"What's good for the n00b might not be right for a poweruser."
Absolutely. The poweruser would be very annoyed if functions were lacking. And remotes, such as shipped with most devices, are notoriously unconfigurable.
So I'll propose a much simpler solution. Send a 'simple' remote with the devices in question, make the device itself capable of outputting its codes for programming (like, an OSD with send-code for each function, or even better, a _standard_ for rapidly programming remotes, or even better than that, a standard for remote codes), clearly write the function codes in the manual, and make sure the specs are widely published and available on the net.
Todays power user will most likely throw the 52-button remote in a box in the closet anyway; they have a much nicer programmable one that actually handles all of their devices.
"Maybe not him but Christie's should definitely get stung for a large wad of cash."
More appropriately, they should be required to, for each item auctioned, state that the item they're selling may or may not be authentic and as far as they're concerned they make more money if you think it is.
"IIRC, their business model rather depends upon the legitimacy of the items going under the hammer."
The actual legitimacy isnt as importance as the appearance of legitimacy. Which may be why auctioning houses have a long and profitable history of making mistakes.
"but do think of me as quite able to tell the difference between HD and SD content."
Well, do a blindtest on yourself and see:) It's not much harder than a script around mplayer and a randomizer.
It does, of course, depend a lot on conditions and how you do the test. One of the tests I saw began with 3.5 meters (11.5 ft) distance, 42 inch LCD, starting with 480p. Everyone in the panel thought it was 720p. After several switches back and forth between different modes some became more capable of distinguising what was what. Further down the row of tests there were still a fair number that couldnt even tell 480p apart from 1080p at a 2 meter (6.5ft) distance, and essentially nobody could tell 720p from 1080p.
It is, of course, not unexpected, as the resolution of the average human eye tops out at about DVD resolution on a 32 inch screen at 2 meters distance (your capacity for that is very easy to demonstrate to yourself, simply make an about millimeter size black/white chessboard in a paintprogram, then back away and see what distance the pixels turn into gray (and remember, this is the best case with maximum contrast, far, far from the case with a moving picture and lower contrasts)). Increase screen size and you'll get a better ability to tell things apart, but the pixel resolution ability scales linearly with distance and screen size.
Many people simply need an eye upgrade to actually appreciate HD.
This is not to say I'm opposed to HD as a concept, I'm just saying it's not worth paying more for. HD quality TV's mean cheaper and larger monitors for computers, and various other situations where you actually can use the resolution.
It's just sad to see the lack of honesty about the issue, and it's also sad to see HD being touted as a great improvement when there are other factors like contrast and color that are much, _much_ more important for picture quality in home theatre situations.
Considering that the vast majority of p2p material isnt even SD I rather doubt it. Heck, ripping my own DVDs to disk, with a good encoder I dont notice enough quality difference to make it worth the 50 cents worth of diskspace to use full SD quality.
In what few blind tests I've seen not even experts can reliably tell what's HD content or upscaled SD content under normal viewing conditions, so why bother? You might as well stick a HD sticker on your old TV and be as amazed with the incredible improvement in picture.
"they can't just pick any PC game off the shelf and expect it to run on this computer as it is not a windows computer"
Of course, were it a 'windows computer' you still couldnt pick any game off the shelf and expect it to run.
"they will probably expect that they can buy something like Age of Empires or Civ II or whatever and be able to run it on their computer."
Of course, for those who want to run Civ II that PC is perfectly qualified to run Freeciv (which the cheap walmarter doesnt even have to buy!). In fact, in the range of games that hardware can be expected to support there is a selection of free Linux games that could easily have the walmart customer wasting a year or three.
The best way to get there is probably to reframe the debate in those terms, how to retain the incentive while allowing competition.
I doubt anyone today implementing an incentive system for innovation would have come up with the current system, but as it's there and it seems like we get it for 'free', appearing to cost the economy no more than the cost of the USPTO paperwork, there are simply so many (boring) things you need to get people to fundamentally reconsider before they rethink their positions.
Sometimes I think the easiest route and first step would be to convince government auditors to take up total economy patent royalties as a government budget item (a zero-sum item of royalties paid and royalties recieved), so we'd actually see the cost of the system. That would be the most difficult part to politically oppose, yet one of the most important for thinking about the problem in constructive terms, cost/benefit, who gets the money and who pays for it. Once that's on paper it becomes much easier to come up with and present creative solutions on how to get the best possible incentive for the money it costs us.
"Look, if I were to invent a super-efficient solar cell, how would I get it manufactured?"
You probably cant. Then again, if you want sole custody of the idea, then you'd have no option but to sit on it 'til the patent expires and get no benefit anyway, because nobody with the means to produce it could risk producing it either. For example, take a look at the huge number of display technologies available that dont get produced every year.
