Adding additional drives to a raidz vdev is not supported, no. Apparently it's a use case that is extremely rare in enterprise, which is where zfs was intended for.
This is largely what has kept me away from zfs, besides it not being in the mainline kernel.
If I had 300 disks then being forced to add/remove them in groups of 5 or so wouldn't be a big deal. When you have just a few disks at 90% capacity, being unable to add/remove them 1 at a time while keeping everything redundant is a much bigger problem (using an n+1 redundancy solution, not an n*2 solution).
One of the things I like about btrfs is that the design is more dynamic in this regard - you can have disks of varying sizes and add/remove them at will. Of course, anything more than raid1 on btrfs isn't production-ready yet (actually, anything on btrfs is only production-ready if you're willing to make some tradeoffs).
But isn't that the whole point, ZFS is designed to avoid the most common failure modes but it relies on reducing the errors in the data it is using for check summing. Thats why ECC is important, RAM errors are the 2nd most common bit flip error and if that's your comparator, it needs to be accurate.
Sure, but what good is having the RAM store the checksum correctly if the CPU calculated it incorrectly, or if the CPU compares the correct checksum to the identical correct checksum and determines that they don't match?
This is just about diminishing returns and where to draw the line. There are lots of things that can go wrong. Believe it or not people use systems that actually detect and correct CPU logic errors, and I'm sure the people selling them could tell you how often they detect errors.
I'd seriously consider ECC RAM in my next system in any case, but you can't just compare the cost of the RAM - if it requires buying a different CPU or motherboard you have to look at all the costs and tradeoffs.
Do you have a link. The last time I looked into this, you could not add a disk to a raid-z. You could add disks to a zpool, or add another raid-z to a zpool. However, a raid-z was basically immutable. This is in contrast to mdadm where you can add/remove individual disks from a raid5.
Google seems to suggest that this has not changed, however I'd certainly be interested in whether this is the case. The last time I chatted with somebody who was using ZFS in a big way they indicated that this was a limitation. He was using it for very large storage systems, and I could see how many of the ZFS features made it much more appropriate in these kinds of situations, especially with things like write intent log on seperate media, having many independent storage units which are individually redundant but otherwise behaving like a big array of disks (which helps to distribute IO which reduces some of the penalties with RAID), etc. I'm more familiar with btrfs and it seems to be evolving more towards being an ext4 replacement, where smaller arrays are the norm, etc. That isn't to say that many of the features on either aren't potentially useful for both.
adaptable to a large variety of mirror/striping/RAID configurations
"Adaptable" is a bit of a stretch here. If you set up a RAID on ZFS, you can't change it, you can only replace individual disks within it, or destroy the entire array.
That isn't a big deal if you're talking about a ZFS filesystem with a very large number of drives, but it is a big limitation for a small ZFS filesystem. That is, if I have 300 disks in 60 arrays of 5 1TB disks each, and I want to move to 3TB disks, then I just need to add 5 3TB disks, turn them into an array, add them to the filesystem, then remove 51TB disks, and then keep repeating this. On the other hand, if I have a motherboard with 5 SATA ports and I have 5 1TB drives in an array, then there is no easy way to replace those with 5 3TB drives one at a time and actually get use out of the extra space. You can do that with mdadm in any of its raid modes, and with btrfs in any of the raid modes that actually work (basically raid 0/1 now - it will work with raid5/6/etc but those aren't production ready).
Granted, a lot of people who are interested in ZFS are interested in using it for SAN/NAS/etc and this isn't likely to be an issue for them. For the average slashdotter with a few TB of drives in a RAID, it could be a problem.
The technical descriptions I've read say that you absolutely should use ECC because ZFS will eventually hit a checksum mismatch. This could result in valid data being flagged as corrupt. ECC RAM is not much more expensive these days but you do need a mobo that supports it.
You should also make sure that your system redundantly performs every CPU instruction like any decent mainframe would, since the same sorts of problems that can cause RAM to fail can cause CPUs to fail (and CPUs contain cache too).
That isn't a big problem, you just need a CPU/motherboard that supports this. IBM would be more than happy to sell one to you if you have a house you're willing to trade for it.
It has SOME features which btrfs has not yet implemented. Btrfs also has some features which ZFS has not yet implemented, including support for dynamically resizing a RAID (not adding/removing a RAID from a zpool).
Identical twins, born to the same parents, but adopted by different families, tend to have extraordinarily unlikely similarities in adult general intelligence scores.
