However, why again do we need spend millions of dollars on these voting machines and training and maintenance, etc. jacking up my taxes just so I can play with some new shiny object?
Because your elected lawmakers decided that you would.
Accessibility for the handicapped was a major driver. Issues with previous balloting systems were another.
As usual the legislative solution wasn't well thought out and had unintended consequences. Now we have to fix the fix.
Given the importance of visibly accurate elections (which is how republics avoid civil war) it's vitally important to get this done.
Given that reverting to pure paper balloting would regress the improved handicapped access, complete elimination of the electronic voting aids is not an option. (Note that some jurisdictions that have decertified the Diebold mahchines - such as Alameda County CA - still allow their use for handicapped voters until a better solution can be deployed.)
By requiring that the entire platform be open source, the well-intentioned legislators just killed the bill.
The version of "open source" required doesn't give away any copyright or patent protection, or transfer rights to USE the code to others - especially the competition. (It does puncture trade secret.)
If the bill had instead required that only the voting software installed on the voting machines be open source...... it would have been ineffective against malware embedded in the operating system - by the OS developers or later black-hats.
This is not a minor issue: With control of the US Government at stake a LOT of engineering effort can be profitably applied to attempts to compromise the system - by political, economic, or foreign governmental interests.
If this bill passes... [Diebold] can then sell new machines to [all their former] customers.
Or printer and software upgrades.
If Diebold fixes the auditability problem I have no further gripe with the use of their machines. If buying an upgrade from them is 'way cheaper than replacing the machines outright, that's just dandy.
"If it's worth doing, it's worth doing at a profit."
- The voting machine result can be used whenever the particular precinct's results are uncontested, leaving all the advantages (except the "advantage" of being able to invisibly rig an election) intact.
- The auditability of the result will virtually eliminate any utility in rigging the machines (while bringing to bear draconian penalties for attempting to rig them and getting caught), greatly improving the reliability of the machines' results - to the point that they CAN be used without creating the appearance of rigged elections.
Near the end of TFA is the important reason for them to get their support from Oracle:
They're running Oracle's application server and database on some of the mission-critical servers.
With Oracle supporting both the application and the OS under it there's no time and money lost to finger-pointing when something gets hosed. Instead a single team buckles down and fixes it immediately.
(Presuming they ever need service. One of the comments from Red Hat indicates that they may never have actually had to USE the service contract. Take THAT, Microsoft! B-) )
PB can buy all the Sealands they want; Hollywood and the rest will just sue the ISPs for providing the bandwidth.
Given their budgets they could also hire some mercenaries and mount an attack on Sealand themselves.
If they filmed it they might make a profit on it, too.
Copyright (c) 2007 by me writing as "Ungrounded Lightning Rod".
Leave a followup to any posting in my journal with a firm offer if you want to do the movie. Otherwise I may sue for copyright infringement if such an attack is made, filmed, and the film shown for profit - even on a news operation under the same umbrella corporation. B-)
you can usually survive 1 non-matching transfusion, even if it's incompatible. After that your immune system is 'primed', much like for a vaccination, and will attack a second transfusion with extreme prejudice, likely resulting in your death.
Which brings up the issue of whether the body would develop an immunity to the enzyme, potentially producing a fatal anaphylactic reaction upon a future transfusion.
On the other hand, if the enzyme remains in the serum rather than attaching to the red cells the reaction would not produce the fatal clumping. Meanwhile the allergy to the enzyme, even if severe, could be handled by other drugs...
Which would also suppress the immune system somewhat - in a hospital "superbug" environment. So artificial type-O will likely remain an emergency measure, and type-O donors will remain in demand.
The problem with OMGPonies-day at/. is that I begin to doubt even pleasant, plausible news by proximity...
Back in the days of the Cold War I occasionally thought that the ideal (for them) time for the Soviet Union to launch a nuclear attack on the US would be early morning on April 1. Response would be slowed (except, perhaps, in the military itself) by the concern that somebody was pranking. Now that it's "War on Terror" time, ditto for the terrorists.
Once such an attack, with a large body count, is (successfully) made, pulling an April Fool prank MAY become sufficiently taboo to put the brakes on the madness. Unfortunately, I can't imagine anything short of that which might slow it down.
