this week my local Microchip representative got me a couple samples of the MCP2150. I'm looking to make an interface between my Palm Pilot and a datalogging microcontroller. When will I get to it? Who knows- it might just add to my stack of sample parts...
One thing I really like about Microchip is that they are pretty committed to distributing free samples and free/low cost tools for development. If you've ever priced the development hardware from Intel or Motorola, you can see why this is such a big attraction to the small time/home developer.
According to the encyclopedia Britanica: one bel= .1 decibles. In other words, for those of you who dont remember high school physics, a 2 bel hard drive is very quiet.
You've got it the wrong way around- 1 bel is 10 decibels. Remember that "deci" is the metric suffix for 10^-1, "deka" is the metric suffix for 10. These drives are 25 dB when seeking.
Monitor all the traffic with a proxy server- including your own surfing, and go over the logs together. The best way you can teach is to set a good example.
Many kids will accidentally go to someplace like whitehouse.com and get the suprise of their life- as long as they understand that is something they should not be looking at- that should be fine, lesson learned, and they know not to go there again. Eventually the kid will see something you find morally troubling- and so will you. Take the time to explain what you think is wrong about things like that. The child needs to know why she shouldn't be browsing something, more than just "bad place- stay away!"
Is thermal output in direct relation with power usage on processors?
Absolutely! The power has to go somewhere. The power can be stored, but usually that's too inefficient to be worthwhile, so we just radiate it into the environment, mostly as heat. The less heat a computer makes, the less power it is using. Yes, some of that power is going into making other forms of lower frequency EM radiation, but that's insignificant. Eventually all the energy we expend ends up as heat. The laws of thermodynamics are kind of depressing. Stolen from the web somewhere:
The simplified Laws of Thermodynamics: 1. You can't win; 2. You always lose; and 3. You can't quit.
Dell, and Compaq. They're not exactly cheap, but with a KVM you would only need one of them for a whole room full of servers.
I'm pretty sure that other manufacturers have similar products.
The easy way to tell what is drawing the most power is to see what is getting hot. The power that you're spending has to go someplace, and >99% of it is going into heat, a tiny fraction into excess RF energy. Look at the parts with the most fans and the biggest heat sinks- those are drawing the most power. Feel the wall-warts/power supplies if they are external. If it's getting warm, there's some ineffiency there. Many times you can't avoid inefficiency (no matter what, entropy always wins) but you can do things like using a powerstrip to switch off unused things on the AC side.
Really- if you want to save power through your computer- turn it off! Yes, places like SETI@Home can take advantage of your "unused cycles" but don't ever forget that somebody still has to pay for it.
I haven't seen any numbers recently, but I wonder what a chart of watts/instruction would look like as we progress through the processors of history. Are we getting more efficient, less, or just holding even?
An L shaped adapter won't work for PCI- Each slot has a few unique lines- you can't share them, unless your right-angle adapter has a PCI-PCI bridge on it.
There are 1U servers from Dell (for sure) and Compaq (I think) that support 2 cards. I'm pretty sure other manufacturers have them as well, but all the mass produced 1U boxes I've seen use custom motherboards.
If it is truly a dusty environment, and you want to keep the maintainence low, seal the box. You can get Peltier junction A/C units- not very efficient or quiet (because of the fans), but they do have the advantage of allowing a sealed box. This one will remove 550 BTU/hr. You don't need to keep it cold, just remove enough heat so that the computer doesn't fry itself. There will be some leakage of air into it- no big deal, but because of it, you will need some way to get rid of the water that condenses on the cold side of the junction- your periodic maintainence will have to include some sort of dessicant, or at least making sure the drain holes are clear.
If you can't afford a lawyer to write your NDA, how will you be able to afford a lawyer to defend your NDA?
Every business has some risk associated with it. If you're so risk-averse that you never disclose it to anyone, it is very likely that nothing will ever come of your idea. Choose carefully, but ultimately you are going to have to trust who you choose, NDA or not.
About 6 is the absolute maximum number of slots you can get on a PCI bus segment. There are a limited number of "loads" that a PCI bus can support, usually about 10-12 max. The connector counts as a load and the card itself is a load. To go greater than that, you have to have a PCI-PCI bridge. Unfortunately, when you have to go across a bridge, your performance goes down. Because of the way PCI works, transfers on the bus must be interruptible. If the transfer is going across a bridge, you will gain latency when the bridge has to re-negoitate the transfer.