Which is it - academic performance, or general intelligence scores? If the twin studies test one, and this genetic study tests the other, then I would think that this would make it very difficult to draw any conclusions by comparing the results, since academic performance and IQ aren't the same thing.
It shows that a large number of specific candidate genes don't do it. Even if it's not a complete refutation of the hypothesis, it is a push to maybe look elsewhere for some of the mechanisms of intelligence development.
Well, it only would show that those genes don't have an individually-detectable affect on whatever marker they looked at. Maybe collectively they have an effect that can't be detected statistically. Maybe the marker they chose doesn't make sense.
Imagine if I tried to identify the gene for "sickness." I took anybody who ever got sick for any reason and studied their DNA and looked for a common link. Most likely I wouldn't find anything. Would this prove that genes can't make you sick? Or is it more likely that "sickness" is such a broad description of a phenotype that it could have a billion different causes.
Academic performance could be the result of MANY factors. Physical attractiveness has been demonstrated to have an impact on academic performance, and you'd hardly expect the same genes to affect that as your ability to do some kind of mental processing. That is just picking one attribute that is obviously going to confound results. Then you get into stuff like whether intelligence is about persistence, or ability to process information, or memory, etc. All of those things are likely to affect academic performance. Then there are cultural factors - let's just assume the sterotype about Asians prodding their kids to study harder is true - I'm sure there are alleles more common in Asians (the fact that they have distinctive appearances makes this obvious for starters), and that is going to confound things.
If you really want to identify the genes responsible for a trait, you have to first come up with a very precise definition for the trait, ideally one tied to some kind of biological mechanism (good luck if it involves the brain).
The study did not demonstrate this at all. It simply failed to find specific genes responsible for intelligence.
You might still be born with a set intelligence which isn't genetically determined, or it might be genetically determined on the basis of genetic factors that were not identified for any number of reasons. Or you could be right.
The point is, this study doesn't provide any evidence one way or another.
Also, equating academic performance with intelligence may be dubious. There could be many factors responsible for academic performance, of which intelligence is just one. We can't even define what intelligence is...
Indeed, if they randomly distributed 2M surfaces for free throughout the population it would probably cost them about the same, and it would immediately create a demand for applications, it would get their product in front of consumers, etc. That would seem like a far better investment if you really wanted to blow this kind of cash.
Now you're quibbling. You can nitpick with me all you want but never try that shit on a clearance form. You wont like it.
And this is why the clearance process should be changed. I have no doubt that everything that happened to this individual happens to people all the time.
Just one more reason why nobody competent wants to work for the government...
It doesn't work on my Nexus 10 either. It works just fine on my Oneplus One. I can't imagine they'd have bothering to whitelist the OPO, so I'm not sure what the logic is.
No, the evidence is not the property of a US company. The evidence is the property of a US citizen and is being held by a foreign subsidiary of a US company and is, therefore, exactly like the hotel example.
No argument there. However, if I leave evidence in a US hotel chain's London hotel, a US court can certainly order the hotel chain to fly it back to the US and threaten to fine it until it complies.
I'm not arguing morality here - just practicality. If a guy walks into your house and manages to hold you at gunpoint and you're unarmed, as far as you're concerned, he's the boss and he gets what he wants until somebody else shows up and bails you out. As Augustine pointed out, pirates and emporers are the same thing on different scales, but in the case of the emporer there usually isn't a bigger fish who is going to come bail you out if you can hold out long enough.
I don't think that is it. Even a small phone tends to be in your pocket all the time.
The point of the smartwatch is that instead of having your phone vibrate to tell you to take it out of your pocket, turn it on, and navigate to a message, you can instead just look down at your wrist and see the message without touching anything.
If I got a lot of notifications at a time I wasn't already in front of my PC it would make sense. The problem is that 99% of the time the notification is just going to be something I could see in another browser tab on the PC I already have open in front of me. If I still worked in an office where I had a meeting every hour in a different conference room maybe having my calendar on my watch might be useful, but I just phone into all my meetings now. I feel like the smart watch is solving a problem that has already been solved in other ways.
So, your argument is that if Microsoft operates in North Korea, Iran, or any other country.. that those countries should also be able to force Microsoft to hand over any and all records on Americans if they see fit?
It isn't a matter of "should." The fact is that those countries can do anything they have the power to do - that is what makes them countries and not corporations. You'd be a fool to store your personal data with a company that needs to keep a country like North Korea happy if you don't want the North Koreans to have access to your data.
