In this case, having one bad guy with the directions really isn't any better than 10 bad guys with it.
In fact, the more bad guys that have it, the more likely the problem will get fixed, thus it's actually better that the most 'bad guys' possible get it. If only one person knows how to rig the election, chances are higher they'll be able to get away with it. If 100 people know and all try to rig the election, chances are none of them will get away with it, because the tampering will be too obvious.
Frankly I think the best thing to happen would be for someone to utterly steal the next election; make "Mickey Mouse" or "Elvis" get 100 seats in Congress or something. The cost of having to repeat a single election is certainly much smaller than continuing for decades with a flawed process, where nobody can tell whether the vote is being rigged or not.
Well, given the price in gas (last winter we were paying upwards of $400 a month in gas alone to heat a single-story, seven-room manufactured house), if I was building, I'd be willing to get a heat pump and use electric oil-filled radiators as a backup if it got so cold that it stopped working completely. (And save the money for the air conditioning in the summertime, when this place turns into a humid suburb of Hell. The fact that I'm not originally from here is probably starting to show...)
As it is, my housemates and I have decided we're not going to use the gas furnace for anything more than the minimum required to keep the pipes from freezing this season. It's old and inefficient enough that by the time the heat gets from the furnace to the living areas of the house, it costs more than just putting a $40 oil-filled electric radiator there does. Plus, you can put the electric rads on timers (some have built-in ones) and control the temperature in different rooms over time. Sure, you can do that with a timed thermostat on your central heat, but the control you get via individual radiators is far more fine-grained. Add an electric blanket on the bed and a ceramic fan-forced heater in the bathroom, and you really don't need to keep the entire house cooking at 72F. (It probably helps that I dislike an ambient temperature greater than about 65F anyway, summer or winter.)
I wonder if as energy prices continue to increase, they'll be more interest in heat pump systems that use heat resovoirs other than the atmosphere; for example the ground below the frost line, or above-freezing water at the bottom of a lake or pond, and additional types of heat pumps designed to function at very low temperatures. Maybe a solution would be to not try and have the same unit do winter heating and summer cooling? It seems like the dual-functionality probably requires some tradeoffs; if you used two separate units, one that was designed for cold-weather operation, pulling heat in, and another that was designed for hot-weather operation, exhausting heat out, maybe they would be more practical in northern climates.
Ignoring someone who's wrong, when arguing with them would only make them more powerful, is both good and correct.
However, ignoring someone who has a gun to your head, and is asking you for a good reason why they shouldn't pull the trigger, seems rather shortsighted.
The videogames industry is, right now, in the second position. Maybe Congress doesn't have the gun to their head yet, but they're fiddling around trying to take the safety off and figure out which end to hold.
Now is not a good time to just ignore the government and hope it'll go away. That sort of attitude earned us the DMCA; wouldn't it have been nice if the EFF had been around back in the 1980s, when its the precursor laws (in particular, the anti-decryption laws regarding satellite TV) were passed?
Unfortunately, when you stay home and ignore what's going on, it doesn't keep what happens in the absence of your attention from affecting you later.
Basically, what I'm trying to say is that the Gaming Industry could show up to an event like this and have God as a witness and no one there will listen to them when they say videogames do not cause children to perform violent acts.
That pretty much sums the whole thing up in one line.
Actually, you could show up in any Congressional subcommittee with God in tow, and unless God happened to be made out of money, I doubt you'd influence any pending piece of legislation.
If the "games lobby" wants to make its voice heard in government, and keep itself from being run over as the Fox News scarecrow-du-jour, then they should take a very good look at what the National Rifle Association does, in terms of communicating with and mobilizing its support base, getting donations, and funneling those donations to where they'll have maximum political impact. I can't think of any organization that is as frankly successful and powerful as they are, and has continuously maintained such a high profile, and has done it while staying within the bounds of the law. (Some corporate lobbies might come close, but I think their cash burn rates are much higher for the effect they achieve.)
You can have logical arguments so beautiful they'd make Plato sit down and weep, enough scientific evidence to unequivocally prove a dozen theories of everything, but the government will still ignore you if you are not either a large force among voters, or have lots of filthy lucre to burn. Preferably, have both.
Ok well then he should donate money to the gaim developers or sponsor them in some other way. There are other ways to help out.
Definitely! This is one reason that I dislike the "code it yourself" response, because I think it turns people off and makes them believe that if you don't read and write and breathe C, you'll never have any impact or value in OSS development. There are lots of ways to help out, including straightforward financial donations, which are open to many more people than actual coding is.
However, you don't hear about them very often, and a lot of open source projects are set up in such a way that it's more difficult to get involved if you don't have the ability to read code. For example, on a commercial software product you can have an army of testers banging away at software even when it's in development, because you have human-readable specifications that you test against. I've yet to see any specifications on an OSS project, and many programmers think they're a waste of time. The net result is that people who can't read code aren't worth a whole lot. (Which surprised me, coming from a commercial development where we probably have a 3:1 ratio of non-coding analysts and testers for every actual developer, without counting management or dead weight.)
So I think there are multiple levels to the problem. People need to be encouraged to help out projects and make them more useful, but projects also need to be designed from a perspective that's scalable and doesn't assume that everyone can check out the code from CVS and start doing useful stuff with it. Because most people just can't.