PCI-X has improved this, but right now, PCI-X devices (and motherboards) are pretty rare. You can't mix PCI and PCI-X on the same bus segment without the bus segment downgrading to PCI functionality.
Also, according to the PCI spec, a card can draw up to 25W, which increases the size of your power supply.
The best bet is to go with a motherboard with some integrated functionality- as others have said, compatability is an issue, but there are motherboard makers out there who use high quality parts (which are well supported in Free-OSes), Adaptec for SCSI, Intel for ethernet, and so forth, but you're going to have to pay more for it.
The info I was originally citing was a bit dated, but the payback time varies greatly- this study: http://www.siemenssolar.com/Paybackstudy.pdf has a wide range, but this look: http://www.archrecord.com/CONTEDUC/ARTICLES/01_01_ 1.asp has estimates ranging up to 40 years for payback on an installation in New York.
But even in Hawaii, which traditionally has some of the highest electricy rates in the US, this report http://www.hawaii.gov/dbedt/ert/pv_hi.html states advantages and disadvantages- notably:
1. Photovoltaic-produced electricity is presently more expensive than power supplied by utilities.
...
4. Batteries need periodic maintenance and replacement.
5. Some of the materials used in the manufacturing of PV panels are toxic.
Most of the references I found when googling the topic were positive when they were trying to sell you PVs, and neutral otherwise- look at the sources of the articles- Siemens has a great interest in you buying their PVs. Really, it gets into the classic case of shifting the pollution to a lower-rent district- not really providing a global solution.
With PVs, not every place is the same- KWh/year of sunlight at installation location, PV conversion efficiency, energy costs, area available all have an effect on the final cost. No payback time estimate is valid unless this data is included. I'll maintain my original proposition- PVs don't make sense if you're trying to save money on energy for the average home.
No. Unfortunately, photovoltaics take more energy to produce than they will produce in their lifetime. Melting down all that silicon takes a lot of Joules, and they're relatively inefficient- a very, very good PV is about 20%, typically near 10%. I remember an estimate of about 1200W/sq-m for sunlight, so you could get maybe 120W/sq-m- In a sunny area, I think you get the equivalen of about 6 hours of sunlight/day- which is about .7 KW/hr per day, per sq. meter of PV. PV's don't last forever, either, they break when things drop on them, they corrode, they get dirty.
From an economic standpoint, if the power grid is available, PVs will never beat the grid. If your home is extremely isolated, then PVs start making more sense, because it can be very expensive to run power out to the middle of nowhere.
Ultrawideband is not exactly new, I worked on a UWB radar almost 10 years ago. It has promise, but it also has problems that you don't see with traditional carrier based methods- since you are talking about picosecond resolution, communication with a reciever that is moving with respect to your transmitter is hard- in a picosecond, light travels about .3 mm. Virtually any motion can cause your reciever to start looking at the wrong part of the data- since the data is all time-coded.
Many have said that you can't detect UWB, but usually, they are looking at it with a traditional carrier based spectrum analyzer- you can't see it, but it is there. It just looks like additional noise. which looks very similar to spread spectum- which is exactly what UWB is, but you're spreading the spectrum very, very wide- an impulse in the time domain is uniform (all frequencies, equal amplitude) in the frequency domain.
Time domain has been around, and pretty big in the UWB arena for a long time, but most of what they have is vapor- big promises, but they fail to deliver. There are other players in the field, Aetherwire is one of the other biggies of UWB. There has been a huge amount of money poured into UWB, but dreadfully little usable technology.
Linus brings up an interesting problem with society, although he doesn't go into length about it. That problem is the reliance of capitalism in our society for intellectual advancement.
Unfortunately, scientists still have to pay the rent, and buy food. Science has never been free. Even when people are doing it just for the love of it, someone still has to support the scientist. You must remember that the scientists of the classical age had patrons, usually aristocrats, that paid for their research, supported in much the same way that artists were. Capitalism has changed the patron from an individual to a corporation, and unfortunately, corporations have stockholders that insist on a return on their money.
We've gotten pretty far away from the patron/scientist model of scientific research, in part because there is little prestige in supporting a scientist, and also because science has gotten so complex and expensive. To do much serious scientific research, you have to have many specialists with complex equipment that is often very, very expensive (the machinists that make the stuff need to feed their families too!). Open source is similar to, but not equal to scientific inquiry. It has been proven many times over the past decade that a single person with an idea can make a big difference. You can start working on open source with a minimum amount of expense- what it takes is a desire to create. Now you can do your coding on your own, without a big corporation to back you. In 300 years, this may not be the case, but by then, we should have other technologies to hack.
