People will stop being paid outrageous salaries for "playing with money" when we stop giving it to them to play with.
Everybody wants to put away $1M by the time they're 50 so that they can retire and coast for the rest of their lives. When millions of people want to have a million dollars in the bank, that means there are TRILLIONS of dollars looking for investments. The entire productive manufacturing and service economy of the planet doesn't need that much money to operate. However, money managers need to do something with the 90% of the cash in their mutual funds that they can't find reasonably-priced stocks to buy with. So, all kinds of artificial investments get created so that they can say that they money is invested in something that could possibly make a buck.
Have a spare couple of billions of dollars burning a hole in your mutual fund? Well, how about buying oil futures and see if that goes up (surprise - it does when everybody does this). Then along comes a financial bump and you need cash so you sell those futures (as does everybody else) and suddenly the cost of gas at the pump is cut in half.
I'm thinking that there is a fundamental problem with the way our savings-based economy works. Millions of people want to work hard today and then not work at all for the last 20 years of their life. Does that even work (especially with first-world populations declining)? All that cash going into mutual funds leads to inflation (money isn't worth much when people who want to borrow it can find thousands of people desperate to loan it for even a paltry rate of return). When all those boomers retire and want to start going out to eat, I wonder if we'll see costs and salaries suddenly start soaring (when there is a 2 hour wait at the local fast food place they can raise prices and all the wealthy retirees will pay it). Massive inflation would then essentially make retirement accounts worthless and then everybody ends up going back to work. In theory the guy who puts nothing in his 401k could end up just as well off as somebody who only spent half their income and stashed the other half for their entire career (since the only wealth will be from current income - which would be the same for either).
Bottom line is that money managers make money because there is more demand for money managers than people who do real work. This is solely because the average 50-year old at the peak of their earnings doesn't want to buy anything with their money, but instead wants to put it in a mutual fund.
Oh well, go figure. On the btrfs wiki it looks like you can already reshape a btrfs. So, in that respect it is ahead of ZFS. Granted, I wouldn't trust it for production use yet.
Actually, if you look at the btrfs planned feature list I don't see much lacking that ZFS has to offer (besides being available today on select operating systems). I'm not going to switch to solaris just for ZFS (although ZFS via fuse might be an option). I don't think anybody wants to reinvent the wheel or anything, but the current ZFS licensing restrictions and patent encumberance does limit its usefulness to a GPL software project.
One thing that ZFS is currently lacking (and which isn't mentioned for btrfs) is raid reshaping. Sure, you can add a new RAID-Z to a zpool or remove a raid-z from a zpool. However, you can't add one drive to a raidZ, or remove one drive from a raidz. Linux software raid supports both offline and (at more risk) online reshaping of an array. This is very useful for home setups, where a user might want to have the advantages of RAID without having to add drives in at least pairs (and adding them only in pairs wastes 50% of your space).
A practical illustration. I have 3x250GB drives in a RAID-5 (500GB usable space). I want to add 250GB more space. With linux I can just buy one more 250GB drive and add it to the existing array, making a 4x250GB RAID-5 with 750GB of usable space. With ZFS I'd have two options. I could buy 2x250GB drives and create a RAID and add it to the zpool (gaining 250GB of space at the cost of an extra dive). The other option is to find someplace to dump my 500GB of data and then recreate the entire raid from scratch (which is only an option if you have access to 500GB of unused storage).
Nothing about the design of ZFS really prevents this - it is just a feature that hasn't been implemented. It would be nice to see the btrfs folks implement this (or the ZFS folks assuming some day I can use it).
Name one drug available in the 1950s that you can't get generic today in any country in the world.
Name one drug available in the 1980s that you can't get generic today in any country in the world. (There is a chance you might find some biologic product somewhere, but it would be more the exception than the rule.)
The stunts you referenced in your link only work for a few years. They're completely immoral and companies that practice them should be punished (believe it or not, not all big pharma companies do this and the ones who don't are put at a disadvantage by these kinds of tactics). However, they're not the reason that anything invented in the 80s is expensive today (maybe some items from the early 90s though).
If you want patent-free medicine just wait ten years. Sure, block stupid legal stunts that allow patent extensions, but everything that is patented today will be patent-free in a decade. The problem with getting rid of patents for drugs is that there is then no motivation to create new drugs (short of government funding - which currently isn't sufficient to do the job). I'm all for increased government funding for full scale drug development (from soup to nuts) - no need to ban patents to make that happen. I consider it the best of both worlds - doctors and patients can choose between new drugs that are cheap and ones that aren't cheap - and pick whatever works best for them. If the government model works well the private sector will go away or subcontract for the government. If the government model doesn't work at least we'll be back to the status quo (as opposed to just having no new drugs at all).
Getting rid of drug patents is like putting your head in a hole. Sure, all the drugs will be cheap, but in 10 years there won't be any new drugs either and all those cheap drugs would have been cheap anyway. You just won't realize what you're missing out on because nobody will be inventing it.
Oh, and don't give me the line about how government R&D already does "most" of the work discovering new drugs. Don't get me wrong - government blue sky research does turn up most of the ideas that eventually become drugs. However, what comes out of government labs is ten thousand ideas that each cost millions of dollars to pursue and only one of which ends up working out. They do create value, but so do the companies that do the grunt work of developing those drugs. I'm all for the governemnt holding onto some of its ideas and pursuing its own drugs (free of all royalties) - but that isn't the exciting kind of basic research that most academics like and I'm skeptical that it will work out.
One way to look at it is this: The NIH is the architect for the skyscraper, and some big pharma company is the general contractor that employs the 50,000 guys who actually put the thing up. Sure, the architect might have come up with the concept, but that doesn't mean it didn't cost a whole lot of money to turn the drawing into reality. The architect has two choices - pay the contractor to put up the building, or sell the drawings to the contractor. Currently the NIH does the latter - the contractor owns the building. I suggest that for some drugs they should do the former - I think that big pharma companies would be willing to just make a modest fee to develop drugs and then not hold the patents. I think that ideally both should happen - until it is shown that one model or the other is superior.
