APC UPS's have a tendency to cook their batteries as they get near the end of their lifetime. The results can be horrifying... bulging batteries, and if allowed to go on long enough, yes, even "sealed" lead acid batteries will rupture and you'll get the lovely sulfur smell.
I recently pulled these APC batteries out of an APC Smart-UPS 1400, which had to be disassembled (including the removal/replacement of rivets) in order to get the batteries out.
http://img221.imageshack.us/img221/171/imageyv.jpg
Do you really mean to imply that throwing a few table scraps at a starving person is a fix for their ongoing hunger?
Little band-aid fixes are not substantive health care reform.
Fundamentally, capitalism suggests a medical care model that looks a bit like what we have now. The problem is, taken to its ultimate conclusion, you wind up with a system where the rich have cadillac care and the poor have no care. We've tried to "fix" that with hospitals that do not turn away anyone from the ER, for example, but that merely shifts costs around.
Any health care system that is designed to promote the health of the society in general will ultimately have aspects that one can point to and scream "socialism!" - even our current one does: if I go to the hospital, part of my bill is going to cover the hospital's costs for free ER services to those who cannot afford it. It's socialism, just disguised under layers of capitalism and charity.
As a moderate, I get ticked off by both sides of this debate. I've had Republicans try to paint me as "more liberal than Obama" for trying to discuss rational reasons for national health care and ways to pay for it, including such radical concepts as moving away from procedure-based billing, but then I've had Democrats try to paint me as a cold, cruel monster for suggesting that maybe we cannot have a national health system that covers people without any limitations, or that there must be compromises, etc.
The reality, folks, is that we're able to keep people alive, healthy, longer. When Grandpa used to get to 60 and had a heart attack, 50 years ago, his remaining life expectancy was horrible compared to the same scenario today... but that progress comes at a cost. Medications, tests, doctor visits, these things all cost money. If he used to pass at 65, but now manages to hang on 'till 75, that's 10 more years of sometimes-expensive treatment to keep him around. This is a reality we have to come to grips with.
Instead of trying to understand and cope with that reality, we have the Republicans screaming "Death Panels!" and the Democrats trying to jam their heads into the sands about the real issues that cause health care costs to grow.
I don't expect to be able to have a rational and open-minded discussion with anyone on this topic anymore. The few times it has happened, it's been very refreshing. So many people have made up their minds and aren't willing to consider all the angles to this problem, I'm just disgusted.
It's nothing like the health care bill we should have had, something to create a health care system comparable to other modern countries. The Democrats have no backbone and kept watering it down and morphing it until it was only vaguely acceptable to just barely enough of them to possibly pass. This sort of thing leads to awful legislation.
The Republicans, of course, are chanting "wait, wait, this is being rushed," but the facts are that they had years in which they could have pushed through health care reform - years where it was clearly necessary. Despite what they say, your average Republican simply doesn't believe in health care reform, which is why it didn't happen under Clinton and wouldn't happen under Obama if they could figure out a way to delay it. So instead of pushing for a fiscally responsible and conservative health care reform, the Republicans are really pushing for the status quo, without trying to seem like they're doing that.
Both parties stink. I'm kind of hoping this passes, but then the Republicans come into power. It'll be impractical for them to repeal this, but perhaps they'll be smart enough to tinker with it to make it better. Past history is not encouraging, though.
We build servers here. First rule is that if you hurry, you're likely to screw something up and end up needing to fix something down the road. The hour of downtime and cost of replacement parts later is not worth any up-front savings in time.
Building a server starts with a bare metal chassis (not just plugging in a few snap-in components to a prebuilt). Integrate your power supply, install system board, from there of course it gets dependent on the specifics... but if you're not spending a fair bit of time trying to be neat, tying your cables down, thinking about how to route the cables for airflow, custom-building cables to fit odd needs, tightening screws with the right torque, and all of that, then you're not really doing a good job. Rack mount servers are especially sensitive to the skills of the designer and builder.
Your rack mount server will end up in a rack somewhere where it may not be pulled out for another five years. There may not be a "second chance" to fix it if you rush through building it - it becomes someone else's problem, perhaps, but they won't be thanking you for it.