"Why spend years of your life working on something, only to have someone else reap the fruits of your labor?"
There are alternate solutions to the reaping of fruits, patents dont have to be the all-or-nothing approach of monopoly rights.
If we want to encourage invention, instead of granting a monopoly right, a patent could merely give the holder the right to a paycheck, for example modified by the number of sold products including that patent. Corporations, instead of having to negotiate a license, would simply report what inventions they use in their products (and would have no financial incentive not to report it).
Financing, instead of being a hidden tax included in licensing costs on new products could simply be a flat innovation VAT (start financing with calculating what the patent system costs today, and reset at the same level over all products). New technology would no longer be at a price disadvantage, so adoption rates of new and presumably better technology and medicines would increase and create a vast wealth increase in the economy.
If you want to reward innovation and innovators there are much better ways to do it than handing them a monopoly they probably cant use anyway. But then again, patents have never actually been about rewarding and promoting innovation, their purpose has been for those who have the financial muscle to use them to avoid competition.
It's really quite telling that we're handing innovators a paper only a lawyer has any use for instead of a check that anyone could use.
If you're after something more advanced than a bog standard cheapo NAS box, maybe Openfiler would be what you're looking for. It's quite a bit more capable than a cheapo NAS or WHS, but at least it's web manageable.
"Patents help to create that possibility - that the inventor will get to make a fortune"
Eh, if you look at the patent system today you'd do financially better serving fries with that and playing the lottery.
"I don't think you can assume that since there are many runners, the prize isn't important."
As the prize is mainly the right to trip anyone running in the next race you're unlikely to end up with a net benefit. Instead you end up with a lot of runners injured on the track, some prize winners offering protection rackets promising not to trip the runners who pay, and a whole lot of angry people.
As long as you retain the monopoly nature of the patent system it's unfixable. It's like arguing about how the Soviet state-run monopoly system could have been made to work if only comittee meetings were open and the public had an idea box.
It's not the ease or difficulty with which you can obtain a patent that causes the problem; it's the fact that it gives you the power to prevent anyone else from doing the same thing.
Fixing the patent system inevitably means you have to stop funding it by handing out monopoly taxation rights and instead fund it within the state budget. If we need to finance invention beyond the free market competetiveness incentives (which I'm not at all certain of), then we should actually put the pricetag on paper instead of pretending it doesnt cost just because it's hidden as a privatized taxation form.
Once the patent system (as an innovation incentive system) as a whole has an actual budget the rest of the problems are easy to work out, as every player in the system would suddenly have an incentive to see an equitable distribution and granting scheme.
"Genius is about intuition. It's about having a massive jumble in your head that you assemble into a coherent system by deduction, then test afterwards."
Of course, lock a genious in a cellar for fifty years of uninterrupted invention and what do you get?
I mean, really. If you had an average genious and stuck them in a cellar in 1957 and let them out today, would they have a bunch of marvellous inventions? Or would they have a bunch of stuff that would have been marvellous inventions in 1958?
Even the best genious requires the input of the entire world to create the massive jumble from which they take the intuitive leaps, and the progress of a million monkeys building upon eachothers advances inevitably outpaces the single geniouses.
"It's about protecting them from people like Tesla"
I'd say it's more about protecting them from competition and further progress. They could always libel, ostracise and marginalize people like Tesla. They might not be able to do the same thing to another businessman who took Teslas ideas and produced competetive and cheaper or better products.
Public advertising of an email address is just one (fast) way to get it harvested tho. Most are probably compromized by viral harvesters or intrusions and/or software misconfigurations (mailinglists, exchange, etc).
But as long as you just have a single active address, it will, sooner or later, get compromized and then it will just spread. And then you're faced with the painful choice of changing address or trying to filter. Having one address per person sending mail to you makes it trivial to change the address, and trivial to trace leaks and inform them of the problem.
Anyways, it's an easy way to avoid spam (for some, it requires you have control over your own mailserver and a client that supports choosing the right reply From: based on who I got the mail from, such as Horde).
The level of spam to any particular email address is strongly dependent on where that email address has been published.
Personally I have about 50 email addresses (merged to a single mailbox), yet most recieve no spam at all. That's because each address is distributed to only one or a few senders, so if one address is compromized I both know who compromized it and I can selectively disable it.
If your particular mailbox recieves little spam, then it simply hasnt been widely enough distributed yet. Put it on a website or two if you feel lonely.:)
Yep, I know, poor form to RTFA, but I wanted to see if it was something new and interesting.
But basically it looks like the ordinary newspeak about 'mind-doping'. So, people use amphetamines and derivatives to enhance mental performance. Well, duh. That's been done for half a century, and it has well known side effects. The only difference is that these days they're prescribed for ADD instead of as diet pills. And the college chem-students twisting a molecule here and there to keep ahead of the DEA have grown up are working for Big Pharma, twisting a molecule here and there to keep ahead of the FDA and USPTO.