Meanwhile, I've stopped even bothering to read Slashdot on (or near) April 1. When the fools are having the fun the site is utterly useless to me.
In some locales, you can't legally call yourself an "engineer" unless you have a PE to your name, much like you can't start working as a doctor or lawyer without appropriate paperwork.
In most states calling yourself a software or computer engineer is an explicit exception to such requirements. It's a job title, not a ceritification. (I understand one state does require some certification to use the E-word even for computer-related stuff.)
This is largely because the industry has a track record of rejecting such regulatory attempts as stifiling of progress and monopolistic. There was one serious attempt to create such a cert, and if it had succeeded you'd have to be a Cobol expert and understand JCL to work in the field now. Instead, certifications are now industrial internal matters, issued by profit-making corporations and worth the paper their printed plus the reputation of the company's certification operation.
Given that some of the best and brightest - including many of the pioneers - didn't even get a degree, it's not surprising. During the early days the experts were the ones who became so obsessed with working/"playing" with the machines that their other schoolwork often suffered, and many of the best ended up dropping out short of a degree. Meanwhile the early coursework in the field tended to be a sidetrack from what was really needed by industry, and for at least the first ten years of computer science education a 4-year "computer science" degree was actually a hinderance to employment in industry. (You'd think the professors thought everybody was going to write his own compiler, just for starters.) The graduate had to "unlearn" a lot of stuff before he could really pick up on what the employer wanted done.
Minsky divided the first thirty years or so of computer science education into three periods:
1) 4-year degree was counter-productive, as above. Faculty didn't really understand what was needed.
2) 4-year degree was actually useful. Faculty had rented a clue about what was needed.
3) 4-year degree was counterproductive again. Faculty was teaching excessive formalisms and software fads.
(It's been long enough for maybe two more "periods" now but I haven't heard whether Marvin has characterized them. B-) )
Radio waves are alternating magnetic fields. The faster the field is alternated the more power it has and the further it travels.
Actually this is wrong. At least on the 'further it travels' aspect.
Also the "just magnetic field" part, (as explained by another poster above.)
And also the "faster the stronger" part.
The rate of alternation is just the frequency of the wave, and has nothing to do with power.
The power for a given waveform and polarization is proportional to the square of the peak field strength. (For an arbitrary waveform it's actually the integral of the magnitude of the cross product of the electric and magnetic field vectors...)
Perhaps more like "maybe a classical explanation is enough, assuming quantum-mechanical approaches are necessary make you miss some useful stuff and work too hard at it".
Or as "classical" as you can get with electromagnetism anyhow...
If you don't welcome us properly we'll get in our longships and row our tall, blond asses over there and.... um....... start buying up your internet infrastructure manufacturers.
Countries were judged on... the extent to which government policy creates a framework...
In other words: Government meddling in the economy and industry gives a "better" score, hands-off gives a "worse" one.
For instance:
- The government-mandated pre-divestiture ATT monopoly was "better" than ATT+Sprint+MCI+... and the "seven dwarves" CLECs.
- Government building the internet infrastructure is "better" than a gaggle of backbone peers and a herd of mom-and-POP ISPs.
- Government picking (and subsidizing) particular hardware and software vendors is "better" than letting them fight it out with each other (and a crowd of of FOSS teams with a handfull of licenses and ideological battles).
Look how well (terribly) that last one worked in Japan.
Do YOU really think The Cathedral is "better" than The Bazaar?
... why not run a test current through it to measure the conductivity instead of using an accelerator?
She did.
But hooking up molecule-sized test prods to an ohmmeter was a pain.
So she used a particle accelerator to inject the electrons. (TFA doesn't say what else she used to measure the current.)
I've contemplated using scanning electron beams for electrical measurements. Say: a low-energy electron beam for the negative supply, a high-energy one (creating more secondary electrons than injected electrons) for the positive, and a third one at an energy that turns it back around near the surface (or gets sucked in, depending on voltage) for a voltage probe.
But that's both too large and too energetic for testing single molecules of plastic.
Going the other way and using a particle accelerator to excite some observable side-effect of conduction is quite the hack. (I'd propose giving her an award but her university already did. Waytago!)
