Let us say for the sake of argument that it is not fully possible to measure all the subtleties of a AUTHENTIC Stradivarius verses an otherwise high grade violin. Then what? What if humans CAN detect things like "warmth" that a scientific measuring instrument can't fully quantify because we aren't able to measure it with scientific instruments?
I once spent several hours helping someone to modify a guitar amplifier to sound "warmer". We achieved what he wanted eventually, and I came to the conclusion that "warmth" is a combination of a smooth, peakless frequency response in the range of 150-1500 Hz, a little attenuation of the frequencies above 2 kHz, and a slight attenuation of odd-order harmonics in the signal. At least in the case of this particular amplifier. Of course, a sample size of one amplifier doesn't make a scientific study, and this may not be a very good comparison, anyway, as one can't exactly change coupling capacitors in a violin to make it sound better. My point is more that I think "warmth" can easily be measured, I just don't think we're particularly good at quantifying the attributes of a sound that we hear as warm and pleasing.
A lot of 5V rectifier diodes are that way because they were evolved from a 4-pin type 80 (a 5Y3 is actually pretty much a type 80 with an octal base), which had a 2.5V/10-watt heater. 4 amps was probably too much for an octal base, so the voltage was doubled. The 6.3V standard was developed initially for automotive applications; cars had 6V batteries at the time. I suspect that you're right; that since a directly-heated rectifer required a separate transformer winding anyway, nobody worried too much about having 5V and 6V secondaries, rather than two 6V secondaries. It wouldn't really change the manufacturing cost of a power transformer either way.
***Each SAGE housed an A/N FSQ-7 computer, which had around 60,000 vacuum tubes. IBM constructed the hardware, and each computer occupied a huge amount of space.***
The sites had two computers, not one. The switched between them once a day so they could check all the vacuum tubes on the off line computer -- of which I'm pretty sure there were only about 6000. Mostly they were 6SN7 dual triodes so there were actually about 12000 switches in each computer. Memory was 68K by 32 bits wide, and software was continually swapped in from drums in the background. Instruction cycle time was 6 microseconds. The specs weren't vastly different from a 1980s IBM PC with 256K of memory.
Great post! The article is sorely lacking in interesting details like these about the old computers pictured.
I can imagine the power and cooling requirements for a computer like this; older octal-base dual-triode voltage amplifier tubes need a fair amount of heater current, as the heaters run at 6 volts. Those tubes also tend to be noisy and susceptible to mechanical interference (the newer 9-pin 12**7 dual-triodes are really much better, the more fragile connecting pins aside); I wonder if you could cause a processing error by tapping on the tubes with a pencil eraser? Maybe not, discrete components of digital computers operate in on/off states, so they handle external interference pretty well.
Obviously, something in our environment is making autism rates climb. But it doesn't look like it's the thimerosol. Even if it is from mercury (which I don't know of any data showing that it is), it seems to be mercury from some other source, not from thimerosol.
Not to mention, worrying about the mercury in your thimerosol causing autism is like worrying about being poisoned by the chlorine in your table salt, or the flammability of the hydrogen in your tap water. Component elements of compounds undergo a chemical reaction when they combine, and don't retain their original properties.
Does anyone remember an article that was posted here several years ago about higher autism rates in areas with a lot of high-tech companies? It's been a while, but I seem to remember that the rates were higher among children of two parents with autistic tendencies themselves, suggesting that the possibility of a genetic link exists.
I must have gotten the same ones. They came in packs of three, don't illuminate much, and the light has a kind of sickly yellow-green cast to it -- they remind me of propane lamps I used to have before I got electricity. I bought two packs of the damn things, too, enough for my whole house, and they weren't exactly cheap.
So why do I have over 100 PSU's in my computer room? Servers with 3 PSU's for redundancy?
Why can't I have a single server room PSU which provides the 12V, 5V, and -5V on some sort of standardized plug? Make each connection a separate fused bus.
Interesting. This seems like quite a good idea on the face of it. I'm not sure if fusing each bus would be a better idea than a single master fuse. If one bus goes down because the fuse overloads, but the rest are fine, I can imagine some possible unpredictable and/or dangerous scenarios, sort of like what happens in a vacuum-tube device when the bias supply fuse (about the dumbest design decision ever) blows. I don't really know enough about computer electronics to know which way is better, though.
That PSU could be situated ideally for cooling, leaving much of the heat out of my server.
I have a few PSU's doing the AC-DC conversion, not well over 100.
Phase two, my PSU is now my 80KVA APC UPS. It's already doing AC-DC then DC back to AC. Then my PSU's go AC back to DC again. Have my APC UPS go AC-DC and run it at 12V, not 48V. Of course you would also need the 5V and -5V step down's too.
Again, a very interesting idea, and the 5V/-5V step-down is easy, just an extra winding on the transformer secondary. As you stated, most of the cooling in a server is for the power supply -- if we move that out of the racks (or even out of the datacenter; put the big UPS and the cooling it needs in another room, and just have power distribution sockets in the datacenter), then cooling the datacenter becomes a lot simpler, and saves more energy than is saved by eliminating a couple of conversion steps.
