Those are the early adoptors of nearly all technology, and drive the prices to points where normal financially sane people can afford them. Half of all technology every invented was driving by one of those three groups.
For me personally, I use my desktop for lots, and lots of compiling. I'd like my desktop to be more responsive under heavy I/O load. I'd like it to do more things in the background. Personally, I've probably got enough CPU for my desktop. I found find a use for more, but I sure could use lots more disk I/O. I could use more memory also.
I'd like my desktop interface to be much more sophisticated, and to have a lot more context sensitive abilities. All that takes more CPU horsepower. The eye candy is nice, but that's not that interesting to me.
I'd like a desktop that supports 10 monitors. I'd like even more of a network transparent desktop then I have using X11. I'd like to have faster then GigaBit, which requires more CPU, more bus, and faster disks.
I'd like to be able to do lots, and lots of streaming realtime encoded TV signal for my home built PVR (encode 3-4, while playing one that has been paused thus having 5-10 minutes of latency in the playback). I'd like to see more ability to have low latency for specific tasks (music, TV encoding, multi-medial in general).
It's really expensive to build just a handful of CPU's. It's much cheaper to make them in huge volumes. Thus Intel and AMD makes them in huge volumes, and they sell pretty well to the consumer, and Intel makes a killing selling the CPU's, the chipsets, and the various electronic devices they either sell or license to others. So literally the thing driving more, faster CPU's is economics. You just can't buy the old CPU's any more. When you want a new CPU to replace a 5 year old computer, you have to buy a new one.
I'd like more ability to parse the natural english language. Thus facilitating better ability to write a content crawler that can point out urgent e-mails, while crawling the google cache finding content that I'd find amusing.
I'd like the ability to do more and better datamining. To correlate more information about my machines and servers. More ability to automate more and more task to with computer.
I could use more CPU power to use to secure my machines. If I could log every incoming packet and analyize more and more correlations, and deduce when people are doing bad things that'd be great. If I could use it to do more analysis of software and code. To develop more, and more sophisticated techniques of telling when software does something stupid.
I'd like it because it means, the construction of larger, and larger software applications is possible.
A. You are being irrational and short sighted (both for the nurse and surgeon). Basing your judgement on an for entire classes of medical fields on a single incident is completely and utterly irrational. Change that from nurse to "White Doctor", and from surgeon to "Black Doctor", and re-read your story. Everbody and their brother would call you a racists.
To make that sort of judgement you'd make, you'd have to do a comprehensive study of surgeons and nurses, and the outcome of their recommendations. I'm willing to bet dollars to donuts, that if you do that, the surgeon's (as an entire medical class) recommendations have a far better survial rate then a nurse's (as an entire medical class) recommendations. I've known a lot of nurses (I used to work in on the IT equipment on hospitals). Nurses make mistakes too. I've known nurses who can't calculate how many drops per minute somebody should get in an hour, because they can't do simple algebra problems.
However, in your case, it could be that the doctor is incompetant and should have his license revoked.
B. You have no idea that she recovered because of the medication. The human body, especially near death is very strange. People have recovered from being dead 45 minutes after being put in a freezer. There is no control to state that she lived soley (at least not in your story).
C. When I said nurse, I meant run of the mill nurse at my personal doctors office, not an ICU nurse. ICU nurses have lots of knowledge I would take seriously if they gave me advice. Generally, any nurse at a family practice, will only give you advice that will make you feel better, but you'll get better without the advice. About the only medical advice you'll ever need is: "Take it easy, drink lots, and lots of fluids". For 95% of everything that is ever wrong with you, that's all you have to do.
At diagnosing the common cold, I'll bet a nurse is, at reading the MRI, and figuring out what is wrong with the internals of my brain, I'll bet a surgeon is much better. For a surgeon, I want a god complex, really steady hands, and lots, and lots of history of doing the surgery.
In my experience, anything a nurse (or General Practitioner) tells you can be ignored, you'll still get better. A nurse generally gives advice that makes you feel better more comfortable, and possible speeds the process along a bit, but inheirently does nothing to fundamentally change the outcome. Generally I stopped seeing a GP unless I need a bone set, or I have been sick for a week.
However, when you have a bleeding brain, nothing but a brain surgeon will do. When you have a pile of bad C code, a really good programmer, or an average programmer will both get the job done (in differing amounts of time). So there isn't as much selection pressue on the job of a programmer.
If I found a brain surgeon who was nice, I wouldn't let them operate on me. Clearly they aren't a real brain surgeon if they are a decent human being:-)
Finally, if you had quoted the following sentence, I pointed out that, comparing programmers to programmers is just as fair as a nurse to a surgeon. If you made a nurse do a surgeons job, there'd be an order of magnitude difference, if you made a surgen do a nurses job, there'd be an order of magnitude differece in quality. If you took a programmer whose really good a job X and make they do job Y, it's not terrible shocking there is fall of. A lot of programmers take work, and do work in areas they lack experience or knowledge, because it is a good job, and the people doing the hiring can't tell the difference.
Skill as a programmer, because programmers have a very, very broad range of skills and abilities that they need to do to accomplish their job, are inheriently incomparable in most ways. Finally, a lot of great programmers are great on the codebases they work on, but they'd be lousy on other codebases.
Sure there are. Take a look at pro-football. There are orders of magnitude difference inbetween the guys on the scout team, and the guys on the field on Sundays.
There is an order of magnitude difference in skill between a nurse, and a brain surgeon. No it's not the same job, but then, coding the Linux TCP stack has very little to do with being a VB programmer developing a new front end for existing applications at a major insurance company.
There is probably an order of magnitude difference between a new carpenter, versus a experienced carpenter who was trained by another very experienced carpenter.
It's only in the computer field where differentiating between the crappy, and the great, and the incredible demand for the skills lets those with so little relavent talent continue to earn a good wage doing so they have no incentive to find another line of work.
I know I'm not kernel coder quality, but I'm a damn fine programmer at C++. So I'm somewhere on the scale, but not at the very high end.
Wait a second... You can't afford a $130 distro, but you can afford a dual opteron based computer? Hell, you probably dropped $1.5-2K on the server, and didn't consider what you'd do for an OS.
I thought that Fedora had a development version of a 64bit version for the AMD processors. I could be wrong.
According to that link, Alan has a BSc in CS. Linus Torvalds has a Bachealors degree in CS, and an honorary Ph.d from the same school in Finland. I'm too lazy to dig up links for that. It's in several of the books about his life.