I don't understand why it's necessary to offer a copy of their hard disk to the RIAA representatives as evidence of innocence. If they're essentially accusing pseudo-random people of piracy, then isn't the onus on them to prove it? (I realise that in civil cases it's balance of probability, but even so...)
If I read that right:
- The original disk became subject to discovery in the case.
- So they bought a replacement and made a copy of it to continue running on and stored the original.
- They offered to let the RIAA's agent, under supervision, buy a fresh disk (with no deleted music files already on it) and make their own clone of the original.
- And meanwhile, because the defendant had to buy the new disk he's now running on (so the old one could be preserved for RIAA's benefit) they want the RIAA to PAY for that one, too. B-)
Ok. I'm as much of an open source advocate as anyone, but I'm not sure I see what all of the hubbub is about or believe the proposition that this upgrade should be free.
Company makes a box that happens to run linux as the base OS. They should therefore redistribute any changes they make to the GPL'ed code they run. That I get.
Not changes.
They must redistribute THE FULL SOURCE, even if it's NOT changed, of whatever they're loading.
For (essentially) free.
Otherwise they have no right to do the copying involved in loading it.
The right to load it comes only from the GPL's terms, which require you make available to the end users the ENTIRE source if you distribute the object - regardless of whether it has been modified.
They went out of business, and they let people who were former subcontractors give away/sell the information needed to update the system so the end user can continue to use the hardware in some fashion.
And the "former subcontractors" ALSO are under the terms of the GPL. So if THEY have GPLed firmware to download in those dongles, THEY must also release the source to THAT firmware. (Not just if they modified it, etc.)
If they're selling the dongles for $30 they have to sell the firmware source for no more than $30 also.
Meanwhile, the previous company distributed the object so they are obligated to distribute the source. If somebody (like the "former subcontractors") bought that part of the operation, they also bought the liability for distributing the source that matches the object already distributed - which they must satisfy to keep the license rights alive. Otherwise they don't have the rights to the modified code.
If USDTV actually went bankrupt it might be argued that the obligation to distribute source to the previous version was extinguished in bankruptcy. But IMHO that doesn't extinguish Cable Communications' obligation to distribute source when THEY upgrade the firmware. If they do this for free they must distribute the source for free - if they do it for $30 they must distribute the source for no more than $30.
Also: The obligation to distribute the source can not be limited to the original purchasers of the box unless the source is distributed WITH the box. Once even one instance of the box has been through first sale without the source the GPL requires the source to be distributed to anyone on demand.
Rand? Seriously? Shittiest self-styled philosopher and author ever to put pen to page. This passage is about as deep and insightful as "a is a," and about as well written as "My Pet Goat."
You misunderstand Rand's works.
She writes in the form of the Russian Epic Novel.
As a friend of mine once remarked: "Nobody really understands the purpose of the Russian Novel - but it's not entertainment."
This quote might be relevant. But the rest of her shitty books and philosophy are not.
Depends on what subject you're measuring relevance against.
Objectivism has a very important use: It explains, in connected logic understandable by even moderately intelligent psycopaths, how obeying some basic rules of civilized behavior are profitable for them personally.
Teaching Objectivism is the ONLY (non-religious) "treatment" that has a track record of reforming a significant fraction of hardened criminals.
I bought a car about 6 months ago, and I ran into this. Even though I was paying straight cash, they demanded my SSN so they could run a credit report. They said that they needed to run a credit report to see if I was on this list. I argued with them that I was paying cash, they didn't needed my SSN, etc., but it was late and I eventually relented just to keep the process moving.
You should have taken the unsigned contract (with the model and price info) and walked out.
It's amazing how fast that turns sales people around.
And if they don't do a 180, just go to another dealer, ask for the manager, tell them what happened, and offer to buy the car for cash from him at the same price as you'd have paid the other guy. (For him it's a better deal: HE gets the sales commission and doesn't have to waste his and a salesman's time.)
Since the tweaked sheep are chimeras, and humans/sheep don't hybridize, mating two of them (from two different donors for the human cells), if it produced anything at all, would produce either fully-sheep or fully-human offspring.