I really dont see much news in this, other than the tendency to call it 'mind doping' and 'mind-enhancing' rather than 'doing drugs'.
"No, it's NOT a free market, nor has it ever been."
I think you need to re-read your Adam Smith.
There are a lot of non-free qualities of the current market, but the two you mention are explicitly pro-free market regulations. The classic purpose of the free market as a concept is to encourage competition. To quote the Wealth of Nations:
"The price of monopoly is upon every occasion the highest which can be got. The natural price, or the price of free competition, on the contrary, is the lowest which can be taken, not upon every occasion, indeed, but for any considerable time together."
Adam Smith was strongly opposed to the kind of monopolies and cartels that form in a completely unregulated market; he considered the main purpose of government in a free market to be to enforce its competetive nature and to regulate and intervene if and when any market player attempts to prevent that competetive freedom.
Feel free to claim that's not your idea of a 'free market' (or that there are much better examples of actual anti-free market interventions like 'intellectual property' or similar monopoly supporting legislation), but (enforced) free competition is the commonly recognized meaning of the term.
As a midterm storage technology, no. As a short-term information carrier it's horrible. The main reason I'm not subscribing to a paper isn't that the news is late or I couldnt spare the time to read a few articles.
It's that it piles up fast and it's a pain to carry to the recycling station.
Perhaps they could deliver it in toilet paper or soft tissue form; that way I'd have a use for it, and it'd be easily disposable after I'm done reading it.
"We need paid people that are going to do the legwork and investigation."
The run up to the Iraq war certainly put the nail in the coffin on that one. Any investigative journalism could, and would, punch holes wide enough to drive trucks through in every whitehouse press release. The appropriate response from legitimate investigative journalists would have been to punch those holes, expose the responsible people as liars and frauds, and perhaps avoided a lot of bloodshed and outright atrocities. Instead many just published the pressreleases like they were facts or 'news'.
If that's what paying for the investigation gives us I'd rather take the press releases straight up and recognized for what they are. And put my trust in the internet to tear them a new one. Which several sites did.
The real problem with newspapers is that the syndicated content is available anywhere for free as the amortized cost is close to zero. Op-ed, well, opinions, everyone's got one. And a substantial number of people will tell you theirs for free. Competition kills, eh. And, investigative reporting is as close to dead as it can get.
It's basic market economics; information isnt a scarce product. In fact, it's so oversupplied that the zero-cost alternatives produced every day would probably fill global consumer appetite for reading material for the next decade.
"They aren't going out of business, but the business is certainly changing."
A lot will. The aggregated copying and carrying inherent in the internet converts a local demand of the writing of a specific number of journalists into a global demand for the same specific number of journalists (modulated by some variance in demand and a few more languages). The game turned from several thousand different rooms playing musical chairs into one room with not that many chairs.
"Furthermore, I dislike your analogy in that it suggests false positives for DNA fingerprinting are in any way, related to race."
You're right, that was overly simplifying and skipping a number of steps. The more extensive reasoning goes like this; as subgroups and subcultures in urban settings often have a short and relatively close familial distribution these factors will strongly affect the statistical probability of matches within that subgroup. Not because of phenotype but because of close interrelations.
"can two people who encode their first and last names via MD5 hash have a collision?"
Apt comparison. Even more so when you consider the same problem in the same urban context; the likelyhood that two people in a closely interrelated group will have the same name, and end up with the same hash, is much higher than a completely random sample. It is, as you say, possible but extremely rare that you get collisions with utterly different names, but the distribution of names, like DNA, simply isnt that random in the real world.
So for law enforcement the problem remains; from a database of samples, when you get collisions you might get the utterly random ones, but you're far, far more likely to get false positives within the same geographical area and within the same subgroup.
For any specific purpose any piece of information is more or less relevant.
The problem with biometric data is that it isnt particularly unique. Biometry salesmen will try to convince you that their identifiers are special, but the fact of the matter is that evolution doesnt necessarily select for unique identifiers. We still have significant amounts of DNA in common with flatworms, nevermind other people.
All biometrics available today have atrocious error rates, in the range of fractions of percentages up to even whole percentages. That's ok when you have, for example, one print and one suspect, compare them and if it's not a match you now have zero suspects. But enter a database with 300 million americans, and you search for your fingerprint in it, and you get 50K matches. You now have fiftythousand suspects instead. For most purposes, apart from job security, that's even worse than zero suspects.
DNA is as bad, the current theoretical best case is the equivalent of taking a the DNA variant of a dozen traits like hair color, height, skin color, etc. If you have a dozen such variables and calculate the number of possible permutiations you'll get a huge number. But these are not random numbers, they are selected for and match to a much larger than random extent within population groups. Get a large enough database and you will find matches. Take a sample from a hispanic and you'll get a dozen other hispanics. Take a sample from a caucasian and you'll get a horde of white people.