I'm not sure I understand the extent of your problem. I've rewire my home (built more recently then 1964). In the end, I just went to the local graybar, and pickeup a bunch of punch down blocks. I setup punch down block "A" that had 3 lines duplicated 8 times (wiring the second column to the first column in the row directly below it for 2-8). I then used metal clips to connect 2-3 columns on bunch down block "A". So now I could take wires from the 4th column on block "A", over to the first column on block "B".
On block "B", in the forth row, I punched down every jack in the house. I labelled every pair, and had a map of where they were in the house.
You then run patch cables from column 4 block "A", to column 1 block "B", and use the metal clips to complete the circut. Thus any time I wanted to change out wires, all it took was a small run of wire from block A to block B. The original wires from the phone company or to the jack, never had to be touched to manipulate the system. You want to get the wires from the phone company, and the wires from the jack so you, never ever touch them. It makes trouble shooting lots easier, and you'll never run out of cable.
I might have introduced some line noise due to the way I duplicated the lines on block A, but I don't use dialup, so that was never an issue.
Now, if you want to get really fancy, wire each cat 5 cable into a patch panel. Thus you can put phone, or ethernet to any jack in your home. Now you can skip block "B" (the patch panel acts like the punch down block). Now, use row 4 of punchdown block A, and take phone jacks straight to the patch panel. You'll have eight copies of each line to distribute around the house as you see fit. When you want to move a line, just change the port it's plugged into on the patch panel. Unless you do phone, and network, this is really overkill. Just running the punchdown blocks will work great.
I figure that if the phone company uses punch down blocks, they are good enough for me. The tools are kinda expensive ($50-80 for a punch down tool, and the heads you'll need to do it). The punchdown blocks themselves are dirt cheap, like $8 bucks a piece. The patch panels are pretty expensive.
Possible the word I was looking for was generation. However, I'd be surprised from the stories I hear from my Uncle's, and from my Dad, if things 20 year olds in the 1950's were truely any better then 20 year olds today.
Uhhh, have you ever watched a collegiate football game? Rampant might be an overstatement for the cheating, but there are things that are clearly rules violations. A lot of obvious penalties happen that aren't called because that is what the game has degenerated to. I believe trash talking was invented during a football game, if not it's been perfected by 22 year atheletes who are treated like gods by their peers.
Watch any game of pickup basketball by any group of 25 year olds. If there aren't a half dozen fouls every time up and down the court, they aren't really playing hard.
Now, I've never seen a pickup game result in a shooting, however, I'm doubtful it's never happened. I'd be very impressed if a gun has never been pulled after a college football game.
I think it has more to do with the nature of the attitude of people of that age, then it does with people who play computer games.
You said it yourself, when you used the expression "take the first row". Using the adjective first, implies that there are many of them. However, you are correct in saying, there is no need to store more then one. If "asdf" and "3146t" have the same crypt (for a given salt), then technically, both of them are your password. By the pigeon hole princeple (you can have more bits in the password, then there bits in the salt + crypt( passwd ) ), at least one hash will have more then one password that hashes to that, hence technically, the parent is correct, it isn't a 1 to 1 lookup.
I've always wondered by nobody did this before, but then realized just how much space it would take to store. In order to store that many keys to be able to brute force them all, is insane.
Kirby
Re:What does it take to get a 1?
on
Online! The Book
·
· Score: 1
I'd bet to get a one, you'd have to print out all the links to slashdot comments, that involved stupid jokes about Hot Grits, Goats.ex, "In Soviet Russia", and "imagine a cluster of those".
Notice, how you'd have to print URL's to it (so you have to type it in), not the actual comments. Someone might find the comments amusing for the first 5 pages. Enough to earn it a 2 by rounding error or something. So even most of all the worthless comments on Slashdot (some of which are mine), wouldn't even rate a 1. Forcing someone to type the links to get them, would rate a 1 however.
What flavor of s/key? Last I heard, s/key was rendered insecure (not sure why, but given one password, you could find the next one). Maybe I should say, a particular implementation of s/key.
So it might be that the random number generator wasn't so random on the one we used. I can't find a reference off hand. Pretty much everybody I can find says that using the MD4 is not as secure as MD5, but still secure as a hash.
I used one once, where you had to have a username, know the current password, plus you had to download this goofy little Java applet and run it locally, giving it a long passphrase that you had to remember permanently. It output your onetime password for you to use. I suppose someday I should take to using S/Key and generating a list of OTP. So if you were out onsite at a Hospital, that was the only way to log in remotely.
All authentication, should be done based on something you have, and something you know. So having a keypad that generate OTP's, or having some type of secure dongle. That, and an additional password should be needed.
I picked up one of those keyfob's from think geek, haven't written down a password in a long time. About my only gripe, is that it doesn't have enough password slots, and it doesn't generate long enough passwords.
Oddly, nearly all of this guys issues are about authentication, none of them are about authorization, or about other forms of security.
I'll point out, that it wasn't a well paying job. If you can't live in a decent area, it's not well paying. Period.
Yes, it sure might be a 6 figure job, but a 6 figure job in Boston, doesn't afford you the lifestyle that a $30K a year job does in rural NE (not where I live, but within an hour or two drive of here).
I laughed at the guy I know who didn't want to move back to Omaha, NE (where I live), from Boston after his.com collapsed. Because he'd have to take a 60% pay cut.
Neverminding the fact, that I lived off less (including my new truck payment, and my nice apartment payment), then he did for rent in his crappy high rise apartment in a crappy neighborhood. Rent out there, costs about 3 times what it does in Omaha, and my rent is about 1/3 of my monthly living expenses. You do the math. I easily lived pretty well off the 60% paycut he didn't want to take.
The only good part about living in a very expensive place, is two fold. Non-necessities are a lot cheaper relative to your salary (so a $1000 computer, would be 1/10th, instead of 1/3 of your monthly salary). Also, if you build equity there, and then move someplace else, you can live like a king. A $80K house in Omaha, would probably run $300K in Boston. You probably couldn't buy as big a lot as the $80K house was in the Boston area.
A lot of this, is a microcosm of why it's so much cheaper to run a business in Omaha, NE then it is in Boston, or SF. No it's not India cheap, but it's pretty damn cheap.
First off,
I have several questions. Do you have an original copy of the partition before you started running recovery tools (after you deleted the file, but before you created new ones)? If not, make an image immediately. You want the most original image you can find.