On the user's side, people need to get rid of the lingering attitude that "if I wanted to pay for software, I'd just use Windows." There's a happy medium between getting screwed through the nose for commercial software, and using somebody's work without compensating or helping them, and making it more likely that the project will die. In the latter case you're really killing the goose that's laying the golden eggs.
Anyway, it's a complex issue, and I didn't mean for my earlier post to oversimplify and put blame on OSS devs unfairly. However, in the places where OSS has become mainstream, it seems like the same issues and conflicts come up again and again between coders and non-coding users, and I think both sides have some responsibility for making it easier on the other.
This goes round and round slashdot nearly everytime HVAC comes up. Heat pumps are more than 100% electrically efficient. As you said, they're moving heat around rather than directly heating the air. This can work in either direction (warming or cooling a building) and be several times more efficient than straight electric heat as far as your power bill is concerned.
This is totally true; I didn't consider heat pumps.
Actually, I've never understood why they're not more popular in my area. Everyone seems to use gas furnaces, even though they cost a fortune to operate and it never really gets cold enough to make heat pumps prohibitive. (Northern Virginia) There seems to be a bit of Luddism to it...everyone always has used gas, therefore they always will. Maybe as energy gets more expensive, people will come around. If I was building a house here, that's the way I'd certainly go. (Currently I rent, so I'm stuck with their POS gas heater; most of the time I don't even turn it on and just use my computers.)
Anyway, I do think my main point regarding the placement of industries that create waste heat, as opposed to those that require heat as an input, still stands. If you're a datacenter, and have to "buy cool" from the power company via your A/C system, then you might as well go where it's cooler (and where, as you pointed out, you might even be able to reuse or sell your 'waste' heat). If you're an iron foundry or auto-body shop, where you "buy heat" in order to actually do your job, then it makes sense to go where it's warmer or at least temperate.
Well, if you have a Bluetooth enabled phone, and it's not crippled by a shoddy cellular carrier (*cough* Verizon *cough*), you can always shoot the video, store it, and then upload it to your computer via BT and go that way onto the net.
I assume that the next generation of WiFi-enabled phones will be able to upload photos and video that way, working basically like network cameras do now (they either upload files to an FTP server somewhere, or they present themselves as an FTP server so another process can download the files). That's probably a little more practical than Bluetooth.
I could definitely see a phone where video and photos were cached locally on the phone's memory when you were just wandering around outside WiFi'ed areas, but when you got into a hotspot, then the phone would batch-upload everything, say to your Flickr site or Google Album.
I think cellphone videos are actually a neater idea than cellphone still-photos, because videos are more communicative in the same way that a telephone is, plus you don't need high resolution or very good lensing or zoom capability. It would be pretty neat to be able to take short "hey, everybody"-type videos and send them around, or video-blog, without getting reamed by the cellular company in charges.
I have a phone now that can technically take videos (Motorola Razr TMobile/GSM), but it has some sort of bizarre software issue that keeps it from shooting more than a few seconds worth of video at a time. (My girlfriend who has a European version can take ~30s clips, so I don't think it's hardware.)
If you put the datacenter in a normally cool locale, then you'd notice an immediate energy savings without having to do much of anything. The equipment would probably keep the building warm (probably overly warm, if you didn't vent to the outside) and if you designed the cooling system to draw in air from the outside, the actual amount of ventilation required (in CFM) might decrease.
I don't know the formula, but I bet that somebody could probably tell you how many CFM of air at 70F you'd need to equal a 1 CFM flow at 40F. Probably depends on the heat capacity of the air, I assume.
In general, cooling is energetically 'cheaper' than heating -- a good air conditioner can "make" 3 or 4 units of cool using one unit of electricity (in other words, they can move several Joules of heat from one place to another using 1 Joule of energy), while even the most efficient heater can only approach 1 to 1. (Electric heaters are 100% efficient, for all practical purposes.) However, if whatever activity you're planning on doing inside the building produces lots of heat anyway that you need to get rid of (computers), rather than requiring heat as an input (painting), then it might make sense to place yourself in a cold location and take advantage of the "free cool."
I bet Canada would be a great place for datacenters, if you had the backbone capacity and redundancy. Right now, those are really the key features in placing big datacenters; they pop up like toadstools wherever you have a lot of fat pipes coming together. Take that out of the mix, though, and an urban or exurban environment is the last place you'd want to build one. The middle of nowhere is actually a far better place.
That's a sunk cost if you have to have the generators either way, as a backup to the mains power.
However the additional expense of running the generators continuously (which is probably in excess of their rated duty cycle, if they're "emergency" generators) versus just having them sitting there and exercising them weekly, ought to be factored into the cost.
I would bet that a datacenter probably uses as much electricity as a small skyscraper. Because of the high equipment densities, a 1-story datacenter filled with racks probably has HVAC requirements that are like a multi-story office building filled with cubes. IT doesn't have the same lighting requirements, but that's not nearly the draw that heating and cooling are.
Actually, I bet that in many situations, if you just pulled the plug on a 'center, very bad things might happen to the equipment, aside from the obvious ones like data loss. Even without the machines actually producing any more heat (because they're off), without cooling air being forced through them, in very high density racks I wonder if the residual heat might not build up to rather high temperatures and become a problem; damaging hard drives or other temperature-sensitive parts.