I used to have a pit stove for summer too -- it was just a hole in the ground with four cinder blocks around it. Could cook in or atop the coals, or warm stuff in the holes in the blocks. Good place to use up the scrap and chips that aren't worth dragging into the house as kindling.
That's pretty much what mine is. Well, I used rocks instead of cinder blocks, but there's little material difference.
Coal actually heats much better than wood, as it puts out way more heat per pound AND the quality of the heat is better -- the room can be the same temperature yet it *feels* a lot warmer (longer wavelength, I think -- more penetrating). And when it's -60 out, wood can't keep up, but coal can. And you can keep a coal fire going continuously all winter. Trouble is, coal is a lot dirtier and more bother, especially with cleaning the chimney! And then there was the year I had to drive to Wyoming and mine my own coal...!!
Interesting -- using coal never actually occurred to me. Though, I have a good woodlot, large enough that I never have to take anything but dead and storm-damaged trees, so using wood makes sense for me. My house isn't large, anyway -- slightly over 1,000 square feet, and with an open layout, so it isn't a challenge to heat. Even when it's -20, I'm usually wearing shorts inside, and I have an upstairs window open a crack.
It's cool to find someone in here who appreciates living like I do -- nice talking to you!
I have something similar, it supplies most of my cooking needs. For the rest, I built a simple fire pit outside for warmer weather -- the stove does heat the house pretty well, undesirably so in July. I do have a tile stove with a larger firebox for my primary heat, though. Nothing beats a wood stove for heat, either. And, with almost no moving parts, they're pretty low-maintenance.
I vastly prefer living out by myself too -- it is indeed far less stressful, not to mention less annoying. Most of what people think of as urban necessities, I do quite well without, or find some other way to manage, or would rather not be bothered with in the first place (this troglodyte doesn't have a cell phone and doesn't WANT one!)
You are right -- "reasonably ubiquitous" really only applies to populated areas and travel corridors, where probably 99% of the population lives. Get very far from settled and traveled areas, and public services of this sort are rare to nil... tho I've noticed that areas settled a long time ago are far more likely to have run wires to remote (unprofitable to the provider) ranches, whereas the "newer" states are less likely to do so. Might be a side effect of rural co-ops.
It's possible. Another possibility is New Deal-era buildup done by the Rural Electrification Administration.
I've lived where neither power nor phone was available, and survived the experience (this was before solar panels and cell phones, too!) much as did most of our ancestors:)
I like it better this way. Much of that is undoubtedly habit, but I found that when I lived in a populated area, and had a television with fifty channels, a computer with a broadband connection, video games, air conditioning, and such, my stress level was higher, I didn't sleep as well, and I was generally unhappy. I became much more centered and relaxed when I gave up all of that and moved.
I do like my solar panel, though -- electric lights are superior to oil lamps in every possible way.:)
There are more issues that make RF communications on power lines very difficult. Growing numbers of devices plugged into the AC line generate RF noise, which must be controlled to meet FCC regulations (and overseas equivalents). Manufacturers of switch-mode power supplies include filtering to meet those requirements. That means there's a filter cap lurking inside, shorting out BPL signal on the power line (in this typical example, it's C1). Every time you plug in another device, the AC line transmission path is further compromised.
In your linked example, filtering is largely handled by the two inductors L2 and L3, just behind the rectifier. The ferrite bead (L1) provides some extra filtering of induced noise as well. C1 is essentially an archaic safety device, and unnecessary, as notebook power supplies are invariably class 2 devices, though it may be providing some filtering, depending on the impedance of the circuit. The value seems too high for that, though --.33 uF is well into the audio range as a filter, unless I'm misreading the circuit.
This is a somewhat odd power supply, anyway. In particular, the ripple current filter (C2) is on the primary side of the transformer, and has no bleeder resistor. It would be much safer, and a bit cheaper, to filter the transformer secondary. Maybe the switch coupling requires the DC supply to be filtered, though, I'm not intimately familiar with switch-mode supplies.
Electricity and basic phone lines have been in most of the American hinterland for decades -- tho there are parts of Montana that got power in my lifetime, and still lack phone service. Some parts of California still lack both. But overall, power and phone lines are reasonably ubiquitous.
Actually, only a few percent of the country's land area is covered by electric and phone service. It just happens that almost the entire population lives within that small land area. My parents have lived in various places around the country, and neither of them have ever lived anywhere with these services. I have neither electric nor telephone service where I live, and I can't get it without paying somewhere in the high six-figure range to have the lines extended. I'm okay with that -- I have a solar panel, which generates all the electricity I need, and I hate telephones anyway (though I'm at a high enough elevation that I could probably get cellular service if I cared to).
Most people put aside something for education as an escape from the news and their own problems. What that entertainment may be and how much someone's willing to spend is going to vary.
Absolutely. Though, people don't always spend wisely. When I originally had a TV subscription, my wife and I would only ever watch the channels which were available with the lower-priced subscription, and we wasted a lot of money on nothing. This somewhat informs my choice of where to spend money on entertainment. Now, I mostly read. I used to buy a fair amount of books (and I still buy them occasionally), but now I mostly borrow them from the local library for nothing. I get identical entertainment for a fraction of the cost. Ultimately, though, I canceled all of that stuff because I wasn't using it anyway.