This won't stop the paranoid from preventing their children from being immunized because some of these same people have interesting theories about how the vaccines are deliberately nefarious in other ways (going as far on out there as mind control, etc). These people and their little theory have done more to damage public health in a short amount of time than a lot of other things...
I-Pass is accepted on EZpass toll roads, so one alternative might be to pick up an I-Pass instead. I don't know if there is any surcharge for out-of-network tolls at this point, though I believe they've been discussed.
At least in Illinois, the last two times I happened to check, the gas prices were competitive. I just checked right now. Gas at the Deerfield Oasis is currently $2.85 (GasBuddy, 4 hours old) while a lot of the stations on US-45 a little to the west are actually $2.89, with a few $2.82's nearby. It's not always like that, but it's been a long time since I've noticed a severe price difference.
In Illinois, the I-Pass system (linked to your EZpass, IIRC) was implemented and to encourage adoption, they doubled the cash tolls. Drivers from surrounding states who do not have an I-Pass end up paying a lot more.
Of course, the Illinois Tollway is a great example of a system where a toll was instituted to pay for the highway, and then not only did it not go away, but they've actually raised the tolls (several times I believe).
They want to determine coverage. You cannot derive street-level coverage of broadband from IP addresses easily. As it stands, one of the problems with broadband is that you do not get universally consistent coverage, for example, at home, the 3/768 DSL offering of one of the CLEC's failed testing and they provisioned it for 1.5/512 instead. Had we been half a mile closer to the CO, 3/768 likely would have worked. There will be someone else a little further out who can only get it as 768/384.
The real problem will be for the FCC to get enough people to run this to get a meaningful map. I doubt that they'll get enough for it to really matter.
An iPhone (yes there's an iPhone app) test and a laptop test on the same wifi reported wildly different numbers.
Selecting a server 800 miles away rather than the one in the same city yielded much improved numbers (by whole number multiples).
Speedtest.net already has an extensive database, and appears to be part of the backend of this. It's too bad the FCC couldn't have just handed them a small pile of cash to summarize the existing data, which would probably have been better at rapidly producing results.
I keep having this vision of Cmdr. Data on that episode of TNG where he suddenly appears on the bridge with a rapidly flashing palm light. Maybe this will not only do our networking, maybe it'll be able to fix brainwashing too. (And if it can fix it, betcha someone can hack it to DO the brainwashing too...)
No, I didn't. In context, the point is that "we" are not really a party to these negotiations.
If I had said "citizens," it might have caused some confusion by those who maintain that government represents the citizens - which confuses the point I was making. By choosing "consumers," a party that might legitimately bargain with businesses, I was pointing out specifically that the "we" used in the previous message was irrelevant; "we" (consumers/mere citizens/lowly peasants/whatever) are not party to these negotiations.
The previous message seemed to suggest that "we" are party to these negotiations and could get what "we" wanted.
In theory, government is supposed to represent our best interests, but in practice, that does not always happen. In particular, the secrecy surrounding ACTA should make it exceedingly clear what sort of value the US government places on input and participation by its citizens in this matter.
If there was a secure way to make that happen, I'd agree. However, how do you determine who is trustworthy enough to supervise the process? Do they supervise the entire process (empty box -> add votes -> count votes -> report votes) or just a part of it? If you have two, three, four political parties, do you have observers from each party? What happens when votes from a precinct just "disappear" on the way to whereever they're counted (or stored)? How many people need to be subverted in order to corrupt the vote?
The potential for electronic voting is that there are clever ways to avoid lots of problems. Lost votes? Submit them in realtime over the network, and keep an electronic copy and paper copy at the precinct for security. With copies held in more than one place, loss becomes much more difficult. Verifiability? There's been a lot of work done on that as well, but a bit more complex than I can summarize in a sentence. There are also lots of new possibilities for fraud; anyone reading Slashdot knows that, but these can (and should) be addressed.
I wish that "e-voting" meant "working towards a technologically superior method for ensuring the integrity of a vote", but it mostly seems to mean "a way to replace the behemoth mechanical voting machines with 10 columns of 40 levers each" that some of us grew up with. To that end, I don't really care if the platform is XP or Linux, most deployed electronic voting systems are probably just about as corruptible - maybe just in different ways.