As far as the supply of doctors goes - I agree with you completely. There is artificial scarcity in the medial world and it costs a LOT of money. In fact, the bulk of the cost of even drug R&D goes towards paying doctors to participate in clinical trials.
Keep in mind that linux will never get to the point where a free system will "just work" if everybody just accepts proprietary software all over the linux desktop.
I have similar problems with flash. Sure, it is a pain to watch flash on a 64-bit linux desktop. However, I see that as a reason to replace adobe flash, not a reason to find 47 hacks to try to get it to work in spite of the vendor's lack of support.
If you just want something that "works," then buy a copy of Vista. But don't complain when you find certain things that you want to do that the vendor has decided you ought not to be able to do.
The point of the FOSS movement is to get FOSS to a point where we don't need to compromise. We haven't arrived yet. However, we won't get there by accepting compromise either. In practice we all do it, but that doesn't mean that we have to like doing it.:)
This is brainwashing into believing there is no alternative, and that anything else on the outside isn't worth mentioning. Hell, it's pretty close to the sect of RMS denying that there even is an outside.
No, it is about saying that if you need to promote non-free software then you're not truly free. RMS doesn't pretend that proprietary software doesn't exist. His goal is that people will be able to live without it.
If your linux distro promotes the use of non-free software, then that means that people who refuse to use proprietary software are going to get a sub-par experience. Why would RMS want to promote a distro that gives users a sub-par experience if they're unwilling to compromise their ideals?
Is it practical today to distribute software that meet's RMS's guidelines? No. And I don't think that RMS considers it completely practical either. However, if you're going to aim for something at least aim for something good. It would be very nice to live in a world where I can choose to use only free software and not suffer negative consequences - RMS is simply saying that we haven't arrived there yet.
This isn't about censorship. This is about compromise. When a "free" distro forces users to install non-free software (and even makes it easy to do so), the distro is free in name only. They might skirt around copyright by having users fetch tarballs direct from vendors or by licensing rights, but if their users want to go around mirroring their machines they're technically violating copyright (unless they obtain a license to do so). Once you start compromising you lose some of the benefits of FOSS. RMS is simply recognizing this.
Just ask yourself the question - why would a distro promote non-free software and make it easy to install? The anwer will amount to something that could be done with free software, but isn't being done. That is something that needs to be fixed - not something that we should just live with permanently.
Uh, you obviously haven't been following RMS for long if you're only coming to that conclusion now.:)
I think that people are missing RMS's point in those policies (which you could probably pick up if you read more of what he writes). His issue isn't with people being free to use proprietary software. His issue is with distros that are free in name only, but in practice you NEED to use proprietary software to really get full functionality out of them.
For example, the system I'm running on makes it very easy to install the usual suspects (flash, nvidia drivers, win32 codecs, etc). If you post on the distro's forums complaining that you can't get gnash to work (or whatever) you just get a helpful one-line response telling you how to install adobe's version. Sure, it is easy to do, but what if I don't want to use adobe flash? The proprietary software has become a crutch to allow distros to not bother getting certain free software applications to work. I accept the compromise out of practicality, but I agree completely with RMS that this is a compromise and that my system is not truly free.
What RMS objects to are distros that might be free in some technical sense of the word, but which practically force their users to install non-free software. When a distro promotes the use of non-free software it is usually because the distro is not fully usable in its absence. If somebody does want a completely free system they shouldn't need to start a research project to figure out what will and won't work. RMS has made it simple - if the distro claims to follow his rules it probably can be used in a completely free manner.
Your experience is exactly the reason that RMS considers Ubuntu non-free. Sure, they distibute software free of charge, but they design their OS so that you can only use it practically if you're willing to accept non-free software which they then make easy to install. So, most people who use Ubuntu end up having non-free systems as a result.
The proper solution would be for ubuntu to distribute a free driver for their wireless card.
Now, I'll admit that this isn't entirely practical today, but it is something worth shooting for. There are very few distributions that RMS would consider free, and Ubuntu isn't one of them. There is nothing wrong with that. What he describes is where he wants the world to be, not how he can compromise his vision so that we can claim that we've already arrived.
So many FOSS advocates go for the second, and then they wonder why the OS has zero userbase on the desktop, why it's painful to use and unstable, etc.
I doubt the people who push for the second wonder why the OS has zero userbase - they're quite aware that their ideals aren't accepted by everybody. I wouldn't call the fruits of their labor "painful to use" or unstable - it isn't meant for your grandmother but anybody who knows what they're doing should do just fine (better than with proprietary software since the source is available).
RMS's goal isn't to get the masses to use linux - it is to promote the use of free software so that there is something out there that HE can use. Quite a few others share his ideals, even if they constitute a very small minority of the people who buy desktop computers.
The FOSS community doesn't need to make a non-free OS that is popular on the desktop - several of those already exist. However, FOSS has its advantages and the people who need them flock to it. And many of those who use FOSS to make money do care about what RMS says. Let's suppose I'm a manager at Google and I've got 10,000 servers in a grid running some linux-based OS. I just realize that those servers depend on some binary firmware blob with questionable licensing. My company might be violating copyright by copying that blob all over the place. Chances are, the next time I modify my hardware specs (they're constantly replacing hardware due to failures) I go ahead and find a replacement part that either doesn't require a blob, or requires a blob that has clear licensing that I can live with. Once I do that in a few years I'm immune to a lawsuit asking me to pay $50 for every server I own.
The kinds of people who don't care at all about licensing are people who generally can get away with violating licensing at will, or who just need one license for $90 and are willing to just pay for it. If software vendors actually made copy protection that worked (assuming that were theoretically possible) you'd see a huge uptick in FOSS software use. FOSS is taking off in the business world because companies have cracked down on the use of pirated software in the wake of expensive audits. In fact, where possible companies minimize the number of proprietary software packages they use to avoid subjecting themselves to the audits in the first place (which are painful even if you are in compliance - and half the pain is in being prepared for one even if you never actually get audited).