And that's ignoring lab error rates which are bad enough to reach double-digit percentages in some cases (what, we're not supposed to mix the samples before running them???).
So fill your database with junk and you'll get huge amounts of junk at any time. It's the forensic equivalent of using google to search for information on a killer by the name of pr0n.
That's not to say biometrics and dna evidence are bad in every case. It's exceptionally useful when used for ruling people out.
And so, I can only conclude that biometric evidence should only be allowed to use for the defense.
Actually migrating some internal networks to v6 only would free up huge swaths of v4 adresses, so I'm not sure it would be that much of a problem by then.
"Will we see an emergence of gateway services"
Effectively, anyone could provide it, but personally I'd say it'd become an ISP or backbone issue.
"That would also require that all inbound services"
To some extent. Accessing legacy sites may be a requisite for migration, but allowing legacy access to your site may not be (compare HTTP 1 vs HTTP 1.1 webbrowsers, one could only access IP, the other named virtual hosts, yet the migration went fairly fast and quickly reached the cutoff point of upgrade-or-go-away). So, yes, that would also require legacy servicing for a while, but soon enough v6 capable client software would be a requirement for _some_ sites, leading to a rapid adoption and feedback loop as more clients have it and more servers feel the ability to require it.
"What operating system the CEO runs is irrelevant as long as he knows how to market a product and keep a company afloat."
Of course, a Redhat CEO who didnt run Linux would most likely be utterly incapable of marketing to that segment, and thus fail to keep the company afloat. Witness the multitude of spectacular failures at various other companies.
"In fact, maybe it would be a good thing to diversify"
Yes, that would be an example of the kinds of errors that lead to spectacular failures. Redhats customers aren't Redhat customers because they want a generic blend of proprietary crapplications. They could get that from any number of other vendors.
Say, about that, how do they diffrentiate between solicited and unsolicited traffic? I mean, if I decided to send a whole bunch of packets to some poor Telstra subscriber, would they actually be charged for those packets? Or is their entire network firewalled and inaccessible from the outside world?
The very idea that anyone on the internet could decide the size of your next bill would make a metered ISP a very dubious proposition imo.
"those speeds without tuning some network parameters or with some serious CPU and RAID setup."
Basic setup is approximately this; CPU's for both servers and clients range between AMD XP 3500+ to AMD X2 4800+. Motherboards are Asus (Nvidia 550 and AMD690) cards, with 2-4GB memory plus an extra SATA card on the iSCSI servers, and extra rtl8168/9 gigabit cards (the forcedeth driver has some issues). Disks on the iSCSI servers are striped with LVM, but not to more than 3 spindles (I dont care that much about maxing speed, I just want close-to-local disk performance). Tuning's been mainly on the iSCSI side with InitialR2T and ImmediateData. I've played around with network buffers, but basically come to the conclusion that it's more efficient in my case to throw RAM on the problem.
The peak rates (90-97MB per sec) have been obtained on completely unloaded systems, and the standard 40-60MB/s read is with disk mirrored against both iSCSI systems (and then striped on the iSCSI systems).
"I have a buddy that has done the same but with NFS"
Ah. Ehm. NFS. Yes.
Well, to tell the truth, I started out this interesting journey towards a home SAN using diskless systems booted over PXE and mounting NFS roots (mainly to silence my mythtv systems, and simplify backups). Lets just say that after testing and switching the first system over to iSCSI I could barely believe my eyes.
NFS is much, much, much harder to get decent performance out of. No matter how much tuning I've done I've rarely managed to get more than 20-40MB/s peaks, and for many small file accesses the performance is horrible. I'd thought that was what I could expect out of a gigabit Lan, after all, NFS saturated a 100Mbps network. I was quite surprised when my first iSCSI tests got 60-70MB/sec.
If your friends setup is such that block devices would work instead of NFS (at least for some parts), I'd really suggest he try running an iSCSI target. I cant vouch for the Solaris version, but I know there is one and that it sounds similar to the Linux one. Linux ietd (iscsi enterprise target) has been very stable and highly performing for me (more than a year running it now with no lost data:) ). It lacks some features like some forms of SCSI-3 reservation, but I can live with that.
Slashdot Mirror
"What's good for the n00b might not be right for a poweruser."
Absolutely. The poweruser would be very annoyed if functions were lacking. And remotes, such as shipped with most devices, are notoriously unconfigurable.
So I'll propose a much simpler solution. Send a 'simple' remote with the devices in question, make the device itself capable of outputting its codes for programming (like, an OSD with send-code for each function, or even better, a _standard_ for rapidly programming remotes, or even better than that, a standard for remote codes), clearly write the function codes in the manual, and make sure the specs are widely published and available on the net.