Cable TV I could easily toss aside, but internet and cell service isn't going to happen (especially since the rise in cell subscriptions has lead to a decline in the availability of pay phones). I don't see a need for a home phone, but some still cling to that as if it's somehow more likely to still work in an emergency or disaster situation than their cell phone.
I don't see a telephone as essential for me. However, I grew up without one, so my atypical worldview no doubt informs my decisions in this area.
I really don't see cutting off the cell phone and internet as entertainment related, though. I see it more as cutting myself off from my family and friends, most of which are on the other side of the country (in fact neither my wife nor myself has any family within an hour's drive, and most of my family is on the opposite coast of the US).
Well, when I had a telephone, I never talked to anyone but telemarketers on it. Actually, I do very little in the way of communication in general -- there just isn't anything interesting left to say about the weather or someone's in-laws/coworkers or how much better we all are than everyone else, in my opinion. I do periodically find a nice conversation on a new subject (this one, for example), though, which makes it all worthwhile.
All that said, I don't disagree with your way of doing things -- if it works for you, more power to you. I'm okay, you're okay.:)
I've actually canceled my TV, internet, and cell/telephone services over the past couple of years. I'm not broke, I just wasn't using them enough to justify the expense to myself. On the rare occasion when I want to use the 'net or call someone, I can use the library or a payphone, respectively. I actually find that I'm happier and less stressed without these things, and that I increasingly prefer isolation as I get older.
I don't think that you're wrong in your assertion, your post was just interesting to me.
I was rear-ended at a fairly high-speed (I was sitting at a stop light) a few years ago in a Volvo, by a Saturn.
The Saturn was a mess (and literally *bounced* off of the Volvo). My car needed a new bumper and a bit of paint*.
The same thing happened to my wife a couple of years ago; she was waiting at a stop light on the first really nice day of the spring, and a teenager rear-ended her at 45 MPH because he wasn't paying enough attention. Her Corolla needed a rear bumper and a couple of other things around it. I drove it home after the accident, and we still have it. His car was ruined -- it was about a meter shorter, and the front wheels were kilted at about a 45-degree angle to eachother. I think it was a Ford Focus, actually.
Note that none of the changes Ford is proposing would have helped in this situation. That kid was driving well under 80 MPH, and he didn't use his brakes, so traction control wasn't relevant. Not that the changes are a bad idea (I think they are actually quite innovative), it just shouldn't be expected that they will completely prevent accidents -- nothing will.
Actually, I think 80 MPH is a pretty generous governor -- I've never driven a car that fast, period, and thinking that extra speed is needed to avoid an accident is flawed -- most accidents occur at 30 MPH or less. If I were designing this system, I'd set it to 50 MPH, along with disabling the stereo and locking the windows to reduce distractions. Actually, why not make the governor configurable, so that the car's owner could set the parameters where they feel is appropriate? People live in different areas with different driving situations -- my idea of what is appropriate isn't likely to work everywhere.
Actually both liberals and conservatives want to protect choice--just on different issues. Conservatives want to protect your choice to spend your money but want to apply their concepts of morality on society. Liberals, meanwhile, don't really care what you do morally speaking as long as you do what they tell you to do with your money.
This is something of a strawman argument. Liberals don't care how anyone spends their money, they (like conservatives) just like to use tax proceeds to fund their pet programs. The difference is really in taxation theories, and liberals do try to push the burden toward the upper end. However, the difference amounts to probably five percent -- who gives a shit?
I'd rather see them cut the tax rates below ten percent somewhere, and lose most of the pet programs, but neither group seems particularly interested in anything like that.
Which is better? In my opinion a moral society in which people can do what they want with their money is desirable to a morally corrupt society where everything goes as long as you're paying extortion money to the liberal government. But that's just my opinion.
Here, I would disagree with you. While I would like to live in a moral society, morality that is forced at the barrel of a gun isn't. Not to mention, if we do that, whose morality gets enforced? Yours, mine, whoever is currently in power? The latter is obviously most likely, and once the precedent is set to allow government to have that power, what happens if, say, what the conservatives are saying is true (not that I think it is), that Barack Obama is a secret muslim. Then he has the power to enforce Sharia law on us if he gets elected. Do you want that? It's a very real possibility in your vision of a moral society.
As tiresome as it would be, I would prefer to live in a society where there the streets are lined with brothels and vendors of mind-altering substances, because I know in that society, everyone would have the right to practice their own morality in the way they see fit.
In order to ignite the gasoline or jet fuel (essentially high-octane kerosene), you need two other things: a spark which brings the temperature of the fuel to its flash point, and oxygen. You're absolutely right, a low-voltage/current sensor would never be able to spark enough to bring the fuel to its flash point, and if it's submerged in the fuel, it doesn't matter anyway, there's no oxygen available for ignition. Properly insulate the sensor, and it's extremely safe.
Actually, gasoline can be used to put out a fire under certain conditions (I wouldn't recommend it as a practical use, as there are plenty of substances better-suited to the task).