Now, the second question, is how much data am I a looking for, and how large is the partition? (How large is the needle, and how large is the haystack?). What type of data am I looking for? Is it a word document? A text file? A gif? A jpg? Some html? A PDF?
The smaller the file, the more likely that if it got overwritten, it all got overwritten. However, the more likely you are to recover all of it. If it was a very large file, it's possible that you can recover pieces and parts, but not all of it.
Now, it's my understanding that you can recover anything written to a harddrive, even if you have overwritten it several times. However, it's very, very expensive to do so. So now the question is how much money is it worth to you?
The guys as ReiserFS probably have the best shot at helping you. They probably don't want to however. The more you know about the order of the files in the directory, the more you know about how the files were constructed, and the order files got put on disk the better. They you can make better educated guesses about the sequence in which the pages got allocated to know where to go look for the file.
Do you have anything on the drive you are worried about posting? Can you post an image of the drive? I'm not an expert in this area, but I've seen people recover mail spools at an ISP using dd. People leave ISP's over losing all their mail, so they worked really hard at it (however, that was an ext2 filesystem).
Kirby
There is such a beast, or at least pretty close. It's I2O. According to what I've read, it's a drivers where the OS writes the high-level driver, for their specific OS, and the device maker writes the low level driver that provides functionality to the high level driver. The low level driver can be plugged into any OS. There is a specification for each major type of hardware.
It's a neat idea, but I'm not sure how popular it is with hardware makers, and it somewhat constrains the implementation in hardware. The basic underlying princepals of the hardware would have to support the way the high level model is written, as opposed to having the software conform to the software.
It has to be a split driver model, as OS's organize themselves differently, so what would be highly efficient in one would be dog slow in another. This is also why various people recommend not porting a Windows driver to Linux, but to instead write a native Linux driver. Somebody presented a paper on the 10 things not to do while writing a Linux driver.
Notice, that I2C is also how lots of Linux drivers are written for block devices, because lots of block devices have a high layer, a mid layer, and a low level. Normally the high level, and mid layer are similar between lots of drivers, and generally get squeezed into a single driver.
Depends on volume. Intel is famous for at various points in their history for producing the same chip, or making very minor adjustments to it, and selling it as 4 different chips. Because it's cheaper to run 1 line of 4 times the volume, then it is to run 4 production lines.
You only have to do Q/A once. Every Engineering tweak you figure out directly applies to all of the chips. You don't have to worry quite as much about not having enough production capacity to make type A chips, because type A has a lot more demand then was originally predicted.
It's my understanding that overclocking sites list which CPU runs of old Celeron's were just sold under a lower speed CPU and could safely be overclocked well outside of the range Intel labelled it as.
There's a reason you get a huge discount when you buy in volume, for similar reasons, it's much cheaper to produce in huge volumes. That's how Intel became the 800lbs gorrila in chip making.
I'm not sure precisely what you mean about "consumer electronics". You can always underclock a chip to the point that it won't generate that much heat. I've heard of people doing that with really an "old" 1.8Ghz, and underclocked it (like you would with a laptop), way, way, way down, and used it as a part in a no moving parts router. (I've got a machine that is a powersupply, and a CPU fan away from a no moving parts CPU). I just haven't taken the time to find a 150Watt powersupply, and underclock the 1.4Ghz processor down to 200Mhz which should be way more then fast enough to handle my dinky routing needs.
No, I'm not interested in paying for support, I don't want, and I don't need. I have no need for support from RedHat, none at all. All I want is the Q/A they did, and the timely security updates. That's it. I'm willing to pay for it. I've bought copies of RedHat for years. Advanced Server 2.1 and 3.0 are the only copies of RedHat Linux, since 5.1 I don't personally own a copy of. I've even got the Deluxe or Professional Edition of several of them.
I'll pay'em $25-50/year/machine for access to binaries that have security updates, and 2 copies of the original media. I will never call them for support. I'll never e-mail them. I'll never request support from them via a web forum.
Google and the errata page has solved every problem I've ever had with RedHat.
I'd like to deploy RedHat to the desktop to 50 machines at work, for people who only use a web browser. It's not worth $170/year/machine for me to do that. Especially not because I have paid for a licensed browser and OS that are already on it. I'll stay with what I have, because RedHat priced themselves out of being an economical alternative.
Especially because they are identical machines, a problem on one of them is a problem on all of them. They only have to fix the problem once. It's not like the other 49 copies of the problem are likely to add too much in term of support burden on them.
I'm gonna use the same distributor for my servers and my desktop if that is at all possible.
I'm just not paying more for the software and support then I am for the hardware. I'd love to give RedHat my money. It'd be essentially free for them to do (+/- bandwidth costs).
I do pay the support for Oracle. I've paid for a copy of RHEL, I have to have one copy for Oracle.
My question is, given that I have one, is there any legal (that RedHat won't ask me for more money for), for me to install more copies.
What's complex about this? My question is can I do follow the rules! What I'd really like is to take advantage of the GPL'ed software. However, I've signed a support agreement which is completely independent from that.
RedHat is the other party in a legal contract with me, I'd like to know what their interpretation of that contract is. I bet plenty of other people on slashdot who have to sign the same contract would like to know the same things.
If I don't agree to the contract, I have no rights to the binaries. The part where they agreed to give me the binaries is the same part where I agreed to the contract. Sure I didn't sign it, but I'm not willing to send in my legal team against RedHat to prove that in court.
I'll go with one of the myriad of people rebuilding the SRPMS as third parties if RedHat says there is no way for me to be in line with their agreement while installing the signed binaries on more machines. However, RedHat did all the work. If they provided me a way to pay for it, and get the binaries from them, I'd happily take them up on it (I don't want the support, just the security updates, and I'm not paying for 24x7 1hr turnaround on support contracts for it). I'd just take them up on it, but I'm not paying RedHat more for the "free as in freedom" software then I am going to pay for the hardware it runs on. I'll go run Debian or Suse (depending on how the Novell stuff shakes out).
RedHat actually has some claim to the copyright on the structure of the ISO, on the layout of the directories, on a lot of the original software. They have copyright claims on all kinds of binaries, and the binaries that have their signatures. Third parties can get themselves into trouble if RedHat decided to stop playing nice.
Right now, I've got a legal agreement with RedHat which allows them to bill me, and fine me, and furthermore, drop my RedHat support, which ruins my support with Oracle. All that's really bad for me. Hence I have a question. I'm not asking for advice on how to get a copy of running copy of the software. I can do that. I've built the binaries from source (most of it successfully, a handful failed to build), mostly because of stupidity in the RPM's overriding the default build area (bash really like to build in/var/tmp, and overrides your settings in the.spec file for instance).