Todays power user will most likely throw the 52-button remote in a box in the closet anyway; they have a much nicer programmable one that actually handles all of their devices.
"Maybe not him but Christie's should definitely get stung for a large wad of cash."
More appropriately, they should be required to, for each item auctioned, state that the item they're selling may or may not be authentic and as far as they're concerned they make more money if you think it is.
"IIRC, their business model rather depends upon the legitimacy of the items going under the hammer."
The actual legitimacy isnt as importance as the appearance of legitimacy. Which may be why auctioning houses have a long and profitable history of making mistakes.
There's a whole host of printing technologies adapted to low/easy maintenance type situations.
Think thermal printers used in point-of-sale equipment. Not inkjet or office laser printer.
Which is why you need scientific experiments to verify that you're not imagining things.
In the case of Microsoft there have been several courts engaged in the legal equivalent, establishing that those patterns do exist.
"but do think of me as quite able to tell the difference between HD and SD content."
:) It's not much harder than a script around mplayer and a randomizer.
Well, do a blindtest on yourself and see
It does, of course, depend a lot on conditions and how you do the test. One of the tests I saw began with 3.5 meters (11.5 ft) distance, 42 inch LCD, starting with 480p. Everyone in the panel thought it was 720p. After several switches back and forth between different modes some became more capable of distinguising what was what. Further down the row of tests there were still a fair number that couldnt even tell 480p apart from 1080p at a 2 meter (6.5ft) distance, and essentially nobody could tell 720p from 1080p.
It is, of course, not unexpected, as the resolution of the average human eye tops out at about DVD resolution on a 32 inch screen at 2 meters distance (your capacity for that is very easy to demonstrate to yourself, simply make an about millimeter size black/white chessboard in a paintprogram, then back away and see what distance the pixels turn into gray (and remember, this is the best case with maximum contrast, far, far from the case with a moving picture and lower contrasts)). Increase screen size and you'll get a better ability to tell things apart, but the pixel resolution ability scales linearly with distance and screen size.
Many people simply need an eye upgrade to actually appreciate HD.
This is not to say I'm opposed to HD as a concept, I'm just saying it's not worth paying more for. HD quality TV's mean cheaper and larger monitors for computers, and various other situations where you actually can use the resolution.
It's just sad to see the lack of honesty about the issue, and it's also sad to see HD being touted as a great improvement when there are other factors like contrast and color that are much, _much_ more important for picture quality in home theatre situations.
"Build it and they will come..."
Considering that the vast majority of p2p material isnt even SD I rather doubt it. Heck, ripping my own DVDs to disk, with a good encoder I dont notice enough quality difference to make it worth the 50 cents worth of diskspace to use full SD quality.
In what few blind tests I've seen not even experts can reliably tell what's HD content or upscaled SD content under normal viewing conditions, so why bother? You might as well stick a HD sticker on your old TV and be as amazed with the incredible improvement in picture.
"they can't just pick any PC game off the shelf and expect it to run on this computer as it is not a windows computer"
Of course, were it a 'windows computer' you still couldnt pick any game off the shelf and expect it to run.
"they will probably expect that they can buy something like Age of Empires or Civ II or whatever and be able to run it on their computer."
Of course, for those who want to run Civ II that PC is perfectly qualified to run Freeciv (which the cheap walmarter doesnt even have to buy!). In fact, in the range of games that hardware can be expected to support there is a selection of free Linux games that could easily have the walmart customer wasting a year or three.
Wish there was a petition :).
The best way to get there is probably to reframe the debate in those terms, how to retain the incentive while allowing competition.
I doubt anyone today implementing an incentive system for innovation would have come up with the current system, but as it's there and it seems like we get it for 'free', appearing to cost the economy no more than the cost of the USPTO paperwork, there are simply so many (boring) things you need to get people to fundamentally reconsider before they rethink their positions.
Sometimes I think the easiest route and first step would be to convince government auditors to take up total economy patent royalties as a government budget item (a zero-sum item of royalties paid and royalties recieved), so we'd actually see the cost of the system. That would be the most difficult part to politically oppose, yet one of the most important for thinking about the problem in constructive terms, cost/benefit, who gets the money and who pays for it. Once that's on paper it becomes much easier to come up with and present creative solutions on how to get the best possible incentive for the money it costs us.
"Look, if I were to invent a super-efficient solar cell, how would I get it manufactured?"
You probably cant. Then again, if you want sole custody of the idea, then you'd have no option but to sit on it 'til the patent expires and get no benefit anyway, because nobody with the means to produce it could risk producing it either. For example, take a look at the huge number of display technologies available that dont get produced every year.
"Why spend years of your life working on something, only to have someone else reap the fruits of your labor?"