No, I already *read* the article and I was commenting on what was *missing* from the article.
Perhaps *you* should have read the article before making yourself look foolish. Had you actually read the article, then my comment, then you would have known that I, too, was essentially saying that the Linux thing is only one small part of the bigger picture.
Yes, after all, desktop deployments are not really the sort of thing you trumpet about with press releases, etc. Who knows how many large-scale Linux desktop deployments there are.
I'm more curious about what's running on top of Linux, though. Any free software OS (Linux, FreeBSD, etc.) is going to be great simply because it'll save the taxpayers licensing fees. However, as we've discovered here in the US, it is usually the voting software itself that is problematic.
The Linux thing is nice, but it'd be more meaningful to me to say "We've deployed new e-voting machines. They're Linux based devices running verifiable voting software, which we gave sample units of to each of our top tech schools, and none of them were successfully subverted." Or something line that.
We (the consumers) aren't really a party to the negotiation; the government is negotiating on our behalf to work on legitimate problems such as counterfeit goods. The real parties to the negotiation are businesses and government. As such, the businesses are pushing to get all the things they'd like to see, even where they're not really in the interests of the public. Government is dazzled by the show, and will tend to go along with a lot of things, especially where the businesses have been successful with propaganda.
It's all about that growth number. As cell phone penetration is slowly reaching the saturation point, the choices for pushing growth are decreasing.
Since the number of potential subscribers in the country is growing fairly slowly, and since corporations cannot (yet?) manufacture new consumers on a production line, the only real option is to find a way to increase the revenue per subscriber.
That used to mean selling minutes. Then it was texting. Then data. Then additional wireless devices. There always has to be a new gimmick, and of course many people will pay up.
The problem is that eventually the charges become outrageous and people stop buying. Growth will hit this wall eventually.
I don't see it as viable to pay for a $60 data plan for the laptop, a $30 data plan for the iPhone, and then another $30 for iPad. As you say, most people don't use two devices at the same time, and in fact the availability of more devices tends to mean each individual device gets used somewhat less than it would otherwise.
I'm sure the dozens of other maintenance items we addressed during the site visit precluded us from sitting on our asses for an extended period of time.
Normally the drive is about 14 hours, and yes, we often drive because it allows us to carry gear and supplies. To arrive on time for a 2AM-6AM maintenance window, we usually leave midmorning, get there around midnight, set up, work like hell through the script of tasks, and then go have breakfast around 7AM prior to the return drive.
If you stick a card in a machine and it doesn't work, you might suspect:
1) Bad cable 2) Bad card 3) Driver problem
If you stick two cards in two identical machines and both exhibit the same problems, you are likely to conclude that the cables aren't bad (especially after a swap), that it's very unlikely for two cards to be defective, and therefore it is probably a driver issue.
Now, in the extensive 20 or 30 minutes that we would have had in slush time in the downtime schedule, we did try to debug the problem, but at a certain point, you draw the conclusion that the OS either needs to be upgraded to a newer version (something that can be done from remote) or that there's a driver problem (something that can be debugged from remote). The idea that cards purchased from a well-known vendor would be counterfeit probably never comes up.
I'd say you have an unreasonable idea of how the data center world works. It's simply ridiculous to think that we'd travel 800 miles in person just to install a few cards; paid smarthands are capable of taking care of that sort of thing. In reality, it does not make financial sense to visit a distant data center without a boatload of work to do. It can be physically intensive, is invariably tiring physically, is demanding mentally, and isn't really the sort of thing where you get to sit "on your ass for an extended period of time."
We wouldn't have left it in if we felt it was questionable. We're frequently debugging various quirks and problems with bleeding edge hardware platforms and it isn't particularly odd for non-commodity "unusual" hardware such as a dual gigabit ethernet card to require some adjustments from the device driver's author in order to work. Some of us actually fund that sort of thing. After all, the cards were acquired from NewEgg, which most people in this thread seem to feel is a fairly reputable vendor. They were boxed cards that came in an Intel box that appeared identical to past products. There was no particular reason to suspect that the cards were "questionable" rather than suffering a driver issue. Had the cards actually caused the servers to fail, there's a reason that we have multiply redundant servers and a smarthands agreement. There's no "lesson" to be learned there.