Do you know what the actual output of the generators used in modern diesel-electric locomotives is?
I wonder what the voltage/frequency is. I assume that the motors are all AC, but if the generators are designed to produce low-frequency AC, then you're going to have issues using that in a standard datacenter, unless you are one of those rare places that does DC distribution (where you'd just run it thought a rectifier and the incoming frequency would be irrelevant, provided your rectifier could cope). The cost of a rotary converter to go from whatever the loco produces up or down to 60Hz could be just as much as the generator itself. (And rotary converters are, at least as I understand them, nothing but motors permanently coupled to generators, with some feedback circuits to keep the output in phase.)
I've heard that many traction power systems for conventional (non-diesel) electric railways use low-frequency AC, at around 25Hz, because this has advantages in terms of the motor designs. If this is such a great advantage that people are willing to create entirely parallel power systems for electric trains in order to use it, I can't imagine that if you were designing a diesel-electric loco, where the power was only going a few feet, that you wouldn't take advantage of it.
Don't most modern BIOSes also have an option to disable booting from devices other than the hard drive, or to specify the boot order, so you can't override the HD by using a floppy?
I know my HP xw5000 does. I've never used it, but I saw it there in the settings. You could tell it to only boot up from a particular IDE device, even if there were other valid options (say, a boot floppy).
I guess if you set the boot order to HD first, someone could still get into the case, unplug the HD, then turn it on and boot from a CD, then maybe plug the HD back in hot, and delete stuff... but again, if you don't notice that going on in a public lab, you've got problems. Like, why aren't you locking the chassis shut, for starters.
If you're using a BIOS that has a well-known master password, then it's not going to help you at all, but if the BIOS is good and you can't get inside the chassis to reset the CMOS or pull the battery or do some other kind of shenanigans, seems like that would keep people from messing around with it too much.
I wish there were a better way to allow people to safely boot from USB devices, without giving them access to the hard drive (so that a USB stick couldn't be used to mess with the native OS installation). Then you wouldn't have to stop legitimate users who want to boot from a USB stick and use their 'virtual computer' type system (which I think is a really neat idea). Unfortunately, because people love to use external boot devices as a way to fuck up shared computers, there's really no good way to allow this (unless you have everything netbooting from a read-only volume).
"Improve the source" not an option for most.
on
A First Look At Gaim 2.0
·
· Score: 1, Insightful
You don't like it, improve the source.
As long as we're getting things out of our system, I'd like to point out how completely unproductive this sentiment is.
First, most people aren't programmers, and even of people who know something about programming, fewer still have the skills required to make any meaningful modification to an open-source program.
Second, even if a person does know how to program, and is familiar with the project's language / graphical environment / architectural style, except in particular instances where there are no good alternatives available, or where the changes are trivial, it's almost always easier (assuming you value your time at all) to just use an alternative -- even a commercial one -- then spend several days or weeks reading somebody else's code in order to change it. Really, the only reason to work on an OSS project is if it's so specialized that it's the only thing going, or you enjoy working on it and are willing to take it on as a project despite it costing more of your time than purchasing an alternative would.
Any time you tell someone to "go fix it themselves," you might as well tell them to go buy the proprietary alternative, because that's the end result anyway.
I love the concept and philosophy of open-source software. But this geek ideal that everyone can just modify the hell out of their own system is false. It's like saying that because your car is made out of steel, if you don't like the design of the Ford Focus this year, you should go to a metal foundry and a machine shop and learn how to design and fabricate car parts so you can turn it into a sports car. Sure, a small number of people can probably do that, but a regular person isn't going to; they're just going to buy a BMW.
Gadu-Gadu is financed by the display of advertisements. As with ICQ, users are identified by their serial numbers. There are numerous add-ons available to provide extra features. The official version provides over 150 smiley icons, and allows off-line messages, data dispatch, and VoIP. From version 6.0, an experimental SSL secure connection mode can be used. The manufacturer is based in Warsaw.
Gadu-Gadu uses its own proprietary protocol.
Gadu-Gadu is the most popular IM in Poland. There are over 5 million registered accounts, and every day approximately 2.5 million users are online.
Many users consider the latest version too overloaded by unnessesary addons (Gadu-Gadu Radio Station etc.). So the older versions (especially 6.1 build 158) are still more popular than the new one.
It's Yet Another Proprietary IM Protocol, albeit one with basically insignificant market share outside of Poland, apparently.
Somebody needs to mod the parent up; I can't believe it's still down at +2.
The article is completely different if it's "FCC Commissioner" versus "FCC Chair." The guy making the remarks quoted isn't the head honcho, he's a minority member.
If the FCC Chairman had actually come out and said stuff like this, it would be holy crap, stop the presses, who are you and what did you do with the turd pile I used to call my government time.
An FCC Commissioner saying it, is still impressive, but it's an order of magnitude less so.
I for one would much sooner welcome some old guy sitting on his porch by the river bank than any number of wireless water-sensing overlords.