There are alternate solutions to the reaping of fruits, patents dont have to be the all-or-nothing approach of monopoly rights.
If we want to encourage invention, instead of granting a monopoly right, a patent could merely give the holder the right to a paycheck, for example modified by the number of sold products including that patent. Corporations, instead of having to negotiate a license, would simply report what inventions they use in their products (and would have no financial incentive not to report it).
Financing, instead of being a hidden tax included in licensing costs on new products could simply be a flat innovation VAT (start financing with calculating what the patent system costs today, and reset at the same level over all products). New technology would no longer be at a price disadvantage, so adoption rates of new and presumably better technology and medicines would increase and create a vast wealth increase in the economy.
If you want to reward innovation and innovators there are much better ways to do it than handing them a monopoly they probably cant use anyway. But then again, patents have never actually been about rewarding and promoting innovation, their purpose has been for those who have the financial muscle to use them to avoid competition.
It's really quite telling that we're handing innovators a paper only a lawyer has any use for instead of a check that anyone could use.
If you're after something more advanced than a bog standard cheapo NAS box, maybe Openfiler would be what you're looking for. It's quite a bit more capable than a cheapo NAS or WHS, but at least it's web manageable.
"Patents help to create that possibility - that the inventor will get to make a fortune"
Eh, if you look at the patent system today you'd do financially better serving fries with that and playing the lottery.
"I don't think you can assume that since there are many runners, the prize isn't important."
As the prize is mainly the right to trip anyone running in the next race you're unlikely to end up with a net benefit. Instead you end up with a lot of runners injured on the track, some prize winners offering protection rackets promising not to trip the runners who pay, and a whole lot of angry people.
"If you want to fix the patent system"
As long as you retain the monopoly nature of the patent system it's unfixable. It's like arguing about how the Soviet state-run monopoly system could have been made to work if only comittee meetings were open and the public had an idea box.
It's not the ease or difficulty with which you can obtain a patent that causes the problem; it's the fact that it gives you the power to prevent anyone else from doing the same thing.
Fixing the patent system inevitably means you have to stop funding it by handing out monopoly taxation rights and instead fund it within the state budget. If we need to finance invention beyond the free market competetiveness incentives (which I'm not at all certain of), then we should actually put the pricetag on paper instead of pretending it doesnt cost just because it's hidden as a privatized taxation form.
Once the patent system (as an innovation incentive system) as a whole has an actual budget the rest of the problems are easy to work out, as every player in the system would suddenly have an incentive to see an equitable distribution and granting scheme.
"Genius is about intuition. It's about having a massive jumble in your head that you assemble into a coherent system by deduction, then test afterwards."
Of course, lock a genious in a cellar for fifty years of uninterrupted invention and what do you get?
I mean, really. If you had an average genious and stuck them in a cellar in 1957 and let them out today, would they have a bunch of marvellous inventions? Or would they have a bunch of stuff that would have been marvellous inventions in 1958?
Even the best genious requires the input of the entire world to create the massive jumble from which they take the intuitive leaps, and the progress of a million monkeys building upon eachothers advances inevitably outpaces the single geniouses.
"It's about protecting them from people like Tesla"
I'd say it's more about protecting them from competition and further progress. They could always libel, ostracise and marginalize people like Tesla. They might not be able to do the same thing to another businessman who took Teslas ideas and produced competetive and cheaper or better products.
"They are uniformly bad for progress."
Without a doubt.
Public advertising of an email address is just one (fast) way to get it harvested tho. Most are probably compromized by viral harvesters or intrusions and/or software misconfigurations (mailinglists, exchange, etc).
But as long as you just have a single active address, it will, sooner or later, get compromized and then it will just spread. And then you're faced with the painful choice of changing address or trying to filter. Having one address per person sending mail to you makes it trivial to change the address, and trivial to trace leaks and inform them of the problem.
Anyways, it's an easy way to avoid spam (for some, it requires you have control over your own mailserver and a client that supports choosing the right reply From: based on who I got the mail from, such as Horde).
"Rules of statistics tell me is grossly inflated"
:)
The level of spam to any particular email address is strongly dependent on where that email address has been published.
Personally I have about 50 email addresses (merged to a single mailbox), yet most recieve no spam at all. That's because each address is distributed to only one or a few senders, so if one address is compromized I both know who compromized it and I can selectively disable it.
If your particular mailbox recieves little spam, then it simply hasnt been widely enough distributed yet. Put it on a website or two if you feel lonely.
Yep, I know, poor form to RTFA, but I wanted to see if it was something new and interesting.
But basically it looks like the ordinary newspeak about 'mind-doping'. So, people use amphetamines and derivatives to enhance mental performance. Well, duh. That's been done for half a century, and it has well known side effects. The only difference is that these days they're prescribed for ADD instead of as diet pills. And the college chem-students twisting a molecule here and there to keep ahead of the DEA have grown up are working for Big Pharma, twisting a molecule here and there to keep ahead of the FDA and USPTO.