If you run a business selling swastikas, you'd be foolish to build all your infrastructure on servers owned by a company that counted on the German government for most of its business. At any time the Germans could ask them for a favor, and you're out of business.
You got that wrong, this is the US providing heavy stimulus for foreign companies creating their own cloud services. They're basically giving free reign to European providers, who do not own a single server in the US, and telling them to go ahead and dominate the market. Europeans should be thanking them, if anything!
They could only dominate the European market though. Those companies wouldn't be able to do business with companies in the US. If they did, then the US government would just threaten them until they gave in - just seize all their US bank accounts, or seize all their payments, harass employees if they travel to the US, etc.
It isn't as simple as you make it out to be. And it isn't just the US doing this sort of thing. Everybody wants the "Internet" to follow their local laws.
If I commit a crime and the evidence is in a Hilton hotel room in the UK nobody expects Hilton
Everybody however expects U.S. police to go the U.K. police and courts who will check if it was a crime in the U.K. in the first place then ask the U.K. Hilton to hand over the data, aka due process.
The problem here is that the evidence is the property of a US company. The US can't send in police to seize the evidence overseas, but it can punish the owner of the evidence until they voluntarily turn it over.
It is like refusing to hand over an encryption key - they may not have the ability to seize it without cooperation, but they can still beat you with a hose.
"We all work so that we can pay the bills... go on and do bigger and better things that 2BN dollars can provide."
Will the work on the things post-2BN also be to pay the bills? Seems like a contradiction. I wonder what proportion of people are actually not working to pay the bills.
Maybe 0.05%? One of the advantages of pocketing the $2B is that now you can take all the risks you want to and do the stuff you REALLY want to do. Maybe that makes you even more money, or maybe it doesn't, but who cares at that point?
Work would be a lot less stressful for me if I had $10M in the bank. I'd probably still be doing a lot of the stuff I'm doing today, though I might not be doing it for the same employers. I'd be more interested in getting good stuff done and less about whether my boss likes me, etc.
Well, in your analogy there is another big difference between IBM and Glaxo. In your example, if IBM does a faulty study that shows that a new way of making microchips is better, then the result will be that they spend a billion dollars on a new fab and it produces faulty microchips that nobody buys. In the other case, it is others who are relying on the results of the study to make a decision about whether to buy Glaxo's products. This is why it is much more important that drug trials be regulated. If it were up to me I'd have them all publicly conducted, but nobody wants to spend tax dollars on clinical trials.
There are lots of things that are messed-up about clinical trials, and the main reason for this is that they are VERY expensive to run. The problem is that you can't give somebody a drug unless their doctor is involved. Doctors make a lot of money, and clinical trials take a lot of time for them to participate in. So, if you want somebody who makes $500k/yr to spend 10 hours per week on a clinical trial, what is the only way to get them to cooperate? You have to pay them a LOT of money. Multiply this by thousands of patients, and then factor in that there are 14 cancer trials going on at the same time, so the doctor is going to pick the trial that pays the best to recommend to their patients, and now you have a bidding war on top of it.
Nobody wants to spend the money needed to either sponsor trials run by a disinterested party, or replicate trials.
(And yes, I do realize that some doctors are among the salt of the earth, and they will participate in trials because it is the right thing for their patients even if they were completely uncompensated. However, for every one of these there are probably 5 others that make it onto the FDA's debarrment list for getting patients to take experimental drugs that they shouldn't be taking just to collect extra fees. The list is posted online - just read it for yourself.)
Think about it. There are a million ways you can analyze any dataset. There are millions of datasets out there to analyze. There are millions of people who can independently decide to go back and do a re-analysis.
So, the issue is that if somebody goes back and does a re-analysis and the results are boring, nobody publishes. However, if the results are controversial, it gets published. Since there are so many permutations, you're guaranteed to find something exciting.
This is why you're supposed to establish your methods BEFORE you collect the data, and then stick to the methods you established to analyze the data. Otherwise your 95% confidence turns into a more realistic 1% confidence.
In practice, though, I'm sure the initial analyses are just as prone to this kind of problem. It just gets REALLY bad when you look backwards.
Slashdot Mirror
Adding additional drives to a raidz vdev is not supported, no. Apparently it's a use case that is extremely rare in enterprise, which is where zfs was intended for.
This is largely what has kept me away from zfs, besides it not being in the mainline kernel.
If I had 300 disks then being forced to add/remove them in groups of 5 or so wouldn't be a big deal. When you have just a few disks at 90% capacity, being unable to add/remove them 1 at a time while keeping everything redundant is a much bigger problem (using an n+1 redundancy solution, not an n*2 solution).