The best part is that the NOAA has a "sensor net" for that type of 'remote data sensor' already. It's called "SKYWARN" (beware, there is some sort of hideous applet or something on their page, it got my machine's HD thrashing for half a minute while FF froze) and it provides some really good coverage of stuff that might not get picked up by mechanical sensors. It wouldn't be cost effective, for example, to put a sensor network that measures the size of hail, all over the Midwest. Yet hail sizes can be an important part of tornado predictions.
Also, some spotter organizations are affiliated with Ham radio clubs and can operate entirely without infrastructure, meaning that you can retain some remote-sensing capability even in the midst of a weather-related disaster: exactly when you need it most.
Sometimes the "Mark 1 eyeball" and its accessories really are the best tools for the job. I suspect, knowing the government, that SKYWARN doesn't get a hundredth of the funding that various pet fancy-gadget-du-jour projects do.
Why our government, of course; the world's biggest insurance company, and the only one dumb enough to underwrite such a policy.
No sane insurance company would write half the policies that the National Flood Insurance Program does, because they know better. They can't just depend on a steady stream of money from nowhere to keep them afloat financially, at the same time that their insureds may be literally; companies in the real world have to at least break even over the long term.
Basically, the NFIP is a giant subsidy, paid by people living in non-flood-prone areas in order to allow other people to live in areas where they really shouldn't be. The clincher is that although defenders of the program always love to wave around the spectre of some poor family being ruined by flooding, most of those people don't know about and don't get the program in the first place. It's mostly people who have money -- and probably could afford to buy private flood insurance on the open market, if it existed (or who would just live somewhere else) -- who have policies and benefit from the program. So not only is it a stupid subsidy, it's a regressive tax on top of it.
I've seen some automated warehouse and inventory-management systems that depend on RFID tags, and (if you're into this kind of stuff) they're the slickest thing you've ever seen. If your full supply chain uses tags, then there's no manual inventorying; as stuff gets unloaded from the trucks at a loading dock (by the pallet-full -- scanners can 'talk' to tens or hundreds of tags at once), it gets noted. When it gets put on a shelf, it gets noted. When an order comes in, the system knows whether it's in stock, and where's it's located. The picker (guys who pull individual items from warehouse shelves) can follow a wrist-mounted computer right to the location, and scan it as they pick it up. As orders get loaded on a truck to go out, they get scanned again at the dock doors. At every step in your supply chain, you can do this.
It's not quite a fully-automated warehouse, but it's pretty close. If you've ever worked in industry or retail, you can appreciate the beauty of such a system. All that real-time data; I won't say there's "no limit" to what you can do, because I don't want to start sounding like an ad, but there's a lot.
So really, don't blame the technology here. The gear is really good. The problem is that a lot of contractors, who want to make a few bucks from Uncle Sam, have convinced some govvies that this sort of data flow -- which is great when you're talking about cases of Rice Krispies or DVD players -- would be nice to have on all of us. The problem with "RFID" as people have come to think of it, is totally a social one. If you could somehow 'uninvent' RFID, put the genie back in the bottle, it wouldn't fix the real issue: that our government is currently obsessed with reaching down into the personal lives of individual citizens, either by accident or by design. A government which took more of an interest in privacy concerns, probably wouldn't think that embedding RFID tags in passports and drivers licenses would be a good idea. That they do, is indicative of a problem in government, not in the tags.
An apt analogy would be Hollerith card sorters and other indexing machines, in the early part of last century. They let people do all sorts of rapid data analysis and were indispensable to industry and government for countless projects. Yet they were also used by the Nazis, to greater or lesser effect depending on who you choose to believe. That a particular technology was used reprehensibly isn't necessarily a valid criticism of the technology itself; virtually anything can be perverted for ill uses.
So in short, don't blame RFID in general. It's a great technology, when used correctly, and its potential for abuse isn't any greater than similarly revolutionary systems were in their day.
Either the Thug Classico, which is a Pint Glass with ice almost to the top, Absolut Mandarin poured in just over half way, then Red Bull to top the glass off (About 2/3 of a can.)
Funny, when I was in the Army we used to drink the same thing, only with unflavored vodka.
One of the reasons people buy Cisco gear is because they trust the company.
Sounds like a really good argument why you should never just blindly trust someone because of a brand name.
If you don't know who's code is actually running on your firewall/router/whatever, and I don't mean "what code is running on that model device, according to the manual," I mean your firewall, that actual metal box in the closet, then you are assuming a certain amount of risk. Any time you blindly swallow what some company that you bought something from tells you, remember that they have a financial motive to make you believe that their farts smell like roses. Some may be more blatant than others, but their goals are not the same as yours, even if they do coincide in certain areas.
By the time you get your hands on a piece of hardware, it's passed through dozens (if not hundreds) of carriers, middlemen, distributors, wholesalers, and the like. You are trusting every one of them to not have messed with it, in ways ranging from an actively hostile backdoor, to petty thievery like the RAM theft that someone discusses further up in the thread. There are some pretty good arguments for using the simplest hardware possible and then loading software yourself. It's still not totally devoid of risk (and with software you get into the whole thing about compiler compromises), but it limits the number of hands the code passes through.
The amount of trust that people put blindly in others is simply astounding. Sometimes it's for good reason, but other times it boils down to calculated laziness. Maybe that calculation needs to be revised a little.