If you're a company that switched to FOSS to get out of tracking software licenses, how happy would you be if you found out the distro you're using includes non-free software that you need to start tracking licensing on?
Yes, RMS does go to an extreme. Yes, it isn't practical to live without any blobs currently. However, it is a goal worth aiming for because it does have the potential to cause problems.
RMS isn't about suing people who use software in ways he doesn't like - just those who distribute his software illegally (such as distributing software containing copyrighted GPL'ed code without a license to do so). When did he ever suggest that it was a violation of the GPL to use a GPL'ed program to load a binary blob onto a piece of hardware?
This isn't about whether it is a violation of the GPL to load a piece of data into memory somewhere. This is about whether a distribution that includes such a piece of data is truly free.
I happen to agree with him, even though I do have some binary blobs that I need to use to get certain hardware to work. I would be willing to jump through hoops to get rid of these if that were an option - perhaps some day it will be. If you look at RMS that is what he does - as the state of FOSS moves on he adopts it for more and more of the world around him. I'm sure his microwave isn't currently powered by FOSS, but perhaps in 10 years it will be.
Sure, it is easy to compromise (ok, just the video drivers, oh, and just the flash plugin, oh, and just the win32 codecs, oh, and just the trademarked web browser, etc...). The problem is that distros that compromise find themselves having to jump through hoops. When they port their distro to 64-bit suddenly they start hitting brick walls, and when they need to backport a security patch they get a cease-and-desist letter. Then, they need to deviate from their philosophy due to using non-free software (ok, normally we backport security patches, but in this one case you need to modify your configuration and take an upgrade, because the world can't live without brand-X).
Sure, RMS is an idealist, but he has been ahead of the curve on the problems with non-free licenses, and many things that people say "would never happen" have in fact happened. When you become dependent on something with a non-free license you're giving control of part of your world away. Now, 95% of those vendors might be nice, but if you have 1000 different pieces of proprietary software in your system you'll find out who the other 5% are fairly quickly.
It makes you wonder if there is any way to do a system compromise at the hardware level (taking over DMA, etc). I remember reading back that there was at least one popular system interface (firewire?) that allowed the foreign device DMA access. If a device made by a non-trusted vendor can access a systen's DMA, then it can be used to completely compromise the system.
Then just take an infected computer, keep changing its clock for the next month, and see where it goes.
Then set up a sting for the appropriate domain names (and don't allow them to actually be registered). The behavior is deterministic so you can stay ahead of the worm.
Agreed that it wouldn't be easy to stay ahead of it indefinitely. However, authorities could register a year's worth of DNS entries to stall for time, and then send emails to ISPs about any infected IPs that connect.
If that was their mindset they probably wouldn't need to visit either NY or DC.
The biggest problem with trains is train advocates. I personally think that the US could benefit substantially from investment in trains/etc.
Here is the problem. A $50B train system that might save $5T in other infrastructure and wasted time (numbers made up for illustrative purposes). Government types will say $50B is too much - let's do it for $5B. Train advocates will lobby on behalf of the proposal and the $5B will get spent. Then, nobody will use the system because it isn't practical (not enough routes, too much waiting, etc). The train advocates will say that people should suck it up and live with inconvenience. All the talking heads agree, but the reality is that everybody still drives to work. Then the government spends the $5T in other infrastructure since nobody is using the trains. Instead of saving $5T now we've managed to waste $50B more.
What we don't need is people arguing how we all just need to suck it up and deal with delays. What we do need is designs for infrastructure that can get people from point A to point B without being crammed in cars with standing room only, without being subjected to dirty conditions, without the need for 3 changeovers with a quarter mile walk at each one, and where it takes LESS time to get from point A to B than it would by driving in current congestion.
I'm sure that with proper investment a good mass-transit solution could do what highways do today better and cheaper. The problem is that the people who push these sorts of things only aim for cheaper, and then nobody uses them.
I've always found the military interesting. On the one hand they'll have the latest and greatest gear that can be found anywhere. On the other hand they'll land it using TACAN.:)
They have a lot of legacy hardware. I was surprised how little guided ordnance was dropped in the first gulf war. Then I was even more surprised to find out that most weapon systems are integrated with the launching platform - a missile that works on an F14 won't necessarily work on an F15 and vice-versa.
I suspect that half the problem is the way military contracts are done. The people making the stuff are in competition with each other, so there isn't incentive to collaborate on something like universal missle communication protocols (so you can hook any missle up to any plane). Also - the original contract includes R&D and deployment of N planes. If you want to change one feature after 10 planes are delivered now you re-open the contract negotiation (with only one bidder since nobody else can make it without R&D).
I've seen similar stuff at work - some favorable contract terms are reached and then everybody is stuck in gridlock without any ability to make changes for a decade.
As others have pointed out, this is already the status quo.
What is your threat model? Terrorists hijack 747 to crash it into a barn near Topeka?
A reasonable air defense network isn't going to try to have 100% coverage of the country.
This system is supposedly very accurate. That means that you could create a data feed from the FAA to NORAD with realtime positioning on all legitimate air traffic. Then NORAD could get a data-feed from military radar. Anything that isn't civilian gets investigated. Anything civilian that behaves outside its published performance envelope (a "747" flying supersonic, for example) also gets investigated. An improved civilian radar system can improve military response, since it can help eliminate the 99.9999% of law-abiding traffic in the air.
In theory the data could go both ways as well, so that civilian controllers can have primary data for non-broadcasting aircraft. Military radar in theory would be much more accurate than what controllers are currently using (I don't think current ATC primary radar gives you anything other position - military radar most likely will give you altitude and speed as well).
While in theory it could be cheaper, I see one practical problem - demand.
When this goes live there will probably be some mandate to install it by a certain date. That means thousands to millions of these things will be flying off the shelves. If the leading avionics manufacturers want to charge for 300% profits are you going to stop flying to avoid buying one?
Yeah, but do you really need to block the whole country?