I really dont see much news in this, other than the tendency to call it 'mind doping' and 'mind-enhancing' rather than 'doing drugs'.
"No, it's NOT a free market, nor has it ever been."
I think you need to re-read your Adam Smith.
There are a lot of non-free qualities of the current market, but the two you mention are explicitly pro-free market regulations. The classic purpose of the free market as a concept is to encourage competition. To quote the Wealth of Nations:
"The price of monopoly is upon every occasion the highest which can be got. The natural price, or the price of free competition, on the contrary, is the lowest which can be taken, not upon every occasion, indeed, but for any considerable time together."
Adam Smith was strongly opposed to the kind of monopolies and cartels that form in a completely unregulated market; he considered the main purpose of government in a free market to be to enforce its competetive nature and to regulate and intervene if and when any market player attempts to prevent that competetive freedom.
Feel free to claim that's not your idea of a 'free market' (or that there are much better examples of actual anti-free market interventions like 'intellectual property' or similar monopoly supporting legislation), but (enforced) free competition is the commonly recognized meaning of the term.
"Paper isn't such a bad technology."
As a midterm storage technology, no. As a short-term information carrier it's horrible. The main reason I'm not subscribing to a paper isn't that the news is late or I couldnt spare the time to read a few articles.
It's that it piles up fast and it's a pain to carry to the recycling station.
Perhaps they could deliver it in toilet paper or soft tissue form; that way I'd have a use for it, and it'd be easily disposable after I'm done reading it.
"We need paid people that are going to do the legwork and investigation."
The run up to the Iraq war certainly put the nail in the coffin on that one. Any investigative journalism could, and would, punch holes wide enough to drive trucks through in every whitehouse press release. The appropriate response from legitimate investigative journalists would have been to punch those holes, expose the responsible people as liars and frauds, and perhaps avoided a lot of bloodshed and outright atrocities. Instead many just published the pressreleases like they were facts or 'news'.
If that's what paying for the investigation gives us I'd rather take the press releases straight up and recognized for what they are. And put my trust in the internet to tear them a new one. Which several sites did.
The real problem with newspapers is that the syndicated content is available anywhere for free as the amortized cost is close to zero. Op-ed, well, opinions, everyone's got one. And a substantial number of people will tell you theirs for free. Competition kills, eh. And, investigative reporting is as close to dead as it can get.
It's basic market economics; information isnt a scarce product. In fact, it's so oversupplied that the zero-cost alternatives produced every day would probably fill global consumer appetite for reading material for the next decade.
"They aren't going out of business, but the business is certainly changing."
A lot will. The aggregated copying and carrying inherent in the internet converts a local demand of the writing of a specific number of journalists into a global demand for the same specific number of journalists (modulated by some variance in demand and a few more languages). The game turned from several thousand different rooms playing musical chairs into one room with not that many chairs.
"Furthermore, I dislike your analogy in that it suggests false positives for DNA fingerprinting are in any way, related to race."
You're right, that was overly simplifying and skipping a number of steps. The more extensive reasoning goes like this; as subgroups and subcultures in urban settings often have a short and relatively close familial distribution these factors will strongly affect the statistical probability of matches within that subgroup. Not because of phenotype but because of close interrelations.
"can two people who encode their first and last names via MD5 hash have a collision?"
Apt comparison. Even more so when you consider the same problem in the same urban context; the likelyhood that two people in a closely interrelated group will have the same name, and end up with the same hash, is much higher than a completely random sample. It is, as you say, possible but extremely rare that you get collisions with utterly different names, but the distribution of names, like DNA, simply isnt that random in the real world.
So for law enforcement the problem remains; from a database of samples, when you get collisions you might get the utterly random ones, but you're far, far more likely to get false positives within the same geographical area and within the same subgroup.
"The data is relevant, don't kid yourself."
For any specific purpose any piece of information is more or less relevant.
The problem with biometric data is that it isnt particularly unique. Biometry salesmen will try to convince you that their identifiers are special, but the fact of the matter is that evolution doesnt necessarily select for unique identifiers. We still have significant amounts of DNA in common with flatworms, nevermind other people.
All biometrics available today have atrocious error rates, in the range of fractions of percentages up to even whole percentages. That's ok when you have, for example, one print and one suspect, compare them and if it's not a match you now have zero suspects. But enter a database with 300 million americans, and you search for your fingerprint in it, and you get 50K matches. You now have fiftythousand suspects instead. For most purposes, apart from job security, that's even worse than zero suspects.