As far as lawyers go, perhaps you didn't realize that an easier method is to have the credit card company do a chargeback. I believe that's what our purchasing dep't did. Compared to what a lawyer might charge, it's free and easy. They could try to sue us for the money, of course, but then they'd be in the position of having to explain the counterfeit.
When we couldn't get them to work on-site, it looked like maybe a driver issue, so we left them in, so we could twiddle with it from remote. Staying on-site just to debug a driver issue is stupid; it costs several hundred dollars per day in expenses and OT. However, once we got back, we still couldn't get it to work.
In other words: We were never happy, they never worked.
So we did pull the cards at the very next chance we had, but it was a few months later. Having bought from NewEgg, a company we had had good experiences with previously, we didn't expect that the cards might be counterfeit. We expected to find that it was a driver problem (in which case it'd eventually be resolved) or possibly a PCI compatibility issue (in which case the cards would still be good). Had either of those cases been true, we would have been fine with keeping the cards, even though we couldn't use them for their intended use.
We had a problem with some Intel dual port gig cards and NewEgg about five years ago. We had ordered them shipped directly out to the data center (800 miles away) because of an urgent need, but upon installation, we were seeing various odd problems. Suspecting a driver issue, we left them in and returned home. Never resolved the problems. Pulled them a few months later on the next visit. Further research showed that they were phony Intel cards. Apparently several resellers had been hit with these. However, NewEgg maintained that it was no longer their problem because of their return policy (30 days, maybe, I don't remember), and refused to make an exception for goods that they knew were knock-offs. This was really too bad, as we started buying less stuff from them after that.
Slashdot Mirror
Computers run on smoke... when the smoke comes out, they stop running.
APC UPS's have a tendency to cook their batteries as they get near the end of their lifetime. The results can be horrifying... bulging batteries, and if allowed to go on long enough, yes, even "sealed" lead acid batteries will rupture and you'll get the lovely sulfur smell.
I recently pulled these APC batteries out of an APC Smart-UPS 1400, which had to be disassembled (including the removal/replacement of rivets) in order to get the batteries out.
http://img221.imageshack.us/img221/171/imageyv.jpg
Do you really mean to imply that throwing a few table scraps at a starving person is a fix for their ongoing hunger?
Little band-aid fixes are not substantive health care reform.
Fundamentally, capitalism suggests a medical care model that looks a bit like what we have now. The problem is, taken to its ultimate conclusion, you wind up with a system where the rich have cadillac care and the poor have no care. We've tried to "fix" that with hospitals that do not turn away anyone from the ER, for example, but that merely shifts costs around.
Any health care system that is designed to promote the health of the society in general will ultimately have aspects that one can point to and scream "socialism!" - even our current one does: if I go to the hospital, part of my bill is going to cover the hospital's costs for free ER services to those who cannot afford it. It's socialism, just disguised under layers of capitalism and charity.
As a moderate, I get ticked off by both sides of this debate. I've had Republicans try to paint me as "more liberal than Obama" for trying to discuss rational reasons for national health care and ways to pay for it, including such radical concepts as moving away from procedure-based billing, but then I've had Democrats try to paint me as a cold, cruel monster for suggesting that maybe we cannot have a national health system that covers people without any limitations, or that there must be compromises, etc.
The reality, folks, is that we're able to keep people alive, healthy, longer. When Grandpa used to get to 60 and had a heart attack, 50 years ago, his remaining life expectancy was horrible compared to the same scenario today... but that progress comes at a cost. Medications, tests, doctor visits, these things all cost money. If he used to pass at 65, but now manages to hang on 'till 75, that's 10 more years of sometimes-expensive treatment to keep him around. This is a reality we have to come to grips with.
Instead of trying to understand and cope with that reality, we have the Republicans screaming "Death Panels!" and the Democrats trying to jam their heads into the sands about the real issues that cause health care costs to grow.
I don't expect to be able to have a rational and open-minded discussion with anyone on this topic anymore. The few times it has happened, it's been very refreshing. So many people have made up their minds and aren't willing to consider all the angles to this problem, I'm just disgusted.