One of the things I like about btrfs is that the design is more dynamic in this regard - you can have disks of varying sizes and add/remove them at will. Of course, anything more than raid1 on btrfs isn't production-ready yet (actually, anything on btrfs is only production-ready if you're willing to make some tradeoffs).
Are you suggesting that if A = B, then B = A? :)
But isn't that the whole point, ZFS is designed to avoid the most common failure modes but it relies on reducing the errors in the data it is using for check summing. Thats why ECC is important, RAM errors are the 2nd most common bit flip error and if that's your comparator, it needs to be accurate.
Sure, but what good is having the RAM store the checksum correctly if the CPU calculated it incorrectly, or if the CPU compares the correct checksum to the identical correct checksum and determines that they don't match?
This is just about diminishing returns and where to draw the line. There are lots of things that can go wrong. Believe it or not people use systems that actually detect and correct CPU logic errors, and I'm sure the people selling them could tell you how often they detect errors.
I'd seriously consider ECC RAM in my next system in any case, but you can't just compare the cost of the RAM - if it requires buying a different CPU or motherboard you have to look at all the costs and tradeoffs.
Do you have a link. The last time I looked into this, you could not add a disk to a raid-z. You could add disks to a zpool, or add another raid-z to a zpool. However, a raid-z was basically immutable. This is in contrast to mdadm where you can add/remove individual disks from a raid5.
Google seems to suggest that this has not changed, however I'd certainly be interested in whether this is the case. The last time I chatted with somebody who was using ZFS in a big way they indicated that this was a limitation. He was using it for very large storage systems, and I could see how many of the ZFS features made it much more appropriate in these kinds of situations, especially with things like write intent log on seperate media, having many independent storage units which are individually redundant but otherwise behaving like a big array of disks (which helps to distribute IO which reduces some of the penalties with RAID), etc. I'm more familiar with btrfs and it seems to be evolving more towards being an ext4 replacement, where smaller arrays are the norm, etc. That isn't to say that many of the features on either aren't potentially useful for both.
adaptable to a large variety of mirror/striping/RAID configurations
"Adaptable" is a bit of a stretch here. If you set up a RAID on ZFS, you can't change it, you can only replace individual disks within it, or destroy the entire array.
That isn't a big deal if you're talking about a ZFS filesystem with a very large number of drives, but it is a big limitation for a small ZFS filesystem. That is, if I have 300 disks in 60 arrays of 5 1TB disks each, and I want to move to 3TB disks, then I just need to add 5 3TB disks, turn them into an array, add them to the filesystem, then remove 51TB disks, and then keep repeating this. On the other hand, if I have a motherboard with 5 SATA ports and I have 5 1TB drives in an array, then there is no easy way to replace those with 5 3TB drives one at a time and actually get use out of the extra space. You can do that with mdadm in any of its raid modes, and with btrfs in any of the raid modes that actually work (basically raid 0/1 now - it will work with raid5/6/etc but those aren't production ready).
Granted, a lot of people who are interested in ZFS are interested in using it for SAN/NAS/etc and this isn't likely to be an issue for them. For the average slashdotter with a few TB of drives in a RAID, it could be a problem.
It is basically the non-BSD, non-GPL version of btrfs. :)
The technical descriptions I've read say that you absolutely should use ECC because ZFS will eventually hit a checksum mismatch. This could result in valid data being flagged as corrupt. ECC RAM is not much more expensive these days but you do need a mobo that supports it.
You should also make sure that your system redundantly performs every CPU instruction like any decent mainframe would, since the same sorts of problems that can cause RAM to fail can cause CPUs to fail (and CPUs contain cache too).
That isn't a big problem, you just need a CPU/motherboard that supports this. IBM would be more than happy to sell one to you if you have a house you're willing to trade for it.
There is always another failure mode...
Because ZFS has far better features than BtrFS
http://rudd-o.com/linux-and-fr...
It has SOME features which btrfs has not yet implemented. Btrfs also has some features which ZFS has not yet implemented, including support for dynamically resizing a RAID (not adding/removing a RAID from a zpool).
Identical twins, born to the same parents, but adopted by different families, tend to have extraordinarily unlikely similarities in adult general intelligence scores.
Which is it - academic performance, or general intelligence scores? If the twin studies test one, and this genetic study tests the other, then I would think that this would make it very difficult to draw any conclusions by comparing the results, since academic performance and IQ aren't the same thing.