They do usually have serial console redirection functionality, which allows boot-time access to CMOS, and works well with my Linux/BSD systems as well. I've never been able to get the thing working on my Windows servers, so I've been using a KVM with those. I only have a few running Windows, so it hasn't been a big deal. To be honest, too, I haven't had a lot of time to put into solving the problem, and I'm a better Unix admin than a Windows admin, so it's possible that it does work, it just may not be as dead easy to get running as it is with Linux/BSD.
A USB floppy works fine (I stopped ordering servers with integrated floppies and bought a couple of USB units -- it saves a few bucks, and I only use the drives at install time anyway). However, maybe you can speak to someone about bringing back the PS/2 keyboard and mouse ports. Those of us with dozens of servers in racks need to use KVM switches, and USB is annoying at best in that capacity, with the hardware redetection delay upon switching consoles. Not only that, USB KVM switches are somewhat harder to come by than PS/2 units, depending on the options required.
CERT has covered vulnerabilities in OSS products, apparently just not since 2004, which is as far back as the online archive you linked goes. I don't remember seeing anything about them discontinuing their coverage of OSS products, did they do that somewhere along the way? I know there have been holes in OSS products since 2004, have they just not been significant enough to warrant CERT's attention?
There's an analogous situation here in Colorado: you can't buy a bottle of liquor on Sunday. The state isn't banning it to save your soul; you're welcome to drink your way to perdition in a bar. The reason? Sunday closing is much more harmful to total by-the-drink sales than it is to total package sales, and business overhead is substantially higher for a 7-day store than for a 6-day store. So bars stay open on Sunday, liquor stores close, and they're both happy. Every attempt to repeal the Sunday-closing law is shot down by the liquor business.
There's a similar law here in New York. A couple of years ago there was a move to repeal it, and it happened exactly as you describe -- the liquor lobby shot it down based on the increased costs that liquor store owners would incur. What I don't get is why a law about this is necessary -- if you incur greater costs than benefits by keeping your store open 7 days a week rather than 6, then don't open the store on the 7th day -- it's not like the removal of the law would force you to do so. I guess they're probably worried that their competition will be able to open on Sunday and they'll lose business to them. If they can't function in a free market, their business model is probably flawed.
The OS or hardware platforms all have faults (with the possible exception of OpenVMS on Alphas).
I seem to remember it was OpenVMS on the VAX platform that had the hardware-enforced security contexts, but it could have been there on the Alpha as well. I used to admin VAXen and Alphas until Unix and WinNT took over, and I assure you that the much-vaunted security didn't mean much, although it was better than many other systems available at the time. For one thing, the OS was unfriendly enough that getting everything configured in a properly secure fashion could be a challenge for some admins, and I always thought it relied a bit too much on security-by-obscurity due to the same unfriendlyness. Also, a lot of applications didn't have much of their own security -- who cares if the admin account on the machine is impervious if the database instance on the machine just got owned by a buffer-overflow attack, compromising all your customer data?
MMORPG bandwidth usage is generally quite low, for simple economic reasons. MMORPG companies have to pay for bandwidth to their datacenter, so they have a financial interest in each of their thousands of clients using as little bandwidth as possible. I can't speak to EQ2 specifically, as I don't have any experience with it, but the MMOs I've played all use modest amounts of bandwidth (average is probably around 20 kilobits/sec, with surges when during large battles or when entering populated areas).
I play on two servers (my clan plays both Alliance and Horde on PvP servers, where you can only have characters of one faction, so we picked one for each.) One server is a medium-population server, and has never had significant problems (the occasional bugged loot node, logging out and back in fixes it). The other one is one of the high-population servers that is now eligible for character transfer. It isn't *that* bad overall -- occasional disconnects, minor "loot lag", and the like. There are login queues at peak times when the realm gets full, but after the starting area congestion the first week, I never saw a queue more than ten minutes long.
I actually think Blizzard did a fine job considering how much they underestimated demand, though I have to say they should have had a clue that it was coming after half a million open beta signups, and the buzz that the beta generated. I went to buy the game on release day when the local Electronics Boutique opened, and I wouldn't have gotten a copy if I hadn't reserved one. I was in line with 50 other people, including, no joke, a man dressed as an orc, complete with green makeup, who said *nothing* but "Zug-zug" and "Lok-tar" during his entire transaction. And I live in a smallish town (around 50,000 people).
Currently, my understanding of MMORPGs is that they keep a copy of the world on each server.
Yeah, each server (more like a small cluster, really -- A WoW session opens connections to three or more distinct IP addresses depending on which zones you visit) holds its own copy of the game world, and what happens on one server has no effect on any of the others.
Why can't MMORPGs set up the server-clusters so that the world is partitioned into seperate zones on the servers, depending on load. So a player can (say) log onto the European server cluster, enter zone Foo, and the zones Foo, Bar, and Baz are currently located on server #5. If zone Bar gets too many players on it, zone Bar is seemlessly transferred to a mostly idle server along with all of its players.