Human offspring seem likely, since the part of the sheep replaced/augmented with human stem cells was the part that produces the internal organs - presumably including the reproductive tract. (Can somebody more familiar with embryogenesis check me on that?)
First generation might be premature or have nasty birth defects due to being raised in the mostly-sheep chimera's body, with most of its biological feedback mechanisms "programmed" to produce lambs and some to produce humans. Second generation would likely be fully normal people or sheep.
These sheep are not 15% human, there is no such thing - they're 15% antigenically identical. There is no percentage at which they will become human, because their basic structure is still of a sheep!
These are not hybrids or the result of genetic engineering. They are chimeras: Organisms composed of two separate clone of cells. Some of the cells are 100% human, some are 100% sheep. The total animal has 15% of its cells being 100% human (surface antigens and all), not 100% of its cells having 15% human traits. (The immune system matured in the presence of both so it doesn't attack either.)
The cells are in coherent lumps, too. Entire organs - including the brain - may be 100% human tissue. (Though they may be morphologically similar to the sheep equivalent because they were exposed mainly to sheep growth factors while forming.)
(emulation will only require MORE computing power than the actual Windows box)
Use WINE for any applications it supports (which is most of 'em these days).
The acronym stands for "Wine Is Not and Emulator". It doesn't "emulate" a machine running Windows.
What it fakes up is the API: The application runs at full CPU speed while the system and library calls are translated from those appropriate for Windows to the Linux/X equivalents or are supported by native replacement routines.
So the application itself is running at the same CPU speed. Any bets on whether Linux and X (even with system call translation) or Windows take longer to get things done when the application needs support? Or whether the Windows or Linux scheduler gives the application more CPU?
Let's assume someone keeps his machine 5 or 6 years with linux installed on it and suddenly something give, like the cpu or the memory, now how much would it cost for that user to get the same cpu to run on his machine? We assume here it's a server and not some home usage, How much do you think it's going to cost him to find replcment parts?
That happened to me a year or so ago.
We'd gotten two Compaq AMD-based machines just before the Y2K transition. I installed RedHat on one of 'em and kept the other for a spare. Foolish person that I was, I left the spare in the box, rather than pull it out and check it.
Needed to install the spare for a workstation (old one is still running fine - except for video card - more later). Pulled it out and tried to boot it. DOA. (Oops!) A little part-swapping with the operating one showed the problem to be the CPU.
Compaq had spare CPUs but still wanted the old price of $700 for 'em. So I went down to Weird Stuff Warehouse. They didn't have exactly the same version so I bought a slightly faster one for something in the two-digit-dollas range. (Had to try two of 'em - first one was TOO fast for the MOBO's RAM so I had to downgrade to one that was only a LITTLE faster. They have a "try it and swap it" exchange policy on such stuff. They also priced 'em by speed, so I got a partial refund.)
(Turns out the old spare-box wasn't up to the performance level needed for the software that was to run on it. So we bought a new machine for that anyhow. Thus the repaired machine is back to being a backup spare (though the person it was intended for has just expressed an interest in using it for a "try out Ubutnu" machine before migrating the newer one to Linux or dual-boot).)
A few months before the above discovery, the RAM on the video card in the in-use (for half a decade) machine had started to flake out intermittently. (Probably heat from not vacuuming out the dust over that long period.) So I got a replacement video card while I was at it. (About $10. Later revision of the same card, worked just fine. Total cost: Well under $100 plus two trips to the store.)
Point is that "trailing edge technology" spares are often available and quite cheap - if you're willing to accept used parts from scrap dealers. (If it's a set of mission-critical machines, of course, you have to be prepared to migrate when the new-part supply chain has shut down and your stock of working spares is getting thin.)
The big problem arises if you want to expand your deployment significantly after the parts have been end-of-life-ed. THEN you can't get what you need and need to move to something more recent, to avoid problems with support.
Slashdot Mirror
However, why again do we need spend millions of dollars on these voting machines and training and maintenance, etc. jacking up my taxes just so I can play with some new shiny object?
Because your elected lawmakers decided that you would.
Accessibility for the handicapped was a major driver. Issues with previous balloting systems were another.
As usual the legislative solution wasn't well thought out and had unintended consequences. Now we have to fix the fix.