It's nothing like the health care bill we should have had, something to create a health care system comparable to other modern countries. The Democrats have no backbone and kept watering it down and morphing it until it was only vaguely acceptable to just barely enough of them to possibly pass. This sort of thing leads to awful legislation.
The Republicans, of course, are chanting "wait, wait, this is being rushed," but the facts are that they had years in which they could have pushed through health care reform - years where it was clearly necessary. Despite what they say, your average Republican simply doesn't believe in health care reform, which is why it didn't happen under Clinton and wouldn't happen under Obama if they could figure out a way to delay it. So instead of pushing for a fiscally responsible and conservative health care reform, the Republicans are really pushing for the status quo, without trying to seem like they're doing that.
Both parties stink. I'm kind of hoping this passes, but then the Republicans come into power. It'll be impractical for them to repeal this, but perhaps they'll be smart enough to tinker with it to make it better. Past history is not encouraging, though.
We build servers here. First rule is that if you hurry, you're likely to screw something up and end up needing to fix something down the road. The hour of downtime and cost of replacement parts later is not worth any up-front savings in time.
Building a server starts with a bare metal chassis (not just plugging in a few snap-in components to a prebuilt). Integrate your power supply, install system board, from there of course it gets dependent on the specifics... but if you're not spending a fair bit of time trying to be neat, tying your cables down, thinking about how to route the cables for airflow, custom-building cables to fit odd needs, tightening screws with the right torque, and all of that, then you're not really doing a good job. Rack mount servers are especially sensitive to the skills of the designer and builder.
Your rack mount server will end up in a rack somewhere where it may not be pulled out for another five years. There may not be a "second chance" to fix it if you rush through building it - it becomes someone else's problem, perhaps, but they won't be thanking you for it.
This won't stop the paranoid from preventing their children from being immunized because some of these same people have interesting theories about how the vaccines are deliberately nefarious in other ways (going as far on out there as mind control, etc). These people and their little theory have done more to damage public health in a short amount of time than a lot of other things...
I-Pass is accepted on EZpass toll roads, so one alternative might be to pick up an I-Pass instead. I don't know if there is any surcharge for out-of-network tolls at this point, though I believe they've been discussed.
At least in Illinois, the last two times I happened to check, the gas prices were competitive. I just checked right now. Gas at the Deerfield Oasis is currently $2.85 (GasBuddy, 4 hours old) while a lot of the stations on US-45 a little to the west are actually $2.89, with a few $2.82's nearby. It's not always like that, but it's been a long time since I've noticed a severe price difference.
http://www.illinoisgasprices.com/map_gas_prices.aspx?z=8&lat=42.254700&long=-87.900024&ft=A&tl=48
Speedtest already makes this sort of information available, take a look under "World Results."
I'm not sure what you're seeing in their privacy policy that would prevent them from sharing the information.
In Illinois, the I-Pass system (linked to your EZpass, IIRC) was implemented and to encourage adoption, they doubled the cash tolls. Drivers from surrounding states who do not have an I-Pass end up paying a lot more.
Of course, the Illinois Tollway is a great example of a system where a toll was instituted to pay for the highway, and then not only did it not go away, but they've actually raised the tolls (several times I believe).
They want to determine coverage. You cannot derive street-level coverage of broadband from IP addresses easily. As it stands, one of the problems with broadband is that you do not get universally consistent coverage, for example, at home, the 3/768 DSL offering of one of the CLEC's failed testing and they provisioned it for 1.5/512 instead. Had we been half a mile closer to the CO, 3/768 likely would have worked. There will be someone else a little further out who can only get it as 768/384.
The real problem will be for the FCC to get enough people to run this to get a meaningful map. I doubt that they'll get enough for it to really matter.
An iPhone (yes there's an iPhone app) test and a laptop test on the same wifi reported wildly different numbers.
Selecting a server 800 miles away rather than the one in the same city yielded much improved numbers (by whole number multiples).
Speedtest.net already has an extensive database, and appears to be part of the backend of this. It's too bad the FCC couldn't have just handed them a small pile of cash to summarize the existing data, which would probably have been better at rapidly producing results.