Dynamic load balancing across the realm cluster is actually a really good idea. It might be overkill for WoW, as the game world really isn't that large -- it would likely become uncomfortably crowded with more than a few thousand people logged into a single realm. For a game with a larger play area and a single realm (the upcoming Wish is like this) the approach makes a lot of sense. Decentralizing it in a p2p fashion might be another way to go, but that approach isn't without its own problems.
Slashdot Mirror
Let us say for the sake of argument that it is not fully possible to measure all the subtleties of a AUTHENTIC Stradivarius verses an otherwise high grade violin. Then what? What if humans CAN detect things like "warmth" that a scientific measuring instrument can't fully quantify because we aren't able to measure it with scientific instruments?
I once spent several hours helping someone to modify a guitar amplifier to sound "warmer". We achieved what he wanted eventually, and I came to the conclusion that "warmth" is a combination of a smooth, peakless frequency response in the range of 150-1500 Hz, a little attenuation of the frequencies above 2 kHz, and a slight attenuation of odd-order harmonics in the signal. At least in the case of this particular amplifier. Of course, a sample size of one amplifier doesn't make a scientific study, and this may not be a very good comparison, anyway, as one can't exactly change coupling capacitors in a violin to make it sound better. My point is more that I think "warmth" can easily be measured, I just don't think we're particularly good at quantifying the attributes of a sound that we hear as warm and pleasing.
A lot of 5V rectifier diodes are that way because they were evolved from a 4-pin type 80 (a 5Y3 is actually pretty much a type 80 with an octal base), which had a 2.5V/10-watt heater. 4 amps was probably too much for an octal base, so the voltage was doubled. The 6.3V standard was developed initially for automotive applications; cars had 6V batteries at the time. I suspect that you're right; that since a directly-heated rectifer required a separate transformer winding anyway, nobody worried too much about having 5V and 6V secondaries, rather than two 6V secondaries. It wouldn't really change the manufacturing cost of a power transformer either way.
***Each SAGE housed an A/N FSQ-7 computer, which had around 60,000 vacuum tubes. IBM constructed the hardware, and each computer occupied a huge amount of space.***
The sites had two computers, not one. The switched between them once a day so they could check all the vacuum tubes on the off line computer -- of which I'm pretty sure there were only about 6000. Mostly they were 6SN7 dual triodes so there were actually about 12000 switches in each computer. Memory was 68K by 32 bits wide, and software was continually swapped in from drums in the background. Instruction cycle time was 6 microseconds. The specs weren't vastly different from a 1980s IBM PC with 256K of memory.
Great post! The article is sorely lacking in interesting details like these about the old computers pictured.
I can imagine the power and cooling requirements for a computer like this; older octal-base dual-triode voltage amplifier tubes need a fair amount of heater current, as the heaters run at 6 volts. Those tubes also tend to be noisy and susceptible to mechanical interference (the newer 9-pin 12**7 dual-triodes are really much better, the more fragile connecting pins aside); I wonder if you could cause a processing error by tapping on the tubes with a pencil eraser? Maybe not, discrete components of digital computers operate in on/off states, so they handle external interference pretty well.
Obviously, something in our environment is making autism rates climb. But it doesn't look like it's the thimerosol. Even if it is from mercury (which I don't know of any data showing that it is), it seems to be mercury from some other source, not from thimerosol.
Not to mention, worrying about the mercury in your thimerosol causing autism is like worrying about being poisoned by the chlorine in your table salt, or the flammability of the hydrogen in your tap water. Component elements of compounds undergo a chemical reaction when they combine, and don't retain their original properties.
Does anyone remember an article that was posted here several years ago about higher autism rates in areas with a lot of high-tech companies? It's been a while, but I seem to remember that the rates were higher among children of two parents with autistic tendencies themselves, suggesting that the possibility of a genetic link exists.
That's funny, I do the same thing when your wife wants to spice things up in the bedroom.
(Sorry, I couldn't resist =P)
I must have gotten the same ones. They came in packs of three, don't illuminate much, and the light has a kind of sickly yellow-green cast to it -- they remind me of propane lamps I used to have before I got electricity. I bought two packs of the damn things, too, enough for my whole house, and they weren't exactly cheap.
Oh, well -- live and learn, I guess.
So why do I have over 100 PSU's in my computer room? Servers with 3 PSU's for redundancy?
Why can't I have a single server room PSU which provides the 12V, 5V, and -5V on some sort of standardized plug? Make each connection a separate fused bus.
Interesting. This seems like quite a good idea on the face of it. I'm not sure if fusing each bus would be a better idea than a single master fuse. If one bus goes down because the fuse overloads, but the rest are fine, I can imagine some possible unpredictable and/or dangerous scenarios, sort of like what happens in a vacuum-tube device when the bias supply fuse (about the dumbest design decision ever) blows. I don't really know enough about computer electronics to know which way is better, though.
That PSU could be situated ideally for cooling, leaving much of the heat out of my server.
I have a few PSU's doing the AC-DC conversion, not well over 100.
Phase two, my PSU is now my 80KVA APC UPS. It's already doing AC-DC then DC back to AC. Then my PSU's go AC back to DC again. Have my APC UPS go AC-DC and run it at 12V, not 48V. Of course you would also need the 5V and -5V step down's too.