Given the importance of visibly accurate elections (which is how republics avoid civil war) it's vitally important to get this done.
Given that reverting to pure paper balloting would regress the improved handicapped access, complete elimination of the electronic voting aids is not an option. (Note that some jurisdictions that have decertified the Diebold mahchines - such as Alameda County CA - still allow their use for handicapped voters until a better solution can be deployed.)
By requiring that the entire platform be open source, the well-intentioned legislators just killed the bill.
... ... it would have been ineffective against malware embedded in the operating system - by the OS developers or later black-hats.
The version of "open source" required doesn't give away any copyright or patent protection, or transfer rights to USE the code to others - especially the competition. (It does puncture trade secret.)
If the bill had instead required that only the voting software installed on the voting machines be open source
This is not a minor issue: With control of the US Government at stake a LOT of engineering effort can be profitably applied to attempts to compromise the system - by political, economic, or foreign governmental interests.
If this bill passes ... [Diebold] can then sell new machines to [all their former] customers.
Or printer and software upgrades.
If Diebold fixes the auditability problem I have no further gripe with the use of their machines. If buying an upgrade from them is 'way cheaper than replacing the machines outright, that's just dandy.
"If it's worth doing, it's worth doing at a profit."
Because:
- The voting machine result can be used whenever the particular precinct's results are uncontested, leaving all the advantages (except the "advantage" of being able to invisibly rig an election) intact.
- The auditability of the result will virtually eliminate any utility in rigging the machines (while bringing to bear draconian penalties for attempting to rig them and getting caught), greatly improving the reliability of the machines' results - to the point that they CAN be used without creating the appearance of rigged elections.
Near the end of TFA is the important reason for them to get their support from Oracle:
They're running Oracle's application server and database on some of the mission-critical servers.
With Oracle supporting both the application and the OS under it there's no time and money lost to finger-pointing when something gets hosed. Instead a single team buckles down and fixes it immediately.
(Presuming they ever need service. One of the comments from Red Hat indicates that they may never have actually had to USE the service contract. Take THAT, Microsoft! B-) )
PB can buy all the Sealands they want; Hollywood and the rest will just sue the ISPs for providing the bandwidth.
Given their budgets they could also hire some mercenaries and mount an attack on Sealand themselves.
If they filmed it they might make a profit on it, too.
Copyright (c) 2007 by me writing as "Ungrounded Lightning Rod".
Leave a followup to any posting in my journal with a firm offer if you want to do the movie. Otherwise I may sue for copyright infringement if such an attack is made, filmed, and the film shown for profit - even on a news operation under the same umbrella corporation. B-)
you can usually survive 1 non-matching transfusion, even if it's incompatible. After that your immune system is 'primed', much like for a vaccination, and will attack a second transfusion with extreme prejudice, likely resulting in your death.
Which brings up the issue of whether the body would develop an immunity to the enzyme, potentially producing a fatal anaphylactic reaction upon a future transfusion.
On the other hand, if the enzyme remains in the serum rather than attaching to the red cells the reaction would not produce the fatal clumping. Meanwhile the allergy to the enzyme, even if severe, could be handled by other drugs...
Which would also suppress the immune system somewhat - in a hospital "superbug" environment. So artificial type-O will likely remain an emergency measure, and type-O donors will remain in demand.
Why not use a program like http://www.truecrypt.org/ to hide your data if your going to download illegally.
Because once you're subpoenaed and served a search warrant you'll have to cough up the keys or you'll be in contempt of court.
Forgot or lost the key? Go to jail. Stay there until you remember it.
That's the whole point of April Fool's day. To remind you to be skeptical.
I thought it was to confine the pranksters to a single day, to keep them from pulling such junk intermittently all year.
Part of the tradition is a little rhyme used on anyone who pulls such a prank on April 2:
"April Fool's come and passed.
You're the bigges fool at last!"
The problem with OMGPonies-day at /. is that I begin to doubt even pleasant, plausible news by proximity ...
Back in the days of the Cold War I occasionally thought that the ideal (for them) time for the Soviet Union to launch a nuclear attack on the US would be early morning on April 1. Response would be slowed (except, perhaps, in the military itself) by the concern that somebody was pranking. Now that it's "War on Terror" time, ditto for the terrorists.