Slashdot Mirror
In this case, having one bad guy with the directions really isn't any better than 10 bad guys with it.
In fact, the more bad guys that have it, the more likely the problem will get fixed, thus it's actually better that the most 'bad guys' possible get it. If only one person knows how to rig the election, chances are higher they'll be able to get away with it. If 100 people know and all try to rig the election, chances are none of them will get away with it, because the tampering will be too obvious.
Frankly I think the best thing to happen would be for someone to utterly steal the next election; make "Mickey Mouse" or "Elvis" get 100 seats in Congress or something. The cost of having to repeat a single election is certainly much smaller than continuing for decades with a flawed process, where nobody can tell whether the vote is being rigged or not.
...how great inventions discovered while making pornography are now carried on space missions?
Wait...it's not for making pornography?
I, for one, welcome our ... damn.
Interesting.
Well, given the price in gas (last winter we were paying upwards of $400 a month in gas alone to heat a single-story, seven-room manufactured house), if I was building, I'd be willing to get a heat pump and use electric oil-filled radiators as a backup if it got so cold that it stopped working completely. (And save the money for the air conditioning in the summertime, when this place turns into a humid suburb of Hell. The fact that I'm not originally from here is probably starting to show...)
As it is, my housemates and I have decided we're not going to use the gas furnace for anything more than the minimum required to keep the pipes from freezing this season. It's old and inefficient enough that by the time the heat gets from the furnace to the living areas of the house, it costs more than just putting a $40 oil-filled electric radiator there does. Plus, you can put the electric rads on timers (some have built-in ones) and control the temperature in different rooms over time. Sure, you can do that with a timed thermostat on your central heat, but the control you get via individual radiators is far more fine-grained. Add an electric blanket on the bed and a ceramic fan-forced heater in the bathroom, and you really don't need to keep the entire house cooking at 72F. (It probably helps that I dislike an ambient temperature greater than about 65F anyway, summer or winter.)
I wonder if as energy prices continue to increase, they'll be more interest in heat pump systems that use heat resovoirs other than the atmosphere; for example the ground below the frost line, or above-freezing water at the bottom of a lake or pond, and additional types of heat pumps designed to function at very low temperatures. Maybe a solution would be to not try and have the same unit do winter heating and summer cooling? It seems like the dual-functionality probably requires some tradeoffs; if you used two separate units, one that was designed for cold-weather operation, pulling heat in, and another that was designed for hot-weather operation, exhausting heat out, maybe they would be more practical in northern climates.
Does anyone know why CACert's root isn't included in Firefox?
Seems like that would be a no-brainer; I can't believe Firefox is really interested in perpetuating the Verisign monopoly. (Or is Verisign a donor?)
Ignoring someone who's wrong, when arguing with them would only make them more powerful, is both good and correct.
However, ignoring someone who has a gun to your head, and is asking you for a good reason why they shouldn't pull the trigger, seems rather shortsighted.
The videogames industry is, right now, in the second position. Maybe Congress doesn't have the gun to their head yet, but they're fiddling around trying to take the safety off and figure out which end to hold.
Now is not a good time to just ignore the government and hope it'll go away. That sort of attitude earned us the DMCA; wouldn't it have been nice if the EFF had been around back in the 1980s, when its the precursor laws (in particular, the anti-decryption laws regarding satellite TV) were passed?
Unfortunately, when you stay home and ignore what's going on, it doesn't keep what happens in the absence of your attention from affecting you later.
Basically, what I'm trying to say is that the Gaming Industry could show up to an event like this and have God as a witness and no one there will listen to them when they say videogames do not cause children to perform violent acts.
That pretty much sums the whole thing up in one line.
Actually, you could show up in any Congressional subcommittee with God in tow, and unless God happened to be made out of money, I doubt you'd influence any pending piece of legislation.
If the "games lobby" wants to make its voice heard in government, and keep itself from being run over as the Fox News scarecrow-du-jour, then they should take a very good look at what the National Rifle Association does, in terms of communicating with and mobilizing its support base, getting donations, and funneling those donations to where they'll have maximum political impact. I can't think of any organization that is as frankly successful and powerful as they are, and has continuously maintained such a high profile, and has done it while staying within the bounds of the law. (Some corporate lobbies might come close, but I think their cash burn rates are much higher for the effect they achieve.)
You can have logical arguments so beautiful they'd make Plato sit down and weep, enough scientific evidence to unequivocally prove a dozen theories of everything, but the government will still ignore you if you are not either a large force among voters, or have lots of filthy lucre to burn. Preferably, have both.
Ok well then he should donate money to the gaim developers or sponsor them in some other way. There are other ways to help out.
Definitely! This is one reason that I dislike the "code it yourself" response, because I think it turns people off and makes them believe that if you don't read and write and breathe C, you'll never have any impact or value in OSS development. There are lots of ways to help out, including straightforward financial donations, which are open to many more people than actual coding is.
However, you don't hear about them very often, and a lot of open source projects are set up in such a way that it's more difficult to get involved if you don't have the ability to read code. For example, on a commercial software product you can have an army of testers banging away at software even when it's in development, because you have human-readable specifications that you test against. I've yet to see any specifications on an OSS project, and many programmers think they're a waste of time. The net result is that people who can't read code aren't worth a whole lot. (Which surprised me, coming from a commercial development where we probably have a 3:1 ratio of non-coding analysts and testers for every actual developer, without counting management or dead weight.)