I keep having this vision of Cmdr. Data on that episode of TNG where he suddenly appears on the bridge with a rapidly flashing palm light. Maybe this will not only do our networking, maybe it'll be able to fix brainwashing too. (And if it can fix it, betcha someone can hack it to DO the brainwashing too...)
Outsourcing doctors offshore.
No, I didn't. In context, the point is that "we" are not really a party to these negotiations.
If I had said "citizens," it might have caused some confusion by those who maintain that government represents the citizens - which confuses the point I was making. By choosing "consumers," a party that might legitimately bargain with businesses, I was pointing out specifically that the "we" used in the previous message was irrelevant; "we" (consumers/mere citizens/lowly peasants/whatever) are not party to these negotiations.
The previous message seemed to suggest that "we" are party to these negotiations and could get what "we" wanted.
In theory, government is supposed to represent our best interests, but in practice, that does not always happen. In particular, the secrecy surrounding ACTA should make it exceedingly clear what sort of value the US government places on input and participation by its citizens in this matter.
Hope that clarifies it...
If there was a secure way to make that happen, I'd agree. However, how do you determine who is trustworthy enough to supervise the process? Do they supervise the entire process (empty box -> add votes -> count votes -> report votes) or just a part of it? If you have two, three, four political parties, do you have observers from each party? What happens when votes from a precinct just "disappear" on the way to whereever they're counted (or stored)? How many people need to be subverted in order to corrupt the vote?
The potential for electronic voting is that there are clever ways to avoid lots of problems. Lost votes? Submit them in realtime over the network, and keep an electronic copy and paper copy at the precinct for security. With copies held in more than one place, loss becomes much more difficult. Verifiability? There's been a lot of work done on that as well, but a bit more complex than I can summarize in a sentence. There are also lots of new possibilities for fraud; anyone reading Slashdot knows that, but these can (and should) be addressed.
I wish that "e-voting" meant "working towards a technologically superior method for ensuring the integrity of a vote", but it mostly seems to mean "a way to replace the behemoth mechanical voting machines with 10 columns of 40 levers each" that some of us grew up with. To that end, I don't really care if the platform is XP or Linux, most deployed electronic voting systems are probably just about as corruptible - maybe just in different ways.
No, I already *read* the article and I was commenting on what was *missing* from the article.
Perhaps *you* should have read the article before making yourself look foolish. Had you actually read the article, then my comment, then you would have known that I, too, was essentially saying that the Linux thing is only one small part of the bigger picture.
Oops.
Yes, after all, desktop deployments are not really the sort of thing you trumpet about with press releases, etc. Who knows how many large-scale Linux desktop deployments there are.
I'm more curious about what's running on top of Linux, though. Any free software OS (Linux, FreeBSD, etc.) is going to be great simply because it'll save the taxpayers licensing fees. However, as we've discovered here in the US, it is usually the voting software itself that is problematic.
The Linux thing is nice, but it'd be more meaningful to me to say "We've deployed new e-voting machines. They're Linux based devices running verifiable voting software, which we gave sample units of to each of our top tech schools, and none of them were successfully subverted." Or something line that.
Is the software open source and based on verifiable voting, too?
We (the consumers) aren't really a party to the negotiation; the government is negotiating on our behalf to work on legitimate problems such as counterfeit goods. The real parties to the negotiation are businesses and government. As such, the businesses are pushing to get all the things they'd like to see, even where they're not really in the interests of the public. Government is dazzled by the show, and will tend to go along with a lot of things, especially where the businesses have been successful with propaganda.
Wasn't there a time when the US led the world in freedom, liberty, and openness?
I know I'm going to get horribly trolled for this, but damn it, it needs to be said.
It's all about that growth number. As cell phone penetration is slowly reaching the saturation point, the choices for pushing growth are decreasing.
Since the number of potential subscribers in the country is growing fairly slowly, and since corporations cannot (yet?) manufacture new consumers on a production line, the only real option is to find a way to increase the revenue per subscriber.
That used to mean selling minutes. Then it was texting. Then data. Then additional wireless devices. There always has to be a new gimmick, and of course many people will pay up.