Once such an attack, with a large body count, is (successfully) made, pulling an April Fool prank MAY become sufficiently taboo to put the brakes on the madness. Unfortunately, I can't imagine anything short of that which might slow it down.
Meanwhile, I've stopped even bothering to read Slashdot on (or near) April 1. When the fools are having the fun the site is utterly useless to me.
In some locales, you can't legally call yourself an "engineer" unless you have a PE to your name, much like you can't start working as a doctor or lawyer without appropriate paperwork.
In most states calling yourself a software or computer engineer is an explicit exception to such requirements. It's a job title, not a ceritification. (I understand one state does require some certification to use the E-word even for computer-related stuff.)
This is largely because the industry has a track record of rejecting such regulatory attempts as stifiling of progress and monopolistic. There was one serious attempt to create such a cert, and if it had succeeded you'd have to be a Cobol expert and understand JCL to work in the field now. Instead, certifications are now industrial internal matters, issued by profit-making corporations and worth the paper their printed plus the reputation of the company's certification operation.
Given that some of the best and brightest - including many of the pioneers - didn't even get a degree, it's not surprising. During the early days the experts were the ones who became so obsessed with working/"playing" with the machines that their other schoolwork often suffered, and many of the best ended up dropping out short of a degree. Meanwhile the early coursework in the field tended to be a sidetrack from what was really needed by industry, and for at least the first ten years of computer science education a 4-year "computer science" degree was actually a hinderance to employment in industry. (You'd think the professors thought everybody was going to write his own compiler, just for starters.) The graduate had to "unlearn" a lot of stuff before he could really pick up on what the employer wanted done.
Minsky divided the first thirty years or so of computer science education into three periods:
1) 4-year degree was counter-productive, as above. Faculty didn't really understand what was needed.
2) 4-year degree was actually useful. Faculty had rented a clue about what was needed.
3) 4-year degree was counterproductive again. Faculty was teaching excessive formalisms and software fads.
(It's been long enough for maybe two more "periods" now but I haven't heard whether Marvin has characterized them. B-) )
Radio waves are alternating magnetic fields. The faster the field is alternated the more power it has and the further it travels.
Actually this is wrong. At least on the 'further it travels' aspect.
Also the "just magnetic field" part, (as explained by another poster above.)
And also the "faster the stronger" part.
The rate of alternation is just the frequency of the wave, and has nothing to do with power.
The power for a given waveform and polarization is proportional to the square of the peak field strength. (For an arbitrary waveform it's actually the integral of the magnitude of the cross product of the electric and magnetic field vectors...)
mmmm.. quantum mechanics..
Perhaps more like "maybe a classical explanation is enough, assuming quantum-mechanical approaches are necessary make you miss some useful stuff and work too hard at it".
Or as "classical" as you can get with electromagnetism anyhow...
If you don't welcome us properly we'll get in our longships and row our tall, blond asses over there and.... um.... ... start buying up your internet infrastructure manufacturers.
Oh, wait. We're already doing that.
Countries were judged on ... the extent to which government policy creates a framework ...
In other words: Government meddling in the economy and industry gives a "better" score, hands-off gives a "worse" one.
For instance:
- The government-mandated pre-divestiture ATT monopoly was "better" than ATT+Sprint+MCI+... and the "seven dwarves" CLECs.
- Government building the internet infrastructure is "better" than a gaggle of backbone peers and a herd of mom-and-POP ISPs.
- Government picking (and subsidizing) particular hardware and software vendors is "better" than letting them fight it out with each other (and a crowd of of FOSS teams with a handfull of licenses and ideological battles).
Look how well (terribly) that last one worked in Japan.
Do YOU really think The Cathedral is "better" than The Bazaar?
... why not run a test current through it to measure the conductivity instead of using an accelerator?
She did.
But hooking up molecule-sized test prods to an ohmmeter was a pain.
So she used a particle accelerator to inject the electrons. (TFA doesn't say what else she used to measure the current.)