I know how to plug those onto an ISO and install them. I knew everything you've had to say in this thread, long before you said to me. I have a question for the other party in a contract a lot of people who are on slashdot are probably one of the two parties of too!
Okay, your not getting it. I signed a contract (or agreed to the terms of the licensing for the support I do have, and continue to want), that if I install the binaries they gave me, as a term of my support agreement, they can bill more.
I understand the GPL. I understand I can take a lot of the binaries and just reuse them. However, some of them, including say the XFree86 server, they could claim copyright on the binaries and cruify me in court.
I need a copy of XFree86 in order to install Oracle. Thanks, but go read the agreement like I said originally. It's crystal clear on several points, including that the terms of my support agreement mean that installing a second "unsupported" version of RHEL isn't something I can do, I agreed to a support contract that says that in a nut shell.
Furthermore, I can't terminate the support contract during the first year unless RedHat is in breach of contract. So I can't just by the thing, get the binaries, terminate the contract and then install it as many times as I like. Otherwise, I'd do just that, and then build it from source.
I'm not going to play games, this is a business I work at. I can't run an unsupported version of Oracle. It'll run fine for 90 days at a streach, but on the 91st day, you really, really, really need Oracle. I'm not going to have Oracle, or RedHat of them drag me into a legal tangle because I didn't have my support ducks in a row.
Furthermore, I'm a programmer, I want everybody to be on the up and up with my licensing, I want to be on the up and up of theirs. They are pretty damn clear that an additional install of RHEL is a billable event. It's really clear they want to bill me by the server, and there isn't any provision for installing an "unsupported" version of RHEL.
Read the licensing agreement again, and this time closely. If RedHat showed up on my doorstep and auditted my place (which they have a contractual right to do), they have a contractual right to invoice me for all of the installed copies (anything which is installed you get billed for the support). That's what you agree to when you get your original copy. I've heard arguments that either, I can pay the invoice or lose my support. I think RedHat could drag me to court and whoop me until I pay them money, plus legal fees, plus the 20% surcharge if I am "cheating" by more then 5%.
Anything that is GPL'ed I could construct my own installer using their GPL'ed software and release my own ISO. However, they could claim copyright on any packages that aren't GPL'ed if I released the copy with their signature.
There are a handful of packages which I would have to replace, including the "RedHat Logos", and a few other minor things. The IBM JRE and I think some Intel profiling (that might be a different RedHat release) tools are included.
I'm more likely to go with the group that is recompiling the whole thing from SRPMS in Germany. I forget the name of the group doing it. Then try my hand at having to deal with an audit.
Is there any legal way for me to get a copy or a RedHat production quality without support?
I have no interest in dial up, e-mail or the web based support. I've had Linux, and RedHat in particular on my personal desktop for 5 years at home, and 3.5 at work. I've supported hundreds of Linux machines over the years. In all that time, access to updated security was all I ever needed. I'm willing to pay for it, I've always purchased copies of new RedHat distributions.
I'll happily pay a reasonable price for the completely unsupported desktops, with access to security updates. However, $170/year/workstation isn't a reasonsable price ($50/year/workstation, and better quantity pricing is getting closer to the mark).
I have no need of most of the support infrastructure you provide. All I really want is the QA. Once RedHat says it is good to go, I've learned that they are generally right, and if they aren't, there will be a fix for it quickly.
Do you have a solution for workstations and file/print servers, that isn't more expensive then buying new PC's every three years (for 3 * $170, I can get a replacement desktop from Dell, for 3 * $349, I can get low end server, and for 3 * $1499, I can get an highend server)?
Bleeding edge technology from Fedora isn't what I want to be placing on desktops and file servers. I surely don't want to deal with updating it 2-3 times a year if Fedora isn't going to support older versions of the OS, especially if security fixes are "upgrade to the newest version". For the same reason your company always recommended never to use RawHide on anything important.
I'm the proud owner of a RedHat 2.1 AS server, and will continue to pay for the support of my servers that require specialized support (Oracle DB/Reports servers), where I do use it. However, for print server, file server, and desktop use, there are no price/quality competitive options to just buying and using Windows, or migrating to a different distribution.
So what your saying is that by installing a small either time based, or heat based variable resistor, I could construct a lightbulb that last forever? If all that blows the thing up is too much current, then using a variable resistor that slowly (in this case over a period of 10-20 seconds, or make it heat based) takes the current from 0 to full current, then no lightbulbs should burn out then right?
I wonder if a guy could build a light fixture adaptor that does just that. Just create an inline light fixture adaptor, that takes roughly 10-60 seconds to step up the voltage. Sure my lights would take 10-60 seconds to light up, but if the bulbs lasted forever, what do I care.
Hmmm, curious, read the links section on the light that has the 96 year old light bulb. It's on the related links section of the HowStuffWorks.com. Clearly it's possible to make a very long lasting light bulb. Then again, I might have misunderstood the instructor.
I've got a minor in physics, I understand the basics of electricity.
When was the last time you actually saw a bulb blow while it was running? I've never seen it happen in all might life on an incandesent light bulb. Ever. It is something about turning the bulb on, he claimed it had to do with where on the sine wave you are when it's turned on, and how many hours it had been run. I didn't have enough material science to argue with him. It's what my college instructor told me, I took in on faith, all his demonstrations of weird properties worked just fine. As some says further down in the thread, it could be done just as I said it could, however, it'd be highly ineffecient powerwise.
Toner, batteries, light bulbs and razors in a capitalism economy, it is in the best interest of a small cartel of people to control the cost and the quality of product. Just as the OPEC nations can. Because it makes the most sense for the consumer of the product to get the longest lasting (if they have the same efficiency), the cheapest toner (that has the same quality), and people buy the blades for the cheap razors they bought.
The reason it generally works out, is that the consumer is relatively uninformed, and doesn't think things thru, and/or can't vote with their money because it's a cartel, there are no options. A small group of companies work this area pretty hard. Companies do in fact make cheaper toner then HP or Lexmark. Last I knew, Lexmark was legally challenging people under the DMCA and patent law.
Oil cartels tightly control the supply of Oil to keep prices high. The RIAA was found guilty of price fixing. Microsoft was found guilty of anti-competitive practices, including using OEM's as leverage to keep people out, and have historically just bought any product they couldn't beat.