The problem is that eventually the charges become outrageous and people stop buying. Growth will hit this wall eventually.
I don't see it as viable to pay for a $60 data plan for the laptop, a $30 data plan for the iPhone, and then another $30 for iPad. As you say, most people don't use two devices at the same time, and in fact the availability of more devices tends to mean each individual device gets used somewhat less than it would otherwise.
I'm sure the dozens of other maintenance items we addressed during the site visit precluded us from sitting on our asses for an extended period of time.
Normally the drive is about 14 hours, and yes, we often drive because it allows us to carry gear and supplies. To arrive on time for a 2AM-6AM maintenance window, we usually leave midmorning, get there around midnight, set up, work like hell through the script of tasks, and then go have breakfast around 7AM prior to the return drive.
If you stick a card in a machine and it doesn't work, you might suspect:
1) Bad cable
2) Bad card
3) Driver problem
If you stick two cards in two identical machines and both exhibit the same problems, you are likely to conclude that the cables aren't bad (especially after a swap), that it's very unlikely for two cards to be defective, and therefore it is probably a driver issue.
Now, in the extensive 20 or 30 minutes that we would have had in slush time in the downtime schedule, we did try to debug the problem, but at a certain point, you draw the conclusion that the OS either needs to be upgraded to a newer version (something that can be done from remote) or that there's a driver problem (something that can be debugged from remote). The idea that cards purchased from a well-known vendor would be counterfeit probably never comes up.
I'd say you have an unreasonable idea of how the data center world works. It's simply ridiculous to think that we'd travel 800 miles in person just to install a few cards; paid smarthands are capable of taking care of that sort of thing. In reality, it does not make financial sense to visit a distant data center without a boatload of work to do. It can be physically intensive, is invariably tiring physically, is demanding mentally, and isn't really the sort of thing where you get to sit "on your ass for an extended period of time."
We wouldn't have left it in if we felt it was questionable. We're frequently debugging various quirks and problems with bleeding edge hardware platforms and it isn't particularly odd for non-commodity "unusual" hardware such as a dual gigabit ethernet card to require some adjustments from the device driver's author in order to work. Some of us actually fund that sort of thing. After all, the cards were acquired from NewEgg, which most people in this thread seem to feel is a fairly reputable vendor. They were boxed cards that came in an Intel box that appeared identical to past products. There was no particular reason to suspect that the cards were "questionable" rather than suffering a driver issue. Had the cards actually caused the servers to fail, there's a reason that we have multiply redundant servers and a smarthands agreement. There's no "lesson" to be learned there.
As far as lawyers go, perhaps you didn't realize that an easier method is to have the credit card company do a chargeback. I believe that's what our purchasing dep't did. Compared to what a lawyer might charge, it's free and easy. They could try to sue us for the money, of course, but then they'd be in the position of having to explain the counterfeit.
Please read again.
They were stuck in machines 800 miles away.
When we couldn't get them to work on-site, it looked like maybe a driver issue, so we left them in, so we could twiddle with it from remote. Staying on-site just to debug a driver issue is stupid; it costs several hundred dollars per day in expenses and OT. However, once we got back, we still couldn't get it to work.
In other words: We were never happy, they never worked.
So we did pull the cards at the very next chance we had, but it was a few months later. Having bought from NewEgg, a company we had had good experiences with previously, we didn't expect that the cards might be counterfeit. We expected to find that it was a driver problem (in which case it'd eventually be resolved) or possibly a PCI compatibility issue (in which case the cards would still be good). Had either of those cases been true, we would have been fine with keeping the cards, even though we couldn't use them for their intended use.
We had a problem with some Intel dual port gig cards and NewEgg about five years ago. We had ordered them shipped directly out to the data center (800 miles away) because of an urgent need, but upon installation, we were seeing various odd problems. Suspecting a driver issue, we left them in and returned home. Never resolved the problems. Pulled them a few months later on the next visit. Further research showed that they were phony Intel cards. Apparently several resellers had been hit with these. However, NewEgg maintained that it was no longer their problem because of their return policy (30 days, maybe, I don't remember), and refused to make an exception for goods that they knew were knock-offs. This was really too bad, as we started buying less stuff from them after that.