I've contemplated using scanning electron beams for electrical measurements. Say: a low-energy electron beam for the negative supply, a high-energy one (creating more secondary electrons than injected electrons) for the positive, and a third one at an energy that turns it back around near the surface (or gets sucked in, depending on voltage) for a voltage probe.
But that's both too large and too energetic for testing single molecules of plastic.
Going the other way and using a particle accelerator to excite some observable side-effect of conduction is quite the hack. (I'd propose giving her an award but her university already did. Waytago!)
I don't understand why it's necessary to offer a copy of their hard disk to the RIAA representatives as evidence of innocence. If they're essentially accusing pseudo-random people of piracy, then isn't the onus on them to prove it? (I realise that in civil cases it's balance of probability, but even so...)
If I read that right:
- The original disk became subject to discovery in the case.
- So they bought a replacement and made a copy of it to continue running on and stored the original.
- They offered to let the RIAA's agent, under supervision, buy a fresh disk (with no deleted music files already on it) and make their own clone of the original.
- And meanwhile, because the defendant had to buy the new disk he's now running on (so the old one could be preserved for RIAA's benefit) they want the RIAA to PAY for that one, too. B-)
Ok. I'm as much of an open source advocate as anyone, but I'm not sure I see what all of the hubbub is about or believe the proposition that this upgrade should be free.
Company makes a box that happens to run linux as the base OS. They should therefore redistribute any changes they make to the GPL'ed code they run. That I get.
Not changes.
They must redistribute THE FULL SOURCE, even if it's NOT changed, of whatever they're loading.
For (essentially) free.
Otherwise they have no right to do the copying involved in loading it.
The right to load it comes only from the GPL's terms, which require you make available to the end users the ENTIRE source if you distribute the object - regardless of whether it has been modified.
They went out of business, and they let people who were former subcontractors give away/sell the information needed to update the system so the end user can continue to use the hardware in some fashion.
And the "former subcontractors" ALSO are under the terms of the GPL. So if THEY have GPLed firmware to download in those dongles, THEY must also release the source to THAT firmware. (Not just if they modified it, etc.)
If they're selling the dongles for $30 they have to sell the firmware source for no more than $30 also.
Meanwhile, the previous company distributed the object so they are obligated to distribute the source. If somebody (like the "former subcontractors") bought that part of the operation, they also bought the liability for distributing the source that matches the object already distributed - which they must satisfy to keep the license rights alive. Otherwise they don't have the rights to the modified code.
If USDTV actually went bankrupt it might be argued that the obligation to distribute source to the previous version was extinguished in bankruptcy. But IMHO that doesn't extinguish Cable Communications' obligation to distribute source when THEY upgrade the firmware. If they do this for free they must distribute the source for free - if they do it for $30 they must distribute the source for no more than $30.
Also: The obligation to distribute the source can not be limited to the original purchasers of the box unless the source is distributed WITH the box. Once even one instance of the box has been through first sale without the source the GPL requires the source to be distributed to anyone on demand.
Rand? Seriously? Shittiest self-styled philosopher and author ever to put pen to page. This passage is about as deep and insightful as "a is a," and about as well written as "My Pet Goat."
You misunderstand Rand's works.
She writes in the form of the Russian Epic Novel.
As a friend of mine once remarked: "Nobody really understands the purpose of the Russian Novel - but it's not entertainment."
This quote might be relevant. But the rest of her shitty books and philosophy are not.
Depends on what subject you're measuring relevance against.
Objectivism has a very important use: It explains, in connected logic understandable by even moderately intelligent psycopaths, how obeying some basic rules of civilized behavior are profitable for them personally.
Teaching Objectivism is the ONLY (non-religious) "treatment" that has a track record of reforming a significant fraction of hardened criminals.
I bought a car about 6 months ago, and I ran into this. Even though I was paying straight cash, they demanded my SSN so they could run a credit report. They said that they needed to run a credit report to see if I was on this list. I argued with them that I was paying cash, they didn't needed my SSN, etc., but it was late and I eventually relented just to keep the process moving.
You should have taken the unsigned contract (with the model and price info) and walked out.
It's amazing how fast that turns sales people around.
And if they don't do a 180, just go to another dealer, ask for the manager, tell them what happened, and offer to buy the car for cash from him at the same price as you'd have paid the other guy. (For him it's a better deal: HE gets the sales commission and doesn't have to waste his and a salesman's time.)