Just because it is in the best intersest of the consumer to produce more Oil. Just becaue it was in the interest of the consumer not to have the RIAA price fix. Just because it was in the interest of the consumer for Microsoft to compete fairly doesn't mean that is precisely what happens. The interest of the consumer isn't what businesses in a capitalistic society are all about.
If a battery producer could theoretically create a set of batteries that had 100 times the power in them, that cost them the same to produce, it wouldn't be in their own best to sell them. If the battery industry was a cartel (I don't know that it is), it wouldn't get released. It is not in the interest of the company to do so, so it doesn't get done.
The claim that a company would willing release a product that would cut profits by 100 times is insane. Sure they'd own all of the market, in smaller market. That's not smart business. Now, if they tried to charge 100 times the price, they couldn't batteries have a sweet spot in terms of price that the price needs to be between $2-$20 for a pack of batteries at a local grocery store.
Lets say they make batteries that will last 100 times longer, and be priced at times times as high, and cost the same to make. Lets say the current industy is worth $10Bil a year. You'll take a $10Bil/year industry, and now turn it into $1Bil/year industry. Oh, did I mention that your fixed costs are the same, so your profits go down faster then linearly. So your profits will be divided by at least 10. If I told my boss I had a great advancement on technology, but as a side affect it would shrink the market by 10 times, he'd fire me, or more likely, he'd pay me all the money I ever wanted to never tell anyone that.
Tell some VC guy, you have a way to make a battery that will last his entire life, and he'll tell you, he won't fund you to make batteries to compete with Duracell, because you
Actually, you are correct, the RMS of american power, is around 120, but peaks at 170 (the RMS means Root Mean Squared, so it's an average, in this case of a sign wave of the power, so 120 isn't the highest point). The bulb is rated for 120v (which is why a bulb will last essentially forever if you turn it on and leave it on, the average voltage is what the bulb is rated for). If you turn a bulb on at a peak, it has issues and blows. It's not like a light a bulb blows because it's too hot when you first turn it on. It clearly ran when you turned it off. You'd think the heat is what would ruin bulb. Never has been that I've seen, it was alway the initial surge of power. I knew I'd get the details wrong, and some clever slashdotter would point it out.
The instructors was somewhat insane, but what he said makes sense. He's also a considered a leading authority on Extra-Terrestrials. For as insane as he was, they classified a bunch of basic research he did into application of nuclear physics. You also couldn't find a problem in the book he couldn't solve. He was really smart about his physics, and it's applications. He taught all of the electronics classes.
It's also backed up by my emperical evidence of picking up the light bulbs that are 130v instead of 120v last longer in fixtures over the last 5 or 6 years. They aren't any more expensive then the bulbs you pick up a Walmart.
Slashdot Mirror
Those are the early adoptors of nearly all technology, and drive the prices to points where normal financially sane people can afford them. Half of all technology every invented was driving by one of those three groups.
For me personally, I use my desktop for lots, and lots of compiling. I'd like my desktop to be more responsive under heavy I/O load. I'd like it to do more things in the background. Personally, I've probably got enough CPU for my desktop. I found find a use for more, but I sure could use lots more disk I/O. I could use more memory also.
I'd like my desktop interface to be much more sophisticated, and to have a lot more context sensitive abilities. All that takes more CPU horsepower. The eye candy is nice, but that's not that interesting to me.
I'd like a desktop that supports 10 monitors. I'd like even more of a network transparent desktop then I have using X11. I'd like to have faster then GigaBit, which requires more CPU, more bus, and faster disks.
I'd like to be able to do lots, and lots of streaming realtime encoded TV signal for my home built PVR (encode 3-4, while playing one that has been paused thus having 5-10 minutes of latency in the playback). I'd like to see more ability to have low latency for specific tasks (music, TV encoding, multi-medial in general).
It's really expensive to build just a handful of CPU's. It's much cheaper to make them in huge volumes. Thus Intel and AMD makes them in huge volumes, and they sell pretty well to the consumer, and Intel makes a killing selling the CPU's, the chipsets, and the various electronic devices they either sell or license to others. So literally the thing driving more, faster CPU's is economics. You just can't buy the old CPU's any more. When you want a new CPU to replace a 5 year old computer, you have to buy a new one.
I'd like more ability to parse the natural english language. Thus facilitating better ability to write a content crawler that can point out urgent e-mails, while crawling the google cache finding content that I'd find amusing.
I'd like the ability to do more and better datamining. To correlate more information about my machines and servers. More ability to automate more and more task to with computer.
I could use more CPU power to use to secure my machines. If I could log every incoming packet and analyize more and more correlations, and deduce when people are doing bad things that'd be great. If I could use it to do more analysis of software and code. To develop more, and more sophisticated techniques of telling when software does something stupid.
I'd like it because it means, the construction of larger, and larger software applications is possible.
Kirby
A. You are being irrational and short sighted (both for the nurse and surgeon). Basing your judgement on an for entire classes of medical fields on a single incident is completely and utterly irrational. Change that from nurse to "White Doctor", and from surgeon to "Black Doctor", and re-read your story. Everbody and their brother would call you a racists.
To make that sort of judgement you'd make, you'd have to do a comprehensive study of surgeons and nurses, and the outcome of their recommendations. I'm willing to bet dollars to donuts, that if you do that, the surgeon's (as an entire medical class) recommendations have a far better survial rate then a nurse's (as an entire medical class) recommendations. I've known a lot of nurses (I used to work in on the IT equipment on hospitals). Nurses make mistakes too. I've known nurses who can't calculate how many drops per minute somebody should get in an hour, because they can't do simple algebra problems.
However, in your case, it could be that the doctor is incompetant and should have his license revoked.
B. You have no idea that she recovered because of the medication. The human body, especially near death is very strange. People have recovered from being dead 45 minutes after being put in a freezer. There is no control to state that she lived soley (at least not in your story).
C. When I said nurse, I meant run of the mill nurse at my personal doctors office, not an ICU nurse. ICU nurses have lots of knowledge I would take seriously if they gave me advice. Generally, any nurse at a family practice, will only give you advice that will make you feel better, but you'll get better without the advice. About the only medical advice you'll ever need is: "Take it easy, drink lots, and lots of fluids". For 95% of everything that is ever wrong with you, that's all you have to do.
Kirby
In my experience, anything a nurse (or General Practitioner) tells you can be ignored, you'll still get better. A nurse generally gives advice that makes you feel better more comfortable, and possible speeds the process along a bit, but inheirently does nothing to fundamentally change the outcome. Generally I stopped seeing a GP unless I need a bone set, or I have been sick for a week.