The bots obviously need to find their home. Most likely this is via either a hard-coded IP, or a DNS lookup. So, just publish whichever one it is and then everybody can blackhole either the DNS entry or the IP address. If the major ISPs do that the bot dies.
Now, if the bot uses IRC or something like that it could get trickier, since blocking that at the protocol level (short of killing an entire irc network) isn't possible. However, the irc network could probably block the appropriate channels.
The problem in Vietnam was that the US fought a very limited war. It used enough force to generally tick everybody off (including its own populace), and not enough to actually do anything significant.
Let's assume for the sake of argument that it was worth being over there in the first place (if we're going to argue about ways to "win" Vietnam you have to assume that it is even winnable). The best approach would have been what was used in Iraq - have a conventional army walk right across North Vietnam. That would have eliminated most of the opposition's conventional army capability (rather than giving them almost complete sanctuary in the North while the US fought only in the South).
The NVA relied on constant resupply of forces and lots of indoctrination of the population. That can't happen to that degree if they're an underground movement. In the North they weren't underground, which game them a huge supply of cannon fodder.
By any conventional measure US forces were very successful in actual engagements. The problem was that there were too few of them and they weren't fighting a strategic battle. They would win tactically (with heavy losses, but the NVA lost far more). The NVA also maximized their propaganda impact with unsustainable offences in the South that made no military impact, but which had huge TV impact. The US made the mistake of doing all the fighting on its "own" turf - the South - rather than taking the fight to the North. A big lesson from the US Civil War is that it is always better to do your fighting in the other guy's cities. Casulties in the newspapers are bad - but casulties in your own streets make for a very fast loss of popular support.
I'm no fan of what the US did in Iraq, but given another 5-10 years the war could possibly be "won". If the US invested sufficiently in modernizing the country chances are that the Iraquis 15 years from now would get along as well with the US as Germans and Japanese do today. That's how long it would take to westernize their culture. Choking off the Iranian border probably wouldn't hurt either.
Again, I'm not actually advocating conquering other countries and then imposing US culture. However, I suspect that it would actually work - if there was no holding back. People will put up with all kinds of stuff as long as they are fed.
I doubt you'll see harsh measures enacted against such a thing.
Cloning a neanderthal isn't much different from:
1. Conceiving a child one can't care for. 2. Aborting a fetus. 3. Reproductive cloning. 4. Cloning entire embryos for the purpose of organ harvesting.
They all entail the same kinds of issues. Only a few of these are even frowed upon universally - some are considered by many to be a fundamental right (the right to have an abortion, for example). I think they all entail significant moral problems as they amount to trivializing the value of human life in some way. However, I'm under no illusions that the majority of people living on this planet will agree with me on this.
The only thing different about cloning Neanderthals is that it isn't the sort of thing that is likely to become commonplace. That will enable the masses to possibly condemn it without worrying too much about the disconnect between that position and others they hold.
Frankly the chief religious concern in all of this isn't the fact that Neanderthals existed, but the moral issues with cloning living beings for the sole purpose of scientific study.
Cloning for reproductive purposes is already frought with ethical and moral issues, and that is generally when parents want to raise a child but can't, and consequently would be expected to do their best to raise it.
Cloning a sentient being for scientific study is a conflict between the collective rights of society (to gain knowledge) against the individual rights of a cloned being. The welfare of the clone doesn't even seem to be much of a consideration (other than in debating whether to do it). Is there any chance that the clone would get a "normal" upbrininging? Would the clone receive citizenship in the nation in which it was born (or based on its "parentage" - and is that the chief scientist conducting the experiment? Who is the parent of a clone that might have been manufactured from DNA from 10 individuals living in 10 different places?)?
Even cloning of whole beings for medical purposes is fraught with controversy, and at least this has a potentially direct impact on the lives of individual humans. It amounts to making a value judgment that the life of an embryo does not have as much value as the life of an adult. I'm not sure I know when exactly human life starts, but drawing a hard line at the birth canal seems a bit arbitrary - there is no question that a child one minute before birth is exactly just as viable as a child one minute after birth.
In my opinion individual rights just about always trump collective rights. About the only exceptions I could think about are when you're talking about minor rights for the individual and very fundamental rights for the society, but I'm hard-pressed to even come up with a good example that doesn't boil down to infringement of the rights of many individuals. The right of society (or even individuals) to advance knowledge does not trump the individual rights to life and liberty, and yet this is exactly what is proposed.
I tend to agree with your point there. In general while individual software packages get redistributed 4700 times, whole distros tend not to be.
Perhaps the FOSS needs a convention. Every product will have two names - the branded name, and the "generic" name (much as is done for drugs). Then there will be a common standard for the name of modified versions of a product, as well as a name for the upstream-guaranteed version.
Just think of all those photos of Roosevelt - the average american would have had no idea he was paralyzed. Reporters and publishers have never been above PR. I don't know who that reflects more poorly on - the pulishers, or those who read what they print.
Yeah, but if every other software package worked that way then you wouldn't be able to function on a typical linux system. Every software package would have a different name on every different distro.
Openoffice - oh, that is starfish here. GCC - oh, that is magnet here. vi - oh, that is supreme-ruler here. emacs - oh, that is uber-editor here.
And so on. The whole point of a distro is to redistribute software. Honestly, I'm half-tempted to create a website to distribute IceWeasel for every platform (just rebranded mozilla builds) and spend $500M on marketing it so that nobody uses the "genuine" Firefox and we can stop playing these games. Too bad I don't have $500M.:)
Debian is right to not play the Mozilla games. Other distros should do the same. Fortunately I use gentoo, which can get around these games since they distribute pristine sources and patches and the user puts them together. Perhaps debian just needs to implement a binary diff utility in their package manager and then they can claim they're distributing genuine "Firefox" while completely bypassing Mozilla's intent in brand control by having the user modify it before using it automatically.:)
Slashdot Mirror
People will stop being paid outrageous salaries for "playing with money" when we stop giving it to them to play with.