Again, a very interesting idea, and the 5V/-5V step-down is easy, just an extra winding on the transformer secondary. As you stated, most of the cooling in a server is for the power supply -- if we move that out of the racks (or even out of the datacenter; put the big UPS and the cooling it needs in another room, and just have power distribution sockets in the datacenter), then cooling the datacenter becomes a lot simpler, and saves more energy than is saved by eliminating a couple of conversion steps.
I used to have a pit stove for summer too -- it was just a hole in the ground with four cinder blocks around it. Could cook in or atop the coals, or warm stuff in the holes in the blocks. Good place to use up the scrap and chips that aren't worth dragging into the house as kindling.
That's pretty much what mine is. Well, I used rocks instead of cinder blocks, but there's little material difference.
Coal actually heats much better than wood, as it puts out way more heat per pound AND the quality of the heat is better -- the room can be the same temperature yet it *feels* a lot warmer (longer wavelength, I think -- more penetrating). And when it's -60 out, wood can't keep up, but coal can. And you can keep a coal fire going continuously all winter. Trouble is, coal is a lot dirtier and more bother, especially with cleaning the chimney! And then there was the year I had to drive to Wyoming and mine my own coal...!!
Interesting -- using coal never actually occurred to me. Though, I have a good woodlot, large enough that I never have to take anything but dead and storm-damaged trees, so using wood makes sense for me. My house isn't large, anyway -- slightly over 1,000 square feet, and with an open layout, so it isn't a challenge to heat. Even when it's -20, I'm usually wearing shorts inside, and I have an upstairs window open a crack.
It's cool to find someone in here who appreciates living like I do -- nice talking to you!
Having lived with propane lamps and candles.. I agree with you, electric light is superior in almost every way :) OTOH, nothing beats a really good wood stove for cooking. http://www.kountrylife.com/cgi-bin/coll_pic.cgi?coll=cookstoves&picfile=ccblk.jpg&mode=All&Parameter=&SelectParameter=All&firstrec=1&lastrec=15 I once lived in a place that had one of this model. Great stove to cook on! Baked stuff really evenly. Heated the house. :)
I have something similar, it supplies most of my cooking needs. For the rest, I built a simple fire pit outside for warmer weather -- the stove does heat the house pretty well, undesirably so in July. I do have a tile stove with a larger firebox for my primary heat, though. Nothing beats a wood stove for heat, either. And, with almost no moving parts, they're pretty low-maintenance.
I vastly prefer living out by myself too -- it is indeed far less stressful, not to mention less annoying. Most of what people think of as urban necessities, I do quite well without, or find some other way to manage, or would rather not be bothered with in the first place (this troglodyte doesn't have a cell phone and doesn't WANT one!)
Agreed, 100%.
You are right -- "reasonably ubiquitous" really only applies to populated areas and travel corridors, where probably 99% of the population lives. Get very far from settled and traveled areas, and public services of this sort are rare to nil... tho I've noticed that areas settled a long time ago are far more likely to have run wires to remote (unprofitable to the provider) ranches, whereas the "newer" states are less likely to do so. Might be a side effect of rural co-ops.
It's possible. Another possibility is New Deal-era buildup done by the Rural Electrification Administration.
I've lived where neither power nor phone was available, and survived the experience (this was before solar panels and cell phones, too!) much as did most of our ancestors :)
I like it better this way. Much of that is undoubtedly habit, but I found that when I lived in a populated area, and had a television with fifty channels, a computer with a broadband connection, video games, air conditioning, and such, my stress level was higher, I didn't sleep as well, and I was generally unhappy. I became much more centered and relaxed when I gave up all of that and moved.
I do like my solar panel, though -- electric lights are superior to oil lamps in every possible way. :)
There are more issues that make RF communications on power lines very difficult. Growing numbers of devices plugged into the AC line generate RF noise, which must be controlled to meet FCC regulations (and overseas equivalents). Manufacturers of switch-mode power supplies include filtering to meet those requirements. That means there's a filter cap lurking inside, shorting out BPL signal on the power line (in this typical example, it's C1). Every time you plug in another device, the AC line transmission path is further compromised.
In your linked example, filtering is largely handled by the two inductors L2 and L3, just behind the rectifier. The ferrite bead (L1) provides some extra filtering of induced noise as well. C1 is essentially an archaic safety device, and unnecessary, as notebook power supplies are invariably class 2 devices, though it may be providing some filtering, depending on the impedance of the circuit. The value seems too high for that, though -- .33 uF is well into the audio range as a filter, unless I'm misreading the circuit.
This is a somewhat odd power supply, anyway. In particular, the ripple current filter (C2) is on the primary side of the transformer, and has no bleeder resistor. It would be much safer, and a bit cheaper, to filter the transformer secondary. Maybe the switch coupling requires the DC supply to be filtered, though, I'm not intimately familiar with switch-mode supplies.
Electricity and basic phone lines have been in most of the American hinterland for decades -- tho there are parts of Montana that got power in my lifetime, and still lack phone service. Some parts of California still lack both. But overall, power and phone lines are reasonably ubiquitous.