However, when you have a bleeding brain, nothing but a brain surgeon will do. When you have a pile of bad C code, a really good programmer, or an average programmer will both get the job done (in differing amounts of time). So there isn't as much selection pressue on the job of a programmer.
If I found a brain surgeon who was nice, I wouldn't let them operate on me. Clearly they aren't a real brain surgeon if they are a decent human being :-)
Finally, if you had quoted the following sentence, I pointed out that, comparing programmers to programmers is just as fair as a nurse to a surgeon. If you made a nurse do a surgeons job, there'd be an order of magnitude difference, if you made a surgen do a nurses job, there'd be an order of magnitude differece in quality. If you took a programmer whose really good a job X and make they do job Y, it's not terrible shocking there is fall of. A lot of programmers take work, and do work in areas they lack experience or knowledge, because it is a good job, and the people doing the hiring can't tell the difference.
Skill as a programmer, because programmers have a very, very broad range of skills and abilities that they need to do to accomplish their job, are inheriently incomparable in most ways. Finally, a lot of great programmers are great on the codebases they work on, but they'd be lousy on other codebases.
Kirby
There is an order of magnitude difference in skill between a nurse, and a brain surgeon. No it's not the same job, but then, coding the Linux TCP stack has very little to do with being a VB programmer developing a new front end for existing applications at a major insurance company.
There is probably an order of magnitude difference between a new carpenter, versus a experienced carpenter who was trained by another very experienced carpenter.
It's only in the computer field where differentiating between the crappy, and the great, and the incredible demand for the skills lets those with so little relavent talent continue to earn a good wage doing so they have no incentive to find another line of work.
I know I'm not kernel coder quality, but I'm a damn fine programmer at C++. So I'm somewhere on the scale, but not at the very high end.
Kirby
I thought that Fedora had a development version of a 64bit version for the AMD processors. I could be wrong.
Kirby
According to that link, Alan has a BSc in CS. Linus Torvalds has a Bachealors degree in CS, and an honorary Ph.d from the same school in Finland. I'm too lazy to dig up links for that. It's in several of the books about his life.
Kirby
On block "B", in the forth row, I punched down every jack in the house. I labelled every pair, and had a map of where they were in the house.
You then run patch cables from column 4 block "A", to column 1 block "B", and use the metal clips to complete the circut. Thus any time I wanted to change out wires, all it took was a small run of wire from block A to block B. The original wires from the phone company or to the jack, never had to be touched to manipulate the system. You want to get the wires from the phone company, and the wires from the jack so you, never ever touch them. It makes trouble shooting lots easier, and you'll never run out of cable.
I might have introduced some line noise due to the way I duplicated the lines on block A, but I don't use dialup, so that was never an issue.
Now, if you want to get really fancy, wire each cat 5 cable into a patch panel. Thus you can put phone, or ethernet to any jack in your home. Now you can skip block "B" (the patch panel acts like the punch down block). Now, use row 4 of punchdown block A, and take phone jacks straight to the patch panel. You'll have eight copies of each line to distribute around the house as you see fit. When you want to move a line, just change the port it's plugged into on the patch panel. Unless you do phone, and network, this is really overkill. Just running the punchdown blocks will work great.
I figure that if the phone company uses punch down blocks, they are good enough for me. The tools are kinda expensive ($50-80 for a punch down tool, and the heads you'll need to do it). The punchdown blocks themselves are dirt cheap, like $8 bucks a piece. The patch panels are pretty expensive.
Kirby
Possible the word I was looking for was generation. However, I'd be surprised from the stories I hear from my Uncle's, and from my Dad, if things 20 year olds in the 1950's were truely any better then 20 year olds today.
Watch any game of pickup basketball by any group of 25 year olds. If there aren't a half dozen fouls every time up and down the court, they aren't really playing hard.
Now, I've never seen a pickup game result in a shooting, however, I'm doubtful it's never happened. I'd be very impressed if a gun has never been pulled after a college football game.
I think it has more to do with the nature of the attitude of people of that age, then it does with people who play computer games.
I've always wondered by nobody did this before, but then realized just how much space it would take to store. In order to store that many keys to be able to brute force them all, is insane.
Kirby
Notice, how you'd have to print URL's to it (so you have to type it in), not the actual comments. Someone might find the comments amusing for the first 5 pages. Enough to earn it a 2 by rounding error or something. So even most of all the worthless comments on Slashdot (some of which are mine), wouldn't even rate a 1. Forcing someone to type the links to get them, would rate a 1 however.
Kirby
So it might be that the random number generator wasn't so random on the one we used. I can't find a reference off hand. Pretty much everybody I can find says that using the MD4 is not as secure as MD5, but still secure as a hash.
I used one once, where you had to have a username, know the current password, plus you had to download this goofy little Java applet and run it locally, giving it a long passphrase that you had to remember permanently. It output your onetime password for you to use. I suppose someday I should take to using S/Key and generating a list of OTP. So if you were out onsite at a Hospital, that was the only way to log in remotely.
All authentication, should be done based on something you have, and something you know. So having a keypad that generate OTP's, or having some type of secure dongle. That, and an additional password should be needed.
I picked up one of those keyfob's from think geek, haven't written down a password in a long time. About my only gripe, is that it doesn't have enough password slots, and it doesn't generate long enough passwords.
Oddly, nearly all of this guys issues are about authentication, none of them are about authorization, or about other forms of security.
Kirby
Yes, it sure might be a 6 figure job, but a 6 figure job in Boston, doesn't afford you the lifestyle that a $30K a year job does in rural NE (not where I live, but within an hour or two drive of here).
I laughed at the guy I know who didn't want to move back to Omaha, NE (where I live), from Boston after his .com collapsed. Because he'd have to take a 60% pay cut.
Neverminding the fact, that I lived off less (including my new truck payment, and my nice apartment payment), then he did for rent in his crappy high rise apartment in a crappy neighborhood. Rent out there, costs about 3 times what it does in Omaha, and my rent is about 1/3 of my monthly living expenses. You do the math. I easily lived pretty well off the 60% paycut he didn't want to take.
The only good part about living in a very expensive place, is two fold. Non-necessities are a lot cheaper relative to your salary (so a $1000 computer, would be 1/10th, instead of 1/3 of your monthly salary). Also, if you build equity there, and then move someplace else, you can live like a king. A $80K house in Omaha, would probably run $300K in Boston. You probably couldn't buy as big a lot as the $80K house was in the Boston area.