Everybody wants to put away $1M by the time they're 50 so that they can retire and coast for the rest of their lives. When millions of people want to have a million dollars in the bank, that means there are TRILLIONS of dollars looking for investments. The entire productive manufacturing and service economy of the planet doesn't need that much money to operate. However, money managers need to do something with the 90% of the cash in their mutual funds that they can't find reasonably-priced stocks to buy with. So, all kinds of artificial investments get created so that they can say that they money is invested in something that could possibly make a buck.
Have a spare couple of billions of dollars burning a hole in your mutual fund? Well, how about buying oil futures and see if that goes up (surprise - it does when everybody does this). Then along comes a financial bump and you need cash so you sell those futures (as does everybody else) and suddenly the cost of gas at the pump is cut in half.
I'm thinking that there is a fundamental problem with the way our savings-based economy works. Millions of people want to work hard today and then not work at all for the last 20 years of their life. Does that even work (especially with first-world populations declining)? All that cash going into mutual funds leads to inflation (money isn't worth much when people who want to borrow it can find thousands of people desperate to loan it for even a paltry rate of return). When all those boomers retire and want to start going out to eat, I wonder if we'll see costs and salaries suddenly start soaring (when there is a 2 hour wait at the local fast food place they can raise prices and all the wealthy retirees will pay it). Massive inflation would then essentially make retirement accounts worthless and then everybody ends up going back to work. In theory the guy who puts nothing in his 401k could end up just as well off as somebody who only spent half their income and stashed the other half for their entire career (since the only wealth will be from current income - which would be the same for either).
Bottom line is that money managers make money because there is more demand for money managers than people who do real work. This is solely because the average 50-year old at the peak of their earnings doesn't want to buy anything with their money, but instead wants to put it in a mutual fund.
Oh well, go figure. On the btrfs wiki it looks like you can already reshape a btrfs. So, in that respect it is ahead of ZFS. Granted, I wouldn't trust it for production use yet.
Actually, if you look at the btrfs planned feature list I don't see much lacking that ZFS has to offer (besides being available today on select operating systems). I'm not going to switch to solaris just for ZFS (although ZFS via fuse might be an option). I don't think anybody wants to reinvent the wheel or anything, but the current ZFS licensing restrictions and patent encumberance does limit its usefulness to a GPL software project.
One thing that ZFS is currently lacking (and which isn't mentioned for btrfs) is raid reshaping. Sure, you can add a new RAID-Z to a zpool or remove a raid-z from a zpool. However, you can't add one drive to a raidZ, or remove one drive from a raidz. Linux software raid supports both offline and (at more risk) online reshaping of an array. This is very useful for home setups, where a user might want to have the advantages of RAID without having to add drives in at least pairs (and adding them only in pairs wastes 50% of your space).
A practical illustration. I have 3x250GB drives in a RAID-5 (500GB usable space). I want to add 250GB more space. With linux I can just buy one more 250GB drive and add it to the existing array, making a 4x250GB RAID-5 with 750GB of usable space. With ZFS I'd have two options. I could buy 2x250GB drives and create a RAID and add it to the zpool (gaining 250GB of space at the cost of an extra dive). The other option is to find someplace to dump my 500GB of data and then recreate the entire raid from scratch (which is only an option if you have access to 500GB of unused storage).
Nothing about the design of ZFS really prevents this - it is just a feature that hasn't been implemented. It would be nice to see the btrfs folks implement this (or the ZFS folks assuming some day I can use it).
Name one drug available in the 1950s that you can't get generic today in any country in the world.
Name one drug available in the 1980s that you can't get generic today in any country in the world. (There is a chance you might find some biologic product somewhere, but it would be more the exception than the rule.)
The stunts you referenced in your link only work for a few years. They're completely immoral and companies that practice them should be punished (believe it or not, not all big pharma companies do this and the ones who don't are put at a disadvantage by these kinds of tactics). However, they're not the reason that anything invented in the 80s is expensive today (maybe some items from the early 90s though).
If you want patent-free medicine just wait ten years. Sure, block stupid legal stunts that allow patent extensions, but everything that is patented today will be patent-free in a decade. The problem with getting rid of patents for drugs is that there is then no motivation to create new drugs (short of government funding - which currently isn't sufficient to do the job). I'm all for increased government funding for full scale drug development (from soup to nuts) - no need to ban patents to make that happen. I consider it the best of both worlds - doctors and patients can choose between new drugs that are cheap and ones that aren't cheap - and pick whatever works best for them. If the government model works well the private sector will go away or subcontract for the government. If the government model doesn't work at least we'll be back to the status quo (as opposed to just having no new drugs at all).
Getting rid of drug patents is like putting your head in a hole. Sure, all the drugs will be cheap, but in 10 years there won't be any new drugs either and all those cheap drugs would have been cheap anyway. You just won't realize what you're missing out on because nobody will be inventing it.
Oh, and don't give me the line about how government R&D already does "most" of the work discovering new drugs. Don't get me wrong - government blue sky research does turn up most of the ideas that eventually become drugs. However, what comes out of government labs is ten thousand ideas that each cost millions of dollars to pursue and only one of which ends up working out. They do create value, but so do the companies that do the grunt work of developing those drugs. I'm all for the governemnt holding onto some of its ideas and pursuing its own drugs (free of all royalties) - but that isn't the exciting kind of basic research that most academics like and I'm skeptical that it will work out.
One way to look at it is this: The NIH is the architect for the skyscraper, and some big pharma company is the general contractor that employs the 50,000 guys who actually put the thing up. Sure, the architect might have come up with the concept, but that doesn't mean it didn't cost a whole lot of money to turn the drawing into reality. The architect has two choices - pay the contractor to put up the building, or sell the drawings to the contractor. Currently the NIH does the latter - the contractor owns the building. I suggest that for some drugs they should do the former - I think that big pharma companies would be willing to just make a modest fee to develop drugs and then not hold the patents. I think that ideally both should happen - until it is shown that one model or the other is superior.