Since the tweaked sheep are chimeras, and humans/sheep don't hybridize, mating two of them (from two different donors for the human cells), if it produced anything at all, would produce either fully-sheep or fully-human offspring.
Human offspring seem likely, since the part of the sheep replaced/augmented with human stem cells was the part that produces the internal organs - presumably including the reproductive tract. (Can somebody more familiar with embryogenesis check me on that?)
First generation might be premature or have nasty birth defects due to being raised in the mostly-sheep chimera's body, with most of its biological feedback mechanisms "programmed" to produce lambs and some to produce humans. Second generation would likely be fully normal people or sheep.
These sheep are not 15% human, there is no such thing - they're 15% antigenically identical. There is no percentage at which they will become human, because their basic structure is still of a sheep!
These are not hybrids or the result of genetic engineering. They are chimeras: Organisms composed of two separate clone of cells. Some of the cells are 100% human, some are 100% sheep. The total animal has 15% of its cells being 100% human (surface antigens and all), not 100% of its cells having 15% human traits. (The immune system matured in the presence of both so it doesn't attack either.)
The cells are in coherent lumps, too. Entire organs - including the brain - may be 100% human tissue. (Though they may be morphologically similar to the sheep equivalent because they were exposed mainly to sheep growth factors while forming.)
(emulation will only require MORE computing power than the actual Windows box)
Use WINE for any applications it supports (which is most of 'em these days).
The acronym stands for "Wine Is Not and Emulator". It doesn't "emulate" a machine running Windows.
What it fakes up is the API: The application runs at full CPU speed while the system and library calls are translated from those appropriate for Windows to the Linux/X equivalents or are supported by native replacement routines.
So the application itself is running at the same CPU speed. Any bets on whether Linux and X (even with system call translation) or Windows take longer to get things done when the application needs support? Or whether the Windows or Linux scheduler gives the application more CPU?
Let's assume someone keeps his machine 5 or 6 years with linux installed on it and suddenly something give, like the cpu or the memory, now how much would it cost for that user to get the same cpu to run on his machine? We assume here it's a server and not some home usage, How much do you think it's going to cost him to find replcment parts?
That happened to me a year or so ago.
We'd gotten two Compaq AMD-based machines just before the Y2K transition. I installed RedHat on one of 'em and kept the other for a spare. Foolish person that I was, I left the spare in the box, rather than pull it out and check it.
Needed to install the spare for a workstation (old one is still running fine - except for video card - more later). Pulled it out and tried to boot it. DOA. (Oops!) A little part-swapping with the operating one showed the problem to be the CPU.
Compaq had spare CPUs but still wanted the old price of $700 for 'em. So I went down to Weird Stuff Warehouse. They didn't have exactly the same version so I bought a slightly faster one for something in the two-digit-dollas range. (Had to try two of 'em - first one was TOO fast for the MOBO's RAM so I had to downgrade to one that was only a LITTLE faster. They have a "try it and swap it" exchange policy on such stuff. They also priced 'em by speed, so I got a partial refund.)
(Turns out the old spare-box wasn't up to the performance level needed for the software that was to run on it. So we bought a new machine for that anyhow. Thus the repaired machine is back to being a backup spare (though the person it was intended for has just expressed an interest in using it for a "try out Ubutnu" machine before migrating the newer one to Linux or dual-boot).)
A few months before the above discovery, the RAM on the video card in the in-use (for half a decade) machine had started to flake out intermittently. (Probably heat from not vacuuming out the dust over that long period.) So I got a replacement video card while I was at it. (About $10. Later revision of the same card, worked just fine. Total cost: Well under $100 plus two trips to the store.)
Point is that "trailing edge technology" spares are often available and quite cheap - if you're willing to accept used parts from scrap dealers. (If it's a set of mission-critical machines, of course, you have to be prepared to migrate when the new-part supply chain has shut down and your stock of working spares is getting thin.)
The big problem arises if you want to expand your deployment significantly after the parts have been end-of-life-ed. THEN you can't get what you need and need to move to something more recent, to avoid problems with support.