So I think there are multiple levels to the problem. People need to be encouraged to help out projects and make them more useful, but projects also need to be designed from a perspective that's scalable and doesn't assume that everyone can check out the code from CVS and start doing useful stuff with it. Because most people just can't.
On the user's side, people need to get rid of the lingering attitude that "if I wanted to pay for software, I'd just use Windows." There's a happy medium between getting screwed through the nose for commercial software, and using somebody's work without compensating or helping them, and making it more likely that the project will die. In the latter case you're really killing the goose that's laying the golden eggs.
Anyway, it's a complex issue, and I didn't mean for my earlier post to oversimplify and put blame on OSS devs unfairly. However, in the places where OSS has become mainstream, it seems like the same issues and conflicts come up again and again between coders and non-coding users, and I think both sides have some responsibility for making it easier on the other.
This goes round and round slashdot nearly everytime HVAC comes up. Heat pumps are more than 100% electrically efficient. As you said, they're moving heat around rather than directly heating the air. This can work in either direction (warming or cooling a building) and be several times more efficient than straight electric heat as far as your power bill is concerned.
This is totally true; I didn't consider heat pumps.
Actually, I've never understood why they're not more popular in my area. Everyone seems to use gas furnaces, even though they cost a fortune to operate and it never really gets cold enough to make heat pumps prohibitive. (Northern Virginia) There seems to be a bit of Luddism to it...everyone always has used gas, therefore they always will. Maybe as energy gets more expensive, people will come around. If I was building a house here, that's the way I'd certainly go. (Currently I rent, so I'm stuck with their POS gas heater; most of the time I don't even turn it on and just use my computers.)
Anyway, I do think my main point regarding the placement of industries that create waste heat, as opposed to those that require heat as an input, still stands. If you're a datacenter, and have to "buy cool" from the power company via your A/C system, then you might as well go where it's cooler (and where, as you pointed out, you might even be able to reuse or sell your 'waste' heat). If you're an iron foundry or auto-body shop, where you "buy heat" in order to actually do your job, then it makes sense to go where it's warmer or at least temperate.
Well, if you have a Bluetooth enabled phone, and it's not crippled by a shoddy cellular carrier (*cough* Verizon *cough*), you can always shoot the video, store it, and then upload it to your computer via BT and go that way onto the net.
I assume that the next generation of WiFi-enabled phones will be able to upload photos and video that way, working basically like network cameras do now (they either upload files to an FTP server somewhere, or they present themselves as an FTP server so another process can download the files). That's probably a little more practical than Bluetooth.
I could definitely see a phone where video and photos were cached locally on the phone's memory when you were just wandering around outside WiFi'ed areas, but when you got into a hotspot, then the phone would batch-upload everything, say to your Flickr site or Google Album.
I think cellphone videos are actually a neater idea than cellphone still-photos, because videos are more communicative in the same way that a telephone is, plus you don't need high resolution or very good lensing or zoom capability. It would be pretty neat to be able to take short "hey, everybody"-type videos and send them around, or video-blog, without getting reamed by the cellular company in charges.
I have a phone now that can technically take videos (Motorola Razr TMobile/GSM), but it has some sort of bizarre software issue that keeps it from shooting more than a few seconds worth of video at a time. (My girlfriend who has a European version can take ~30s clips, so I don't think it's hardware.)
If you put the datacenter in a normally cool locale, then you'd notice an immediate energy savings without having to do much of anything. The equipment would probably keep the building warm (probably overly warm, if you didn't vent to the outside) and if you designed the cooling system to draw in air from the outside, the actual amount of ventilation required (in CFM) might decrease.
I don't know the formula, but I bet that somebody could probably tell you how many CFM of air at 70F you'd need to equal a 1 CFM flow at 40F. Probably depends on the heat capacity of the air, I assume.
In general, cooling is energetically 'cheaper' than heating -- a good air conditioner can "make" 3 or 4 units of cool using one unit of electricity (in other words, they can move several Joules of heat from one place to another using 1 Joule of energy), while even the most efficient heater can only approach 1 to 1. (Electric heaters are 100% efficient, for all practical purposes.) However, if whatever activity you're planning on doing inside the building produces lots of heat anyway that you need to get rid of (computers), rather than requiring heat as an input (painting), then it might make sense to place yourself in a cold location and take advantage of the "free cool."
I bet Canada would be a great place for datacenters, if you had the backbone capacity and redundancy. Right now, those are really the key features in placing big datacenters; they pop up like toadstools wherever you have a lot of fat pipes coming together. Take that out of the mix, though, and an urban or exurban environment is the last place you'd want to build one. The middle of nowhere is actually a far better place.
That's a sunk cost if you have to have the generators either way, as a backup to the mains power.
However the additional expense of running the generators continuously (which is probably in excess of their rated duty cycle, if they're "emergency" generators) versus just having them sitting there and exercising them weekly, ought to be factored into the cost.