As far as the supply of doctors goes - I agree with you completely. There is artificial scarcity in the medial world and it costs a LOT of money. In fact, the bulk of the cost of even drug R&D goes towards paying doctors to participate in clinical trials.
Keep in mind that linux will never get to the point where a free system will "just work" if everybody just accepts proprietary software all over the linux desktop.
I have similar problems with flash. Sure, it is a pain to watch flash on a 64-bit linux desktop. However, I see that as a reason to replace adobe flash, not a reason to find 47 hacks to try to get it to work in spite of the vendor's lack of support.
If you just want something that "works," then buy a copy of Vista. But don't complain when you find certain things that you want to do that the vendor has decided you ought not to be able to do.
The point of the FOSS movement is to get FOSS to a point where we don't need to compromise. We haven't arrived yet. However, we won't get there by accepting compromise either. In practice we all do it, but that doesn't mean that we have to like doing it. :)
This is brainwashing into believing there is no alternative, and that anything else on the outside isn't worth mentioning. Hell, it's pretty close to the sect of RMS denying that there even is an outside.
No, it is about saying that if you need to promote non-free software then you're not truly free. RMS doesn't pretend that proprietary software doesn't exist. His goal is that people will be able to live without it.
If your linux distro promotes the use of non-free software, then that means that people who refuse to use proprietary software are going to get a sub-par experience. Why would RMS want to promote a distro that gives users a sub-par experience if they're unwilling to compromise their ideals?
Is it practical today to distribute software that meet's RMS's guidelines? No. And I don't think that RMS considers it completely practical either. However, if you're going to aim for something at least aim for something good. It would be very nice to live in a world where I can choose to use only free software and not suffer negative consequences - RMS is simply saying that we haven't arrived there yet.
This isn't about censorship. This is about compromise. When a "free" distro forces users to install non-free software (and even makes it easy to do so), the distro is free in name only. They might skirt around copyright by having users fetch tarballs direct from vendors or by licensing rights, but if their users want to go around mirroring their machines they're technically violating copyright (unless they obtain a license to do so). Once you start compromising you lose some of the benefits of FOSS. RMS is simply recognizing this.
Just ask yourself the question - why would a distro promote non-free software and make it easy to install? The anwer will amount to something that could be done with free software, but isn't being done. That is something that needs to be fixed - not something that we should just live with permanently.
Uh, you obviously haven't been following RMS for long if you're only coming to that conclusion now. :)
I think that people are missing RMS's point in those policies (which you could probably pick up if you read more of what he writes). His issue isn't with people being free to use proprietary software. His issue is with distros that are free in name only, but in practice you NEED to use proprietary software to really get full functionality out of them.
For example, the system I'm running on makes it very easy to install the usual suspects (flash, nvidia drivers, win32 codecs, etc). If you post on the distro's forums complaining that you can't get gnash to work (or whatever) you just get a helpful one-line response telling you how to install adobe's version. Sure, it is easy to do, but what if I don't want to use adobe flash? The proprietary software has become a crutch to allow distros to not bother getting certain free software applications to work. I accept the compromise out of practicality, but I agree completely with RMS that this is a compromise and that my system is not truly free.
What RMS objects to are distros that might be free in some technical sense of the word, but which practically force their users to install non-free software. When a distro promotes the use of non-free software it is usually because the distro is not fully usable in its absence. If somebody does want a completely free system they shouldn't need to start a research project to figure out what will and won't work. RMS has made it simple - if the distro claims to follow his rules it probably can be used in a completely free manner.
Your experience is exactly the reason that RMS considers Ubuntu non-free. Sure, they distibute software free of charge, but they design their OS so that you can only use it practically if you're willing to accept non-free software which they then make easy to install. So, most people who use Ubuntu end up having non-free systems as a result.
The proper solution would be for ubuntu to distribute a free driver for their wireless card.
Now, I'll admit that this isn't entirely practical today, but it is something worth shooting for. There are very few distributions that RMS would consider free, and Ubuntu isn't one of them. There is nothing wrong with that. What he describes is where he wants the world to be, not how he can compromise his vision so that we can claim that we've already arrived.
So many FOSS advocates go for the second, and then they wonder why the OS has zero userbase on the desktop, why it's painful to use and unstable, etc.
I doubt the people who push for the second wonder why the OS has zero userbase - they're quite aware that their ideals aren't accepted by everybody. I wouldn't call the fruits of their labor "painful to use" or unstable - it isn't meant for your grandmother but anybody who knows what they're doing should do just fine (better than with proprietary software since the source is available).
RMS's goal isn't to get the masses to use linux - it is to promote the use of free software so that there is something out there that HE can use. Quite a few others share his ideals, even if they constitute a very small minority of the people who buy desktop computers.
The FOSS community doesn't need to make a non-free OS that is popular on the desktop - several of those already exist. However, FOSS has its advantages and the people who need them flock to it. And many of those who use FOSS to make money do care about what RMS says. Let's suppose I'm a manager at Google and I've got 10,000 servers in a grid running some linux-based OS. I just realize that those servers depend on some binary firmware blob with questionable licensing. My company might be violating copyright by copying that blob all over the place. Chances are, the next time I modify my hardware specs (they're constantly replacing hardware due to failures) I go ahead and find a replacement part that either doesn't require a blob, or requires a blob that has clear licensing that I can live with. Once I do that in a few years I'm immune to a lawsuit asking me to pay $50 for every server I own.
The kinds of people who don't care at all about licensing are people who generally can get away with violating licensing at will, or who just need one license for $90 and are willing to just pay for it. If software vendors actually made copy protection that worked (assuming that were theoretically possible) you'd see a huge uptick in FOSS software use. FOSS is taking off in the business world because companies have cracked down on the use of pirated software in the wake of expensive audits. In fact, where possible companies minimize the number of proprietary software packages they use to avoid subjecting themselves to the audits in the first place (which are painful even if you are in compliance - and half the pain is in being prepared for one even if you never actually get audited).
If you're a company that switched to FOSS to get out of tracking software licenses, how happy would you be if you found out the distro you're using includes non-free software that you need to start tracking licensing on?