Not quite true.
It shows that a large number of specific candidate genes don't do it. Even if it's not a complete refutation of the hypothesis, it is a push to maybe look elsewhere for some of the mechanisms of intelligence development.
Well, it only would show that those genes don't have an individually-detectable affect on whatever marker they looked at. Maybe collectively they have an effect that can't be detected statistically. Maybe the marker they chose doesn't make sense.
Imagine if I tried to identify the gene for "sickness." I took anybody who ever got sick for any reason and studied their DNA and looked for a common link. Most likely I wouldn't find anything. Would this prove that genes can't make you sick? Or is it more likely that "sickness" is such a broad description of a phenotype that it could have a billion different causes.
Academic performance could be the result of MANY factors. Physical attractiveness has been demonstrated to have an impact on academic performance, and you'd hardly expect the same genes to affect that as your ability to do some kind of mental processing. That is just picking one attribute that is obviously going to confound results. Then you get into stuff like whether intelligence is about persistence, or ability to process information, or memory, etc. All of those things are likely to affect academic performance. Then there are cultural factors - let's just assume the sterotype about Asians prodding their kids to study harder is true - I'm sure there are alleles more common in Asians (the fact that they have distinctive appearances makes this obvious for starters), and that is going to confound things.
If you really want to identify the genes responsible for a trait, you have to first come up with a very precise definition for the trait, ideally one tied to some kind of biological mechanism (good luck if it involves the brain).
The study did not demonstrate this at all. It simply failed to find specific genes responsible for intelligence.
You might still be born with a set intelligence which isn't genetically determined, or it might be genetically determined on the basis of genetic factors that were not identified for any number of reasons. Or you could be right.
The point is, this study doesn't provide any evidence one way or another.
Also, equating academic performance with intelligence may be dubious. There could be many factors responsible for academic performance, of which intelligence is just one. We can't even define what intelligence is...
Indeed, if they randomly distributed 2M surfaces for free throughout the population it would probably cost them about the same, and it would immediately create a demand for applications, it would get their product in front of consumers, etc. That would seem like a far better investment if you really wanted to blow this kind of cash.
Now you're quibbling. You can nitpick with me all you want but never try that shit on a clearance form. You wont like it.
And this is why the clearance process should be changed. I have no doubt that everything that happened to this individual happens to people all the time.
Just one more reason why nobody competent wants to work for the government...
So, DON'T LIE.
Giving an incorrect answer isn't a lie. Deliberately giving an incorrect answer is a lie.
How do I know if the math club I was in during college 20 years ago is now considered a terrorist organization?
This sort of thing is over the top.
It doesn't work on my Nexus 10 either. It works just fine on my Oneplus One. I can't imagine they'd have bothering to whitelist the OPO, so I'm not sure what the logic is.
It doesn't work on my Nexus 10 (amazon for tablets app doesn't support it, regular amazon app isn't supported). It works fine on my Oneplus One.
I suspect that if I hack things to bypass the store device compatibility list it would work, but I haven't tried that yet.
No, the evidence is not the property of a US company. The evidence is the property of a US citizen and is being held by a foreign subsidiary of a US company and is, therefore, exactly like the hotel example.
No argument there. However, if I leave evidence in a US hotel chain's London hotel, a US court can certainly order the hotel chain to fly it back to the US and threaten to fine it until it complies.
I'm not arguing morality here - just practicality. If a guy walks into your house and manages to hold you at gunpoint and you're unarmed, as far as you're concerned, he's the boss and he gets what he wants until somebody else shows up and bails you out. As Augustine pointed out, pirates and emporers are the same thing on different scales, but in the case of the emporer there usually isn't a bigger fish who is going to come bail you out if you can hold out long enough.
I don't think that is it. Even a small phone tends to be in your pocket all the time.
The point of the smartwatch is that instead of having your phone vibrate to tell you to take it out of your pocket, turn it on, and navigate to a message, you can instead just look down at your wrist and see the message without touching anything.
If I got a lot of notifications at a time I wasn't already in front of my PC it would make sense. The problem is that 99% of the time the notification is just going to be something I could see in another browser tab on the PC I already have open in front of me. If I still worked in an office where I had a meeting every hour in a different conference room maybe having my calendar on my watch might be useful, but I just phone into all my meetings now. I feel like the smart watch is solving a problem that has already been solved in other ways.