DNA is as bad, the current theoretical best case is the equivalent of taking a the DNA variant of a dozen traits like hair color, height, skin color, etc. If you have a dozen such variables and calculate the number of possible permutiations you'll get a huge number. But these are not random numbers, they are selected for and match to a much larger than random extent within population groups. Get a large enough database and you will find matches. Take a sample from a hispanic and you'll get a dozen other hispanics. Take a sample from a caucasian and you'll get a horde of white people.
And that's ignoring lab error rates which are bad enough to reach double-digit percentages in some cases (what, we're not supposed to mix the samples before running them???).
So fill your database with junk and you'll get huge amounts of junk at any time. It's the forensic equivalent of using google to search for information on a killer by the name of pr0n.
That's not to say biometrics and dna evidence are bad in every case. It's exceptionally useful when used for ruling people out.
And so, I can only conclude that biometric evidence should only be allowed to use for the defense.
"But what happens in 2011"
Actually migrating some internal networks to v6 only would free up huge swaths of v4 adresses, so I'm not sure it would be that much of a problem by then.
"Will we see an emergence of gateway services"
Effectively, anyone could provide it, but personally I'd say it'd become an ISP or backbone issue.
"That would also require that all inbound services"
To some extent. Accessing legacy sites may be a requisite for migration, but allowing legacy access to your site may not be (compare HTTP 1 vs HTTP 1.1 webbrowsers, one could only access IP, the other named virtual hosts, yet the migration went fairly fast and quickly reached the cutoff point of upgrade-or-go-away). So, yes, that would also require legacy servicing for a while, but soon enough v6 capable client software would be a requirement for _some_ sites, leading to a rapid adoption and feedback loop as more clients have it and more servers feel the ability to require it.
"What operating system the CEO runs is irrelevant as long as he knows how to market a product and keep a company afloat."
Of course, a Redhat CEO who didnt run Linux would most likely be utterly incapable of marketing to that segment, and thus fail to keep the company afloat. Witness the multitude of spectacular failures at various other companies.
"In fact, maybe it would be a good thing to diversify"
Yes, that would be an example of the kinds of errors that lead to spectacular failures. Redhats customers aren't Redhat customers because they want a generic blend of proprietary crapplications. They could get that from any number of other vendors.
"The ISP in question meters its users' usage"
Say, about that, how do they diffrentiate between solicited and unsolicited traffic? I mean, if I decided to send a whole bunch of packets to some poor Telstra subscriber, would they actually be charged for those packets? Or is their entire network firewalled and inaccessible from the outside world?
The very idea that anyone on the internet could decide the size of your next bill would make a metered ISP a very dubious proposition imo.
Mmm. Troll (but as there are real 'studies' that make the same error, I'll point it out). Your links say this at the top of the pages:
View Topics > Underlying OS > Linux (Any)
View Topics > Category > OS (Microsoft)
You're comparing security issues in applications that run on linux with security issues in Windows itself.
"those speeds without tuning some network parameters or with some serious CPU and RAID setup."
:) ). It lacks some features like some forms of SCSI-3 reservation, but I can live with that.
Basic setup is approximately this; CPU's for both servers and clients range between AMD XP 3500+ to AMD X2 4800+. Motherboards are Asus (Nvidia 550 and AMD690) cards, with 2-4GB memory plus an extra SATA card on the iSCSI servers, and extra rtl8168/9 gigabit cards (the forcedeth driver has some issues). Disks on the iSCSI servers are striped with LVM, but not to more than 3 spindles (I dont care that much about maxing speed, I just want close-to-local disk performance). Tuning's been mainly on the iSCSI side with InitialR2T and ImmediateData. I've played around with network buffers, but basically come to the conclusion that it's more efficient in my case to throw RAM on the problem.
The peak rates (90-97MB per sec) have been obtained on completely unloaded systems, and the standard 40-60MB/s read is with disk mirrored against both iSCSI systems (and then striped on the iSCSI systems).
"I have a buddy that has done the same but with NFS"
Ah. Ehm. NFS. Yes.
Well, to tell the truth, I started out this interesting journey towards a home SAN using diskless systems booted over PXE and mounting NFS roots (mainly to silence my mythtv systems, and simplify backups). Lets just say that after testing and switching the first system over to iSCSI I could barely believe my eyes.
NFS is much, much, much harder to get decent performance out of. No matter how much tuning I've done I've rarely managed to get more than 20-40MB/s peaks, and for many small file accesses the performance is horrible. I'd thought that was what I could expect out of a gigabit Lan, after all, NFS saturated a 100Mbps network. I was quite surprised when my first iSCSI tests got 60-70MB/sec.
If your friends setup is such that block devices would work instead of NFS (at least for some parts), I'd really suggest he try running an iSCSI target. I cant vouch for the Solaris version, but I know there is one and that it sounds similar to the Linux one. Linux ietd (iscsi enterprise target) has been very stable and highly performing for me (more than a year running it now with no lost data