Yes, RMS does go to an extreme. Yes, it isn't practical to live without any blobs currently. However, it is a goal worth aiming for because it does have the potential to cause problems.
RMS isn't about suing people who use software in ways he doesn't like - just those who distribute his software illegally (such as distributing software containing copyrighted GPL'ed code without a license to do so). When did he ever suggest that it was a violation of the GPL to use a GPL'ed program to load a binary blob onto a piece of hardware?
This isn't about whether it is a violation of the GPL to load a piece of data into memory somewhere. This is about whether a distribution that includes such a piece of data is truly free.
I happen to agree with him, even though I do have some binary blobs that I need to use to get certain hardware to work. I would be willing to jump through hoops to get rid of these if that were an option - perhaps some day it will be. If you look at RMS that is what he does - as the state of FOSS moves on he adopts it for more and more of the world around him. I'm sure his microwave isn't currently powered by FOSS, but perhaps in 10 years it will be.
Sure, it is easy to compromise (ok, just the video drivers, oh, and just the flash plugin, oh, and just the win32 codecs, oh, and just the trademarked web browser, etc...). The problem is that distros that compromise find themselves having to jump through hoops. When they port their distro to 64-bit suddenly they start hitting brick walls, and when they need to backport a security patch they get a cease-and-desist letter. Then, they need to deviate from their philosophy due to using non-free software (ok, normally we backport security patches, but in this one case you need to modify your configuration and take an upgrade, because the world can't live without brand-X).
Sure, RMS is an idealist, but he has been ahead of the curve on the problems with non-free licenses, and many things that people say "would never happen" have in fact happened. When you become dependent on something with a non-free license you're giving control of part of your world away. Now, 95% of those vendors might be nice, but if you have 1000 different pieces of proprietary software in your system you'll find out who the other 5% are fairly quickly.
It makes you wonder if there is any way to do a system compromise at the hardware level (taking over DMA, etc). I remember reading back that there was at least one popular system interface (firewire?) that allowed the foreign device DMA access. If a device made by a non-trusted vendor can access a systen's DMA, then it can be used to completely compromise the system.
Then just take an infected computer, keep changing its clock for the next month, and see where it goes.
Then set up a sting for the appropriate domain names (and don't allow them to actually be registered). The behavior is deterministic so you can stay ahead of the worm.
Agreed that it wouldn't be easy to stay ahead of it indefinitely. However, authorities could register a year's worth of DNS entries to stall for time, and then send emails to ISPs about any infected IPs that connect.
If that was their mindset they probably wouldn't need to visit either NY or DC.
The biggest problem with trains is train advocates. I personally think that the US could benefit substantially from investment in trains/etc.
Here is the problem. A $50B train system that might save $5T in other infrastructure and wasted time (numbers made up for illustrative purposes). Government types will say $50B is too much - let's do it for $5B. Train advocates will lobby on behalf of the proposal and the $5B will get spent. Then, nobody will use the system because it isn't practical (not enough routes, too much waiting, etc). The train advocates will say that people should suck it up and live with inconvenience. All the talking heads agree, but the reality is that everybody still drives to work. Then the government spends the $5T in other infrastructure since nobody is using the trains. Instead of saving $5T now we've managed to waste $50B more.
What we don't need is people arguing how we all just need to suck it up and deal with delays. What we do need is designs for infrastructure that can get people from point A to point B without being crammed in cars with standing room only, without being subjected to dirty conditions, without the need for 3 changeovers with a quarter mile walk at each one, and where it takes LESS time to get from point A to B than it would by driving in current congestion.
I'm sure that with proper investment a good mass-transit solution could do what highways do today better and cheaper. The problem is that the people who push these sorts of things only aim for cheaper, and then nobody uses them.
I've always found the military interesting. On the one hand they'll have the latest and greatest gear that can be found anywhere. On the other hand they'll land it using TACAN. :)
They have a lot of legacy hardware. I was surprised how little guided ordnance was dropped in the first gulf war. Then I was even more surprised to find out that most weapon systems are integrated with the launching platform - a missile that works on an F14 won't necessarily work on an F15 and vice-versa.
I suspect that half the problem is the way military contracts are done. The people making the stuff are in competition with each other, so there isn't incentive to collaborate on something like universal missle communication protocols (so you can hook any missle up to any plane). Also - the original contract includes R&D and deployment of N planes. If you want to change one feature after 10 planes are delivered now you re-open the contract negotiation (with only one bidder since nobody else can make it without R&D).
I've seen similar stuff at work - some favorable contract terms are reached and then everybody is stuck in gridlock without any ability to make changes for a decade.
As others have pointed out, this is already the status quo.
What is your threat model? Terrorists hijack 747 to crash it into a barn near Topeka?
A reasonable air defense network isn't going to try to have 100% coverage of the country.
This system is supposedly very accurate. That means that you could create a data feed from the FAA to NORAD with realtime positioning on all legitimate air traffic. Then NORAD could get a data-feed from military radar. Anything that isn't civilian gets investigated. Anything civilian that behaves outside its published performance envelope (a "747" flying supersonic, for example) also gets investigated. An improved civilian radar system can improve military response, since it can help eliminate the 99.9999% of law-abiding traffic in the air.
In theory the data could go both ways as well, so that civilian controllers can have primary data for non-broadcasting aircraft. Military radar in theory would be much more accurate than what controllers are currently using (I don't think current ATC primary radar gives you anything other position - military radar most likely will give you altitude and speed as well).
While in theory it could be cheaper, I see one practical problem - demand.
When this goes live there will probably be some mandate to install it by a certain date. That means thousands to millions of these things will be flying off the shelves. If the leading avionics manufacturers want to charge for 300% profits are you going to stop flying to avoid buying one?
Yeah, but do you really need to block the whole country?
The bots obviously need to find their home. Most likely this is via either a hard-coded IP, or a DNS lookup. So, just publish whichever one it is and then everybody can blackhole either the DNS entry or the IP address. If the major ISPs do that the bot dies.