A lot of this, is a microcosm of why it's so much cheaper to run a business in Omaha, NE then it is in Boston, or SF. No it's not India cheap, but it's pretty damn cheap.
Kirby
First off, I have several questions. Do you have an original copy of the partition before you started running recovery tools (after you deleted the file, but before you created new ones)? If not, make an image immediately. You want the most original image you can find. Now, the second question, is how much data am I a looking for, and how large is the partition? (How large is the needle, and how large is the haystack?). What type of data am I looking for? Is it a word document? A text file? A gif? A jpg? Some html? A PDF? The smaller the file, the more likely that if it got overwritten, it all got overwritten. However, the more likely you are to recover all of it. If it was a very large file, it's possible that you can recover pieces and parts, but not all of it. Now, it's my understanding that you can recover anything written to a harddrive, even if you have overwritten it several times. However, it's very, very expensive to do so. So now the question is how much money is it worth to you? The guys as ReiserFS probably have the best shot at helping you. They probably don't want to however. The more you know about the order of the files in the directory, the more you know about how the files were constructed, and the order files got put on disk the better. They you can make better educated guesses about the sequence in which the pages got allocated to know where to go look for the file. Do you have anything on the drive you are worried about posting? Can you post an image of the drive? I'm not an expert in this area, but I've seen people recover mail spools at an ISP using dd. People leave ISP's over losing all their mail, so they worked really hard at it (however, that was an ext2 filesystem). Kirby
Here is a link to a page about it.
It's a neat idea, but I'm not sure how popular it is with hardware makers, and it somewhat constrains the implementation in hardware. The basic underlying princepals of the hardware would have to support the way the high level model is written, as opposed to having the software conform to the software.
It has to be a split driver model, as OS's organize themselves differently, so what would be highly efficient in one would be dog slow in another. This is also why various people recommend not porting a Windows driver to Linux, but to instead write a native Linux driver. Somebody presented a paper on the 10 things not to do while writing a Linux driver.
Notice, that I2C is also how lots of Linux drivers are written for block devices, because lots of block devices have a high layer, a mid layer, and a low level. Normally the high level, and mid layer are similar between lots of drivers, and generally get squeezed into a single driver.
Kirby
You only have to do Q/A once. Every Engineering tweak you figure out directly applies to all of the chips. You don't have to worry quite as much about not having enough production capacity to make type A chips, because type A has a lot more demand then was originally predicted.
It's my understanding that overclocking sites list which CPU runs of old Celeron's were just sold under a lower speed CPU and could safely be overclocked well outside of the range Intel labelled it as.
There's a reason you get a huge discount when you buy in volume, for similar reasons, it's much cheaper to produce in huge volumes. That's how Intel became the 800lbs gorrila in chip making.
Kirby
Kirby
I'll pay'em $25-50/year/machine for access to binaries that have security updates, and 2 copies of the original media. I will never call them for support. I'll never e-mail them. I'll never request support from them via a web forum.
Google and the errata page has solved every problem I've ever had with RedHat.
I'd like to deploy RedHat to the desktop to 50 machines at work, for people who only use a web browser. It's not worth $170/year/machine for me to do that. Especially not because I have paid for a licensed browser and OS that are already on it. I'll stay with what I have, because RedHat priced themselves out of being an economical alternative.
Especially because they are identical machines, a problem on one of them is a problem on all of them. They only have to fix the problem once. It's not like the other 49 copies of the problem are likely to add too much in term of support burden on them.
I'm gonna use the same distributor for my servers and my desktop if that is at all possible.
I'm just not paying more for the software and support then I am for the hardware. I'd love to give RedHat my money. It'd be essentially free for them to do (+/- bandwidth costs).
Kirby
My question is, given that I have one, is there any legal (that RedHat won't ask me for more money for), for me to install more copies.
What's complex about this? My question is can I do follow the rules! What I'd really like is to take advantage of the GPL'ed software. However, I've signed a support agreement which is completely independent from that.
RedHat is the other party in a legal contract with me, I'd like to know what their interpretation of that contract is. I bet plenty of other people on slashdot who have to sign the same contract would like to know the same things.
If I don't agree to the contract, I have no rights to the binaries. The part where they agreed to give me the binaries is the same part where I agreed to the contract. Sure I didn't sign it, but I'm not willing to send in my legal team against RedHat to prove that in court.
I'll go with one of the myriad of people rebuilding the SRPMS as third parties if RedHat says there is no way for me to be in line with their agreement while installing the signed binaries on more machines. However, RedHat did all the work. If they provided me a way to pay for it, and get the binaries from them, I'd happily take them up on it (I don't want the support, just the security updates, and I'm not paying for 24x7 1hr turnaround on support contracts for it). I'd just take them up on it, but I'm not paying RedHat more for the "free as in freedom" software then I am going to pay for the hardware it runs on. I'll go run Debian or Suse (depending on how the Novell stuff shakes out).
RedHat actually has some claim to the copyright on the structure of the ISO, on the layout of the directories, on a lot of the original software. They have copyright claims on all kinds of binaries, and the binaries that have their signatures. Third parties can get themselves into trouble if RedHat decided to stop playing nice.
Right now, I've got a legal agreement with RedHat which allows them to bill me, and fine me, and furthermore, drop my RedHat support, which ruins my support with Oracle. All that's really bad for me. Hence I have a question. I'm not asking for advice on how to get a copy of running copy of the software. I can do that. I've built the binaries from source (most of it successfully, a handful failed to build), mostly because of stupidity in the RPM's overriding the default build area (bash really like to build in /var/tmp, and overrides your settings in the .spec file for instance).
I know how to plug those onto an ISO and install them. I knew everything you've had to say in this thread, long before you said to me. I have a question for the other party in a contract a lot of people who are on slashdot are probably one of the two parties of too!
Kirby
I understand the GPL. I understand I can take a lot of the binaries and just reuse them. However, some of them, including say the XFree86 server, they could claim copyright on the binaries and cruify me in court.
I need a copy of XFree86 in order to install Oracle. Thanks, but go read the agreement like I said originally. It's crystal clear on several points, including that the terms of my support agreement mean that installing a second "unsupported" version of RHEL isn't something I can do, I agreed to a support contract that says that in a nut shell.
Furthermore, I can't terminate the support contract during the first year unless RedHat is in breach of contract. So I can't just by the thing, get the binaries, terminate the contract and then install it as many times as I like. Otherwise, I'd do just that, and then build it from source.