So, your argument is that if Microsoft operates in North Korea, Iran, or any other country .. that those countries should also be able to force Microsoft to hand over any and all records on Americans if they see fit?
It isn't a matter of "should." The fact is that those countries can do anything they have the power to do - that is what makes them countries and not corporations. You'd be a fool to store your personal data with a company that needs to keep a country like North Korea happy if you don't want the North Koreans to have access to your data.
If you run a business selling swastikas, you'd be foolish to build all your infrastructure on servers owned by a company that counted on the German government for most of its business. At any time the Germans could ask them for a favor, and you're out of business.
You got that wrong, this is the US providing heavy stimulus for foreign companies creating their own cloud services. They're basically giving free reign to European providers, who do not own a single server in the US, and telling them to go ahead and dominate the market. Europeans should be thanking them, if anything!
They could only dominate the European market though. Those companies wouldn't be able to do business with companies in the US. If they did, then the US government would just threaten them until they gave in - just seize all their US bank accounts, or seize all their payments, harass employees if they travel to the US, etc.
It isn't as simple as you make it out to be. And it isn't just the US doing this sort of thing. Everybody wants the "Internet" to follow their local laws.
If I commit a crime and the evidence is in a Hilton hotel room in the UK nobody expects Hilton
Everybody however expects U.S. police to go the U.K. police and courts who will check if it was a crime in the U.K. in the first place then ask the U.K. Hilton to hand over the data, aka due process .
The problem here is that the evidence is the property of a US company. The US can't send in police to seize the evidence overseas, but it can punish the owner of the evidence until they voluntarily turn it over.
It is like refusing to hand over an encryption key - they may not have the ability to seize it without cooperation, but they can still beat you with a hose.
"We all work so that we can pay the bills... go on and do bigger and better things that 2BN dollars can provide."
Will the work on the things post-2BN also be to pay the bills? Seems like a contradiction. I wonder what proportion of people are actually not working to pay the bills.
Maybe 0.05%? One of the advantages of pocketing the $2B is that now you can take all the risks you want to and do the stuff you REALLY want to do. Maybe that makes you even more money, or maybe it doesn't, but who cares at that point?
Work would be a lot less stressful for me if I had $10M in the bank. I'd probably still be doing a lot of the stuff I'm doing today, though I might not be doing it for the same employers. I'd be more interested in getting good stuff done and less about whether my boss likes me, etc.
Well, in your analogy there is another big difference between IBM and Glaxo. In your example, if IBM does a faulty study that shows that a new way of making microchips is better, then the result will be that they spend a billion dollars on a new fab and it produces faulty microchips that nobody buys. In the other case, it is others who are relying on the results of the study to make a decision about whether to buy Glaxo's products. This is why it is much more important that drug trials be regulated. If it were up to me I'd have them all publicly conducted, but nobody wants to spend tax dollars on clinical trials.
There are lots of things that are messed-up about clinical trials, and the main reason for this is that they are VERY expensive to run. The problem is that you can't give somebody a drug unless their doctor is involved. Doctors make a lot of money, and clinical trials take a lot of time for them to participate in. So, if you want somebody who makes $500k/yr to spend 10 hours per week on a clinical trial, what is the only way to get them to cooperate? You have to pay them a LOT of money. Multiply this by thousands of patients, and then factor in that there are 14 cancer trials going on at the same time, so the doctor is going to pick the trial that pays the best to recommend to their patients, and now you have a bidding war on top of it.
Nobody wants to spend the money needed to either sponsor trials run by a disinterested party, or replicate trials.
(And yes, I do realize that some doctors are among the salt of the earth, and they will participate in trials because it is the right thing for their patients even if they were completely uncompensated. However, for every one of these there are probably 5 others that make it onto the FDA's debarrment list for getting patients to take experimental drugs that they shouldn't be taking just to collect extra fees. The list is posted online - just read it for yourself.)
Anytime you re-analyze data you run into this.
Think about it. There are a million ways you can analyze any dataset. There are millions of datasets out there to analyze. There are millions of people who can independently decide to go back and do a re-analysis.
So, the issue is that if somebody goes back and does a re-analysis and the results are boring, nobody publishes. However, if the results are controversial, it gets published. Since there are so many permutations, you're guaranteed to find something exciting.
This is why you're supposed to establish your methods BEFORE you collect the data, and then stick to the methods you established to analyze the data. Otherwise your 95% confidence turns into a more realistic 1% confidence.
In practice, though, I'm sure the initial analyses are just as prone to this kind of problem. It just gets REALLY bad when you look backwards.