I would bet that a datacenter probably uses as much electricity as a small skyscraper. Because of the high equipment densities, a 1-story datacenter filled with racks probably has HVAC requirements that are like a multi-story office building filled with cubes. IT doesn't have the same lighting requirements, but that's not nearly the draw that heating and cooling are.
Actually, I bet that in many situations, if you just pulled the plug on a 'center, very bad things might happen to the equipment, aside from the obvious ones like data loss. Even without the machines actually producing any more heat (because they're off), without cooling air being forced through them, in very high density racks I wonder if the residual heat might not build up to rather high temperatures and become a problem; damaging hard drives or other temperature-sensitive parts.
Do you know what the actual output of the generators used in modern diesel-electric locomotives is?
r k
:)
I wonder what the voltage/frequency is. I assume that the motors are all AC, but if the generators are designed to produce low-frequency AC, then you're going to have issues using that in a standard datacenter, unless you are one of those rare places that does DC distribution (where you'd just run it thought a rectifier and the incoming frequency would be irrelevant, provided your rectifier could cope). The cost of a rotary converter to go from whatever the loco produces up or down to 60Hz could be just as much as the generator itself. (And rotary converters are, at least as I understand them, nothing but motors permanently coupled to generators, with some feedback circuits to keep the output in phase.)
I've heard that many traction power systems for conventional (non-diesel) electric railways use low-frequency AC, at around 25Hz, because this has advantages in terms of the motor designs. If this is such a great advantage that people are willing to create entirely parallel power systems for electric trains in order to use it, I can't imagine that if you were designing a diesel-electric loco, where the power was only going a few feet, that you wouldn't take advantage of it.
See:
http://en.wikipedia.org/wiki/Traction_power_netwo
http://en.wikipedia.org/wiki/Utility_frequency
But you're right, if you could park an old diesel loco outside your datacenter, that would be infinite nerd points.
You forgot to mention Web 2.0, Ruby on Rails, the blogosphere, and whether it will run Linux.
Don't most modern BIOSes also have an option to disable booting from devices other than the hard drive, or to specify the boot order, so you can't override the HD by using a floppy?
... but again, if you don't notice that going on in a public lab, you've got problems. Like, why aren't you locking the chassis shut, for starters.
I know my HP xw5000 does. I've never used it, but I saw it there in the settings. You could tell it to only boot up from a particular IDE device, even if there were other valid options (say, a boot floppy).
I guess if you set the boot order to HD first, someone could still get into the case, unplug the HD, then turn it on and boot from a CD, then maybe plug the HD back in hot, and delete stuff
If you're using a BIOS that has a well-known master password, then it's not going to help you at all, but if the BIOS is good and you can't get inside the chassis to reset the CMOS or pull the battery or do some other kind of shenanigans, seems like that would keep people from messing around with it too much.
I wish there were a better way to allow people to safely boot from USB devices, without giving them access to the hard drive (so that a USB stick couldn't be used to mess with the native OS installation). Then you wouldn't have to stop legitimate users who want to boot from a USB stick and use their 'virtual computer' type system (which I think is a really neat idea). Unfortunately, because people love to use external boot devices as a way to fuck up shared computers, there's really no good way to allow this (unless you have everything netbooting from a read-only volume).
You don't like it, improve the source.
As long as we're getting things out of our system, I'd like to point out how completely unproductive this sentiment is.
First, most people aren't programmers, and even of people who know something about programming, fewer still have the skills required to make any meaningful modification to an open-source program.
Second, even if a person does know how to program, and is familiar with the project's language / graphical environment / architectural style, except in particular instances where there are no good alternatives available, or where the changes are trivial, it's almost always easier (assuming you value your time at all) to just use an alternative -- even a commercial one -- then spend several days or weeks reading somebody else's code in order to change it. Really, the only reason to work on an OSS project is if it's so specialized that it's the only thing going, or you enjoy working on it and are willing to take it on as a project despite it costing more of your time than purchasing an alternative would.
Any time you tell someone to "go fix it themselves," you might as well tell them to go buy the proprietary alternative, because that's the end result anyway.
I love the concept and philosophy of open-source software. But this geek ideal that everyone can just modify the hell out of their own system is false. It's like saying that because your car is made out of steel, if you don't like the design of the Ford Focus this year, you should go to a metal foundry and a machine shop and learn how to design and fabricate car parts so you can turn it into a sports car. Sure, a small number of people can probably do that, but a regular person isn't going to; they're just going to buy a BMW.
Somebody needs to mod the parent up; I can't believe it's still down at +2.
The article is completely different if it's "FCC Commissioner" versus "FCC Chair." The guy making the remarks quoted isn't the head honcho, he's a minority member.
If the FCC Chairman had actually come out and said stuff like this, it would be holy crap, stop the presses, who are you and what did you do with the turd pile I used to call my government time.
An FCC Commissioner saying it, is still impressive, but it's an order of magnitude less so.
I for one would much sooner welcome some old guy sitting on his porch by the river bank than any number of wireless water-sensing overlords.
The best part is that the NOAA has a "sensor net" for that type of 'remote data sensor' already. It's called "SKYWARN" (beware, there is some sort of hideous applet or something on their page, it got my machine's HD thrashing for half a minute while FF froze) and it provides some really good coverage of stuff that might not get picked up by mechanical sensors. It wouldn't be cost effective, for example, to put a sensor network that measures the size of hail, all over the Midwest. Yet hail sizes can be an important part of tornado predictions.