I'm not going to play games, this is a business I work at. I can't run an unsupported version of Oracle. It'll run fine for 90 days at a streach, but on the 91st day, you really, really, really need Oracle. I'm not going to have Oracle, or RedHat of them drag me into a legal tangle because I didn't have my support ducks in a row.
Furthermore, I'm a programmer, I want everybody to be on the up and up with my licensing, I want to be on the up and up of theirs. They are pretty damn clear that an additional install of RHEL is a billable event. It's really clear they want to bill me by the server, and there isn't any provision for installing an "unsupported" version of RHEL.
Kirby
Anything that is GPL'ed I could construct my own installer using their GPL'ed software and release my own ISO. However, they could claim copyright on any packages that aren't GPL'ed if I released the copy with their signature.
There are a handful of packages which I would have to replace, including the "RedHat Logos", and a few other minor things. The IBM JRE and I think some Intel profiling (that might be a different RedHat release) tools are included.
I'm more likely to go with the group that is recompiling the whole thing from SRPMS in Germany. I forget the name of the group doing it. Then try my hand at having to deal with an audit.
Kirby
Is there any legal way for me to get a copy or a RedHat production quality without support?
I have no interest in dial up, e-mail or the web based support. I've had Linux, and RedHat in particular on my personal desktop for 5 years at home, and 3.5 at work. I've supported hundreds of Linux machines over the years. In all that time, access to updated security was all I ever needed. I'm willing to pay for it, I've always purchased copies of new RedHat distributions.
I'll happily pay a reasonable price for the completely unsupported desktops, with access to security updates. However, $170/year/workstation isn't a reasonsable price ($50/year/workstation, and better quantity pricing is getting closer to the mark).
I have no need of most of the support infrastructure you provide. All I really want is the QA. Once RedHat says it is good to go, I've learned that they are generally right, and if they aren't, there will be a fix for it quickly.
Do you have a solution for workstations and file/print servers, that isn't more expensive then buying new PC's every three years (for 3 * $170, I can get a replacement desktop from Dell, for 3 * $349, I can get low end server, and for 3 * $1499, I can get an highend server)?
Bleeding edge technology from Fedora isn't what I want to be placing on desktops and file servers. I surely don't want to deal with updating it 2-3 times a year if Fedora isn't going to support older versions of the OS, especially if security fixes are "upgrade to the newest version". For the same reason your company always recommended never to use RawHide on anything important.
I'm the proud owner of a RedHat 2.1 AS server, and will continue to pay for the support of my servers that require specialized support (Oracle DB/Reports servers), where I do use it. However, for print server, file server, and desktop use, there are no price/quality competitive options to just buying and using Windows, or migrating to a different distribution.
Kirby
I wonder if a guy could build a light fixture adaptor that does just that. Just create an inline light fixture adaptor, that takes roughly 10-60 seconds to step up the voltage. Sure my lights would take 10-60 seconds to light up, but if the bulbs lasted forever, what do I care.
Kirby
I've got a minor in physics, I understand the basics of electricity.
When was the last time you actually saw a bulb blow while it was running? I've never seen it happen in all might life on an incandesent light bulb. Ever. It is something about turning the bulb on, he claimed it had to do with where on the sine wave you are when it's turned on, and how many hours it had been run. I didn't have enough material science to argue with him. It's what my college instructor told me, I took in on faith, all his demonstrations of weird properties worked just fine. As some says further down in the thread, it could be done just as I said it could, however, it'd be highly ineffecient powerwise.
Toner, batteries, light bulbs and razors in a capitalism economy, it is in the best interest of a small cartel of people to control the cost and the quality of product. Just as the OPEC nations can. Because it makes the most sense for the consumer of the product to get the longest lasting (if they have the same efficiency), the cheapest toner (that has the same quality), and people buy the blades for the cheap razors they bought.
The reason it generally works out, is that the consumer is relatively uninformed, and doesn't think things thru, and/or can't vote with their money because it's a cartel, there are no options. A small group of companies work this area pretty hard. Companies do in fact make cheaper toner then HP or Lexmark. Last I knew, Lexmark was legally challenging people under the DMCA and patent law.
Oil cartels tightly control the supply of Oil to keep prices high. The RIAA was found guilty of price fixing. Microsoft was found guilty of anti-competitive practices, including using OEM's as leverage to keep people out, and have historically just bought any product they couldn't beat.
Just because it is in the best intersest of the consumer to produce more Oil. Just becaue it was in the interest of the consumer not to have the RIAA price fix. Just because it was in the interest of the consumer for Microsoft to compete fairly doesn't mean that is precisely what happens. The interest of the consumer isn't what businesses in a capitalistic society are all about.
If a battery producer could theoretically create a set of batteries that had 100 times the power in them, that cost them the same to produce, it wouldn't be in their own best to sell them. If the battery industry was a cartel (I don't know that it is), it wouldn't get released. It is not in the interest of the company to do so, so it doesn't get done.
The claim that a company would willing release a product that would cut profits by 100 times is insane. Sure they'd own all of the market, in smaller market. That's not smart business. Now, if they tried to charge 100 times the price, they couldn't batteries have a sweet spot in terms of price that the price needs to be between $2-$20 for a pack of batteries at a local grocery store.
Lets say they make batteries that will last 100 times longer, and be priced at times times as high, and cost the same to make. Lets say the current industy is worth $10Bil a year. You'll take a $10Bil/year industry, and now turn it into $1Bil/year industry. Oh, did I mention that your fixed costs are the same, so your profits go down faster then linearly. So your profits will be divided by at least 10. If I told my boss I had a great advancement on technology, but as a side affect it would shrink the market by 10 times, he'd fire me, or more likely, he'd pay me all the money I ever wanted to never tell anyone that.
Tell some VC guy, you have a way to make a battery that will last his entire life, and he'll tell you, he won't fund you to make batteries to compete with Duracell, because you
The instructors was somewhat insane, but what he said makes sense. He's also a considered a leading authority on Extra-Terrestrials. For as insane as he was, they classified a bunch of basic research he did into application of nuclear physics. You also couldn't find a problem in the book he couldn't solve. He was really smart about his physics, and it's applications. He taught all of the electronics classes.
It's also backed up by my emperical evidence of picking up the light bulbs that are 130v instead of 120v last longer in fixtures over the last 5 or 6 years. They aren't any more expensive then the bulbs you pick up a Walmart.
Kirby