Now, if the bot uses IRC or something like that it could get trickier, since blocking that at the protocol level (short of killing an entire irc network) isn't possible. However, the irc network could probably block the appropriate channels.
The problem in Vietnam was that the US fought a very limited war. It used enough force to generally tick everybody off (including its own populace), and not enough to actually do anything significant.
Let's assume for the sake of argument that it was worth being over there in the first place (if we're going to argue about ways to "win" Vietnam you have to assume that it is even winnable). The best approach would have been what was used in Iraq - have a conventional army walk right across North Vietnam. That would have eliminated most of the opposition's conventional army capability (rather than giving them almost complete sanctuary in the North while the US fought only in the South).
The NVA relied on constant resupply of forces and lots of indoctrination of the population. That can't happen to that degree if they're an underground movement. In the North they weren't underground, which game them a huge supply of cannon fodder.
By any conventional measure US forces were very successful in actual engagements. The problem was that there were too few of them and they weren't fighting a strategic battle. They would win tactically (with heavy losses, but the NVA lost far more). The NVA also maximized their propaganda impact with unsustainable offences in the South that made no military impact, but which had huge TV impact. The US made the mistake of doing all the fighting on its "own" turf - the South - rather than taking the fight to the North. A big lesson from the US Civil War is that it is always better to do your fighting in the other guy's cities. Casulties in the newspapers are bad - but casulties in your own streets make for a very fast loss of popular support.
I'm no fan of what the US did in Iraq, but given another 5-10 years the war could possibly be "won". If the US invested sufficiently in modernizing the country chances are that the Iraquis 15 years from now would get along as well with the US as Germans and Japanese do today. That's how long it would take to westernize their culture. Choking off the Iranian border probably wouldn't hurt either.
Again, I'm not actually advocating conquering other countries and then imposing US culture. However, I suspect that it would actually work - if there was no holding back. People will put up with all kinds of stuff as long as they are fed.
I doubt you'll see harsh measures enacted against such a thing.
Cloning a neanderthal isn't much different from:
1. Conceiving a child one can't care for.
2. Aborting a fetus.
3. Reproductive cloning.
4. Cloning entire embryos for the purpose of organ harvesting.
They all entail the same kinds of issues. Only a few of these are even frowed upon universally - some are considered by many to be a fundamental right (the right to have an abortion, for example). I think they all entail significant moral problems as they amount to trivializing the value of human life in some way. However, I'm under no illusions that the majority of people living on this planet will agree with me on this.
The only thing different about cloning Neanderthals is that it isn't the sort of thing that is likely to become commonplace. That will enable the masses to possibly condemn it without worrying too much about the disconnect between that position and others they hold.
Frankly the chief religious concern in all of this isn't the fact that Neanderthals existed, but the moral issues with cloning living beings for the sole purpose of scientific study.
Cloning for reproductive purposes is already frought with ethical and moral issues, and that is generally when parents want to raise a child but can't, and consequently would be expected to do their best to raise it.
Cloning a sentient being for scientific study is a conflict between the collective rights of society (to gain knowledge) against the individual rights of a cloned being. The welfare of the clone doesn't even seem to be much of a consideration (other than in debating whether to do it). Is there any chance that the clone would get a "normal" upbrininging? Would the clone receive citizenship in the nation in which it was born (or based on its "parentage" - and is that the chief scientist conducting the experiment? Who is the parent of a clone that might have been manufactured from DNA from 10 individuals living in 10 different places?)?
Even cloning of whole beings for medical purposes is fraught with controversy, and at least this has a potentially direct impact on the lives of individual humans. It amounts to making a value judgment that the life of an embryo does not have as much value as the life of an adult. I'm not sure I know when exactly human life starts, but drawing a hard line at the birth canal seems a bit arbitrary - there is no question that a child one minute before birth is exactly just as viable as a child one minute after birth.
In my opinion individual rights just about always trump collective rights. About the only exceptions I could think about are when you're talking about minor rights for the individual and very fundamental rights for the society, but I'm hard-pressed to even come up with a good example that doesn't boil down to infringement of the rights of many individuals. The right of society (or even individuals) to advance knowledge does not trump the individual rights to life and liberty, and yet this is exactly what is proposed.
If nothing else they'll wonder why our great culture had such a fascination with falling into pits...
I tend to agree with your point there. In general while individual software packages get redistributed 4700 times, whole distros tend not to be.
Perhaps the FOSS needs a convention. Every product will have two names - the branded name, and the "generic" name (much as is done for drugs). Then there will be a common standard for the name of modified versions of a product, as well as a name for the upstream-guaranteed version.
Just think of all those photos of Roosevelt - the average american would have had no idea he was paralyzed. Reporters and publishers have never been above PR. I don't know who that reflects more poorly on - the pulishers, or those who read what they print.
Yeah, but if every other software package worked that way then you wouldn't be able to function on a typical linux system. Every software package would have a different name on every different distro.
Openoffice - oh, that is starfish here.
GCC - oh, that is magnet here.
vi - oh, that is supreme-ruler here.
emacs - oh, that is uber-editor here.
And so on. The whole point of a distro is to redistribute software. Honestly, I'm half-tempted to create a website to distribute IceWeasel for every platform (just rebranded mozilla builds) and spend $500M on marketing it so that nobody uses the "genuine" Firefox and we can stop playing these games. Too bad I don't have $500M. :)
Debian is right to not play the Mozilla games. Other distros should do the same. Fortunately I use gentoo, which can get around these games since they distribute pristine sources and patches and the user puts them together. Perhaps debian just needs to implement a binary diff utility in their package manager and then they can claim they're distributing genuine "Firefox" while completely bypassing Mozilla's intent in brand control by having the user modify it before using it automatically. :)
Yup - I tried that script a while ago actually. :) That script works better than any I've tried, but it isn't 100% effective.
Last I looked into it the devs were actually looking into this very scenario, so we might see this in mythtranscode sometime soon.