Actually, only a few percent of the country's land area is covered by electric and phone service. It just happens that almost the entire population lives within that small land area. My parents have lived in various places around the country, and neither of them have ever lived anywhere with these services. I have neither electric nor telephone service where I live, and I can't get it without paying somewhere in the high six-figure range to have the lines extended. I'm okay with that -- I have a solar panel, which generates all the electricity I need, and I hate telephones anyway (though I'm at a high enough elevation that I could probably get cellular service if I cared to).
Most people put aside something for education as an escape from the news and their own problems. What that entertainment may be and how much someone's willing to spend is going to vary.
Absolutely. Though, people don't always spend wisely. When I originally had a TV subscription, my wife and I would only ever watch the channels which were available with the lower-priced subscription, and we wasted a lot of money on nothing. This somewhat informs my choice of where to spend money on entertainment. Now, I mostly read. I used to buy a fair amount of books (and I still buy them occasionally), but now I mostly borrow them from the local library for nothing. I get identical entertainment for a fraction of the cost. Ultimately, though, I canceled all of that stuff because I wasn't using it anyway.
Cable TV I could easily toss aside, but internet and cell service isn't going to happen (especially since the rise in cell subscriptions has lead to a decline in the availability of pay phones). I don't see a need for a home phone, but some still cling to that as if it's somehow more likely to still work in an emergency or disaster situation than their cell phone.
I don't see a telephone as essential for me. However, I grew up without one, so my atypical worldview no doubt informs my decisions in this area.
I really don't see cutting off the cell phone and internet as entertainment related, though. I see it more as cutting myself off from my family and friends, most of which are on the other side of the country (in fact neither my wife nor myself has any family within an hour's drive, and most of my family is on the opposite coast of the US).
Well, when I had a telephone, I never talked to anyone but telemarketers on it. Actually, I do very little in the way of communication in general -- there just isn't anything interesting left to say about the weather or someone's in-laws/coworkers or how much better we all are than everyone else, in my opinion. I do periodically find a nice conversation on a new subject (this one, for example), though, which makes it all worthwhile. All that said, I don't disagree with your way of doing things -- if it works for you, more power to you. I'm okay, you're okay. :)
I've actually canceled my TV, internet, and cell/telephone services over the past couple of years. I'm not broke, I just wasn't using them enough to justify the expense to myself. On the rare occasion when I want to use the 'net or call someone, I can use the library or a payphone, respectively. I actually find that I'm happier and less stressed without these things, and that I increasingly prefer isolation as I get older.
I don't think that you're wrong in your assertion, your post was just interesting to me.
I was rear-ended at a fairly high-speed (I was sitting at a stop light) a few years ago in a Volvo, by a Saturn.
The Saturn was a mess (and literally *bounced* off of the Volvo). My car needed a new bumper and a bit of paint*.
The same thing happened to my wife a couple of years ago; she was waiting at a stop light on the first really nice day of the spring, and a teenager rear-ended her at 45 MPH because he wasn't paying enough attention. Her Corolla needed a rear bumper and a couple of other things around it. I drove it home after the accident, and we still have it. His car was ruined -- it was about a meter shorter, and the front wheels were kilted at about a 45-degree angle to eachother. I think it was a Ford Focus, actually.
Note that none of the changes Ford is proposing would have helped in this situation. That kid was driving well under 80 MPH, and he didn't use his brakes, so traction control wasn't relevant. Not that the changes are a bad idea (I think they are actually quite innovative), it just shouldn't be expected that they will completely prevent accidents -- nothing will.
Actually, I think 80 MPH is a pretty generous governor -- I've never driven a car that fast, period, and thinking that extra speed is needed to avoid an accident is flawed -- most accidents occur at 30 MPH or less. If I were designing this system, I'd set it to 50 MPH, along with disabling the stereo and locking the windows to reduce distractions. Actually, why not make the governor configurable, so that the car's owner could set the parameters where they feel is appropriate? People live in different areas with different driving situations -- my idea of what is appropriate isn't likely to work everywhere.
Actually both liberals and conservatives want to protect choice--just on different issues. Conservatives want to protect your choice to spend your money but want to apply their concepts of morality on society. Liberals, meanwhile, don't really care what you do morally speaking as long as you do what they tell you to do with your money.
This is something of a strawman argument. Liberals don't care how anyone spends their money, they (like conservatives) just like to use tax proceeds to fund their pet programs. The difference is really in taxation theories, and liberals do try to push the burden toward the upper end. However, the difference amounts to probably five percent -- who gives a shit?
I'd rather see them cut the tax rates below ten percent somewhere, and lose most of the pet programs, but neither group seems particularly interested in anything like that.
Which is better? In my opinion a moral society in which people can do what they want with their money is desirable to a morally corrupt society where everything goes as long as you're paying extortion money to the liberal government. But that's just my opinion.
Here, I would disagree with you. While I would like to live in a moral society, morality that is forced at the barrel of a gun isn't. Not to mention, if we do that, whose morality gets enforced? Yours, mine, whoever is currently in power? The latter is obviously most likely, and once the precedent is set to allow government to have that power, what happens if, say, what the conservatives are saying is true (not that I think it is), that Barack Obama is a secret muslim. Then he has the power to enforce Sharia law on us if he gets elected. Do you want that? It's a very real possibility in your vision of a moral society.