Competition is supposed to push better products to the fore. If that worked, we'd have light bulbs that laster for 30 years, cars you bought for a lifetime, and software for life.
We would have all those things right now, if it were not for the people that are unwilling to spend $100 on a light-bulb, $200,000 on a car, and $500,000 for a word processor. One way or another, things have to be paid for, otherwise there will be no incentive to produce more. Yeah, capitalism sucks in a lot of ways, but so do all the other -isms. Capitalism has been very successful in inspiring people to create, because by creating they can make a better life for themselves and their progeny.
Until we can divorce the pursuit of capital from advances in science, we are doomed to have any advance kepted restained by the barriers of the a accumulation of that wealth. If at any point, an advancement is deemed to be a money killer, it will be abandoned.
This is not the fault of capitalism, it is the failing of an oligarchy- where the rich have the political power. At that point it is in the best interest of the rich to keep around the mechanisms that keep them rich. The recent attacks on technologies haven't been counter-attacked on the technology front, they have been attacked on the legislative front. Don't try to tear down capitalism, try to build a democracy that isn't financed by the rich, for the rich.
As long as you are keeping the speed down (under 20 MHz or so) wire wrap works well, but as you go higher and higher in frequency, the added inductance and stubs of wire wrapping will kill you. Many of the newer parts are getting harder to find in DIP- especially as the pin count goes up. I don't know of any (modern) common part in a DIP package > 40 pins. On the other hand, 300-400 ball BGAs are getting common- imagine trying to wire wrap that! There is no real market in making a new quad XOR gate in DIP- that technology is mature- they won't really be able to improve on it, except possibly to make it faster, but a DIP package has leads, and leads have inductance, and inductance kills high frequencies, so by default, if you are improving the speed on an IC, you need to shrink it and go to SMT. They are also trying to cram more into a chip, but that implies a greater pin count- again, which pushes you to SMT.
But, as the person who posed the question implied, SMT is difficut for a hobbiest to work with. One alternative is to make your own PCBs, either by etching them at home, or go to some place like APCircuits where you can get prototype PCBs made pretty cheaply. And one hint- if you are doing SMT, get the solder mask, it will make your life very much easier.
I still wire wrap when I need to- and Digi-Key has had most of what I've needed, though the little pins to mount discrete components are outrageously expensive. (In particular, the T68A bifurcated pins, with room for 3 levels of wrap, at $92.25/1000.)
First, in a nutshell, look for PCI-X. It makes PCI a switch based medium instead of a bus, will let you utilize your existing PCI devices..
PCI-X does have some very good advantages, but unfortunately if you plug a PCI device into a PCI-X bus, you've just turned that bus (segment) into a PCI bus, no PCI-X advantages. Plus there are serious limitations on the number of loads (cards) on a PCI-X bus- which makes for more bridges on the motherboard, and a performance hit when you have to cross it.
It _should_ be possible the create a PCI-Specification with let's say 133Mhz, and 64 Bits...
It already exists- its called PCI-X, coming to a computer near you, but it has its own problems, like a limited number of targets per bus (very limited, 1 or 2 I think). One problem with PCI is that it was designed as a low power bus- which is a good idea when you have 64 bits of data coming down the line. Unfortunately that makes it difficult to deal with when you're getting to higher speeds.
PCI has deeper limitations- when a transfer is initated, the target has no idea who is the master of the transaction, which makes it harder to restart a transaction that has been pre-empted. PCI-X attempts to solve this, and other issues that have come up, but maintaining backwards compatibility will only result in band-aid fixes, the fundamental flaws are still there. Sometimes you need to cut your losses and run- like the ISA/EISA to PCI jump- no compatibility there, but higher speed processors make different demands on the bus. As we've built faster and faster systems, we've found new limitations.
This would make transition easier, as old PCI cards would run in the new slots (at least this works for 64Bits Pci-Slots, which can also run 32Bit Card). If this doesn't work for Slots using a higher frequency, the chipsets could include 2 Pci-Controllers, each driving 3 or 4 slots, and each controller could fall back to 33Mhz if there is one card which doesn't support higher speeds.
This technology exists right now- if you have a PCI "bridge" in your system that supports 66MHz/64 bit, that is how it is supposed to work. Unfortunately if your data has to cross a bridge, you get a performance hit.
We can continue patching PCI, or we can learn from our experience and design a new bus.