Also, some spotter organizations are affiliated with Ham radio clubs and can operate entirely without infrastructure, meaning that you can retain some remote-sensing capability even in the midst of a weather-related disaster: exactly when you need it most.
Sometimes the "Mark 1 eyeball" and its accessories really are the best tools for the job. I suspect, knowing the government, that SKYWARN doesn't get a hundredth of the funding that various pet fancy-gadget-du-jour projects do.
/me points in the direction of Capitol Hill.
Why our government, of course; the world's biggest insurance company, and the only one dumb enough to underwrite such a policy.
No sane insurance company would write half the policies that the National Flood Insurance Program does, because they know better. They can't just depend on a steady stream of money from nowhere to keep them afloat financially, at the same time that their insureds may be literally; companies in the real world have to at least break even over the long term.
Basically, the NFIP is a giant subsidy, paid by people living in non-flood-prone areas in order to allow other people to live in areas where they really shouldn't be. The clincher is that although defenders of the program always love to wave around the spectre of some poor family being ruined by flooding, most of those people don't know about and don't get the program in the first place. It's mostly people who have money -- and probably could afford to buy private flood insurance on the open market, if it existed (or who would just live somewhere else) -- who have policies and benefit from the program. So not only is it a stupid subsidy, it's a regressive tax on top of it.
RFID is a great technology in its place.
I've seen some automated warehouse and inventory-management systems that depend on RFID tags, and (if you're into this kind of stuff) they're the slickest thing you've ever seen. If your full supply chain uses tags, then there's no manual inventorying; as stuff gets unloaded from the trucks at a loading dock (by the pallet-full -- scanners can 'talk' to tens or hundreds of tags at once), it gets noted. When it gets put on a shelf, it gets noted. When an order comes in, the system knows whether it's in stock, and where's it's located. The picker (guys who pull individual items from warehouse shelves) can follow a wrist-mounted computer right to the location, and scan it as they pick it up. As orders get loaded on a truck to go out, they get scanned again at the dock doors. At every step in your supply chain, you can do this.
It's not quite a fully-automated warehouse, but it's pretty close. If you've ever worked in industry or retail, you can appreciate the beauty of such a system. All that real-time data; I won't say there's "no limit" to what you can do, because I don't want to start sounding like an ad, but there's a lot.
So really, don't blame the technology here. The gear is really good. The problem is that a lot of contractors, who want to make a few bucks from Uncle Sam, have convinced some govvies that this sort of data flow -- which is great when you're talking about cases of Rice Krispies or DVD players -- would be nice to have on all of us. The problem with "RFID" as people have come to think of it, is totally a social one. If you could somehow 'uninvent' RFID, put the genie back in the bottle, it wouldn't fix the real issue: that our government is currently obsessed with reaching down into the personal lives of individual citizens, either by accident or by design. A government which took more of an interest in privacy concerns, probably wouldn't think that embedding RFID tags in passports and drivers licenses would be a good idea. That they do, is indicative of a problem in government, not in the tags.
An apt analogy would be Hollerith card sorters and other indexing machines, in the early part of last century. They let people do all sorts of rapid data analysis and were indispensable to industry and government for countless projects. Yet they were also used by the Nazis, to greater or lesser effect depending on who you choose to believe. That a particular technology was used reprehensibly isn't necessarily a valid criticism of the technology itself; virtually anything can be perverted for ill uses.
So in short, don't blame RFID in general. It's a great technology, when used correctly, and its potential for abuse isn't any greater than similarly revolutionary systems were in their day.
Either the Thug Classico, which is a Pint Glass with ice almost to the top, Absolut Mandarin poured in just over half way, then Red Bull to top the glass off (About 2/3 of a can.)
Funny, when I was in the Army we used to drink the same thing, only with unflavored vodka.
And without the Red Bull.
Or the ice.
One of the reasons people buy Cisco gear is because they trust the company.
Sounds like a really good argument why you should never just blindly trust someone because of a brand name.
If you don't know who's code is actually running on your firewall/router/whatever, and I don't mean "what code is running on that model device, according to the manual," I mean your firewall, that actual metal box in the closet, then you are assuming a certain amount of risk. Any time you blindly swallow what some company that you bought something from tells you, remember that they have a financial motive to make you believe that their farts smell like roses. Some may be more blatant than others, but their goals are not the same as yours, even if they do coincide in certain areas.
By the time you get your hands on a piece of hardware, it's passed through dozens (if not hundreds) of carriers, middlemen, distributors, wholesalers, and the like. You are trusting every one of them to not have messed with it, in ways ranging from an actively hostile backdoor, to petty thievery like the RAM theft that someone discusses further up in the thread. There are some pretty good arguments for using the simplest hardware possible and then loading software yourself. It's still not totally devoid of risk (and with software you get into the whole thing about compiler compromises), but it limits the number of hands the code passes through.
The amount of trust that people put blindly in others is simply astounding. Sometimes it's for good reason, but other times it boils down to calculated laziness. Maybe that calculation needs to be revised a little.