As tiresome as it would be, I would prefer to live in a society where there the streets are lined with brothels and vendors of mind-altering substances, because I know in that society, everyone would have the right to practice their own morality in the way they see fit.
In order to ignite the gasoline or jet fuel (essentially high-octane kerosene), you need two other things: a spark which brings the temperature of the fuel to its flash point, and oxygen. You're absolutely right, a low-voltage/current sensor would never be able to spark enough to bring the fuel to its flash point, and if it's submerged in the fuel, it doesn't matter anyway, there's no oxygen available for ignition. Properly insulate the sensor, and it's extremely safe.
Actually, gasoline can be used to put out a fire under certain conditions (I wouldn't recommend it as a practical use, as there are plenty of substances better-suited to the task).
They do usually have serial console redirection functionality, which allows boot-time access to CMOS, and works well with my Linux/BSD systems as well. I've never been able to get the thing working on my Windows servers, so I've been using a KVM with those. I only have a few running Windows, so it hasn't been a big deal. To be honest, too, I haven't had a lot of time to put into solving the problem, and I'm a better Unix admin than a Windows admin, so it's possible that it does work, it just may not be as dead easy to get running as it is with Linux/BSD.
A USB floppy works fine (I stopped ordering servers with integrated floppies and bought a couple of USB units -- it saves a few bucks, and I only use the drives at install time anyway). However, maybe you can speak to someone about bringing back the PS/2 keyboard and mouse ports. Those of us with dozens of servers in racks need to use KVM switches, and USB is annoying at best in that capacity, with the hardware redetection delay upon switching consoles. Not only that, USB KVM switches are somewhat harder to come by than PS/2 units, depending on the options required.
CERT has covered vulnerabilities in OSS products, apparently just not since 2004, which is as far back as the online archive you linked goes. I don't remember seeing anything about them discontinuing their coverage of OSS products, did they do that somewhere along the way? I know there have been holes in OSS products since 2004, have they just not been significant enough to warrant CERT's attention?
There's a similar law here in New York. A couple of years ago there was a move to repeal it, and it happened exactly as you describe -- the liquor lobby shot it down based on the increased costs that liquor store owners would incur. What I don't get is why a law about this is necessary -- if you incur greater costs than benefits by keeping your store open 7 days a week rather than 6, then don't open the store on the 7th day -- it's not like the removal of the law would force you to do so. I guess they're probably worried that their competition will be able to open on Sunday and they'll lose business to them. If they can't function in a free market, their business model is probably flawed.
I seem to remember it was OpenVMS on the VAX platform that had the hardware-enforced security contexts, but it could have been there on the Alpha as well. I used to admin VAXen and Alphas until Unix and WinNT took over, and I assure you that the much-vaunted security didn't mean much, although it was better than many other systems available at the time. For one thing, the OS was unfriendly enough that getting everything configured in a properly secure fashion could be a challenge for some admins, and I always thought it relied a bit too much on security-by-obscurity due to the same unfriendlyness. Also, a lot of applications didn't have much of their own security -- who cares if the admin account on the machine is impervious if the database instance on the machine just got owned by a buffer-overflow attack, compromising all your customer data?
MMORPG bandwidth usage is generally quite low, for simple economic reasons. MMORPG companies have to pay for bandwidth to their datacenter, so they have a financial interest in each of their thousands of clients using as little bandwidth as possible. I can't speak to EQ2 specifically, as I don't have any experience with it, but the MMOs I've played all use modest amounts of bandwidth (average is probably around 20 kilobits/sec, with surges when during large battles or when entering populated areas).
I actually think Blizzard did a fine job considering how much they underestimated demand, though I have to say they should have had a clue that it was coming after half a million open beta signups, and the buzz that the beta generated. I went to buy the game on release day when the local Electronics Boutique opened, and I wouldn't have gotten a copy if I hadn't reserved one. I was in line with 50 other people, including, no joke, a man dressed as an orc, complete with green makeup, who said *nothing* but "Zug-zug" and "Lok-tar" during his entire transaction. And I live in a smallish town (around 50,000 people).
Yeah, each server (more like a small cluster, really -- A WoW session opens connections to three or more distinct IP addresses depending on which zones you visit) holds its own copy of the game world, and what happens on one server has no effect on any of the others.
Why can't MMORPGs set up the server-clusters so that the world is partitioned into seperate zones on the servers, depending on load. So a player can (say) log onto the European server cluster, enter zone Foo, and the zones Foo, Bar, and Baz are currently located on server #5. If zone Bar gets too many players on it, zone Bar is seemlessly transferred to a mostly idle server along with all of its players.
Dynamic load balancing across the realm cluster is actually a really good idea. It might be overkill for WoW, as the game world really isn't that large -- it would likely become uncomfortably crowded with more than a few thousand people logged into a single realm. For a game with a larger play area and a single realm (the upcoming Wish is like this) the approach makes a lot of sense. Decentralizing it in a p2p fashion might be another way to go, but that approach isn't without its own problems.