I wonder if Roman ships may have been innately more flammable than that 80 year old boat. The use of tar or pitch to seal rough-hewn planks on the sides of the Roman ships would have made them more susceptible to fire. Any oiled cloth would also have made these ancient boats more flammable.
With only a few k of memory and a few thousand or few tens of thousands of transistors, these old machine were 100% comprehensible. A hobbyist could readily learn the purpose and functioning of every instruction, every chip, and every circuit trace. In contrast, modern machines are largely inscrutable black boxes with millions of lines of code in deeply layered architectures.
I'l gladly give up knowledge of 100% of the internals in exchange for the power of OS X on a G5, but those old machines do provide a pleasant simplicity.
Flu viruses can evolve resistance to antivirals. Already, some flu drugs have less effectiveness because of presumed overuse in Asia. If Tamiflu becomes as cheap as the other flu medications it will be dispensed too much. It might even encourage China to mass produce the drug for use in livestock - a factor that may have contributed to increased viral resistance to older antivirals.
The moral dimensions to mass produced antibiotics and antivirals are more complex than just the issue of patents.
If you have an old (pre-USB) computer and an old dot matrix printer, then you can use it to create two axes of motion control for moving dismembered hands, dangling spiders, etc. Sending the right bytes out through the serial or parallel port will cause the print head and feed rollers to move on command. For extra credit, one can wire switches into a keyboard so that the system detects when a door opens, etc. With a bit of programming, the system can be quite interactive.
The modern stuff is not as easy to hack in this way because its hard to talk directly to the hardware -- too many drivers, libraries, and embedded smarts between the CPU and the printer's motors.
Physical disks are just a means to an end. Why buy a physical disk player and physical disks when bandwidth provides the same experience? Physical formats add bulk without adding much value (in most cases). I'd bet that most people want the content and relatively few people want the artifact.
I wonder how CD player and disk sales are doing? Last I heard both were flat or declining. Once people realized that they wanted their music on an iPod, the CD became an added hassle. The same process will occur with DVDs.
But DVDs won't die for 10-20 years because some collectors will be willing to pay handsomely for the "Extended Platinum Director's Super Secret Cut Anniversary Re-release edition with matching book-ends." What will occur is that fewer B-list titles will appear on DVD because video-on-demand/pay-per-view/download services will offer a larger play list with lower distribution costs.
Most people don't need (or want) the functionality of a PDA. They don't need a contact manager, mobile calendar program, wireless web browser, email, document reader, word processor, etc. Yes, some fraction of the world (road warriors and geeks) may want a multifunctional highly utilitarian device, but most people can't even program their VCRs and just want to carry around some music.
The iPod sells because people do want portable entertainment. PDAs are too much work and remind people too much of work. Personally, I love my Psion 5mx but its obvious that I am part of an economically unsustainable niche market.
Social bookmarking would seem to be an ideal target for spammers and other malefactors of the net. How do systems such as Flock keep spammers from touting commercial links?
These systems would also make ideal phishing grounds. Posting a fake "eBay" link ("look at this cool auction!!!") would take the target person to a faked eBay auction page (e.g with an IDN exploit) or just a scam domain (ebbay.com, etc.) that then asks for a eBay or Paypal password. Since many of the people that would follow a socially bookmarked eBay link are eBay/Paypal users the phisher would get a high hit rate.
Even if the system relies on some form of accumulated reputation or trust networks, its still possible for someone to cultivate a great reputation before abusing the system with spam or phishing.
This study of $384 billion in R&D spending by 1,000 companies finds no correlation between R&D spending or patents and a company's growth, profitability, or shareholder return. Part of the problem probably stems from too much R and too little D.
What's interesting is that companies with extremely strong R&D foundations such as IBM and Lucent haven't done as well as low R&D companies such as Dell or Wal-Mart. Companies such as Dell and Wal-Mart show the power of very tightly managed business processes without a lot of the traditional science-based R&D.
I'm not saying that new materials aren't essential to the future, only that these new materials are useless without highly efficient business processes to commercialize them. I hope that space race R&D takes this fact into account.
Once the code is cracked, anyone can add a pattern of yellow dots that say anything. Assuming someone can tweeze the overlapping codes, they would discover that the document was printed 10/10/05 by printer 2721272 or 5/8/05 by printer 8798798 or 11/2/05 by printer 9813982, etc. If one can get the alignment right, one could even fill-in the printer's native dot pattern so that all pages are printed on FF/FF/FF by printer FFFFFFF.
It's easy to buy enough flash NVRAM to equal the RAM of a PC (1 GB is well under $100 retail). The only real issue is the read bandwidth (it need to be at least 100 MB/sec to load RAM quickly enough. Write bandwidth is less of an issue if the user doesn't wait around "while Windows is shutting down" (the computer might "sleep" instantly and off-load RAM over a period of minutes). And read/write cycle life is a non-issue if you don't turn-off the computer more than a half-dozen times per day (27 cycles per day for 10 years is 100,000 cycles).
I wonder if Intel has created a Flash memory architecture that has massive internal parallelism on the read circuits of the flash RAM cells to feed a high-speed interface to RAM? It might read on the order of a thousand bits in parallel at slow speed (say 1 Mb/sec on each line) and reassemble in to output at high speed (1000 Mb/sec = 125 MB/sec). Seems feasible to me.
Tabbed browsing is a miracle for the time-starved. I love to throw up a bunch of pages in tabs loading in the background and then visit each one AFTER it loads -- no waiting while someone's painful graphics loads from some battered server. I often quickly scan a site for likely follow-on links and start those loading before I fully read the page I'm on. I hate hate hate hate sites that use javascript or Flash navigation that interferes with Cmd-clicking a link to open it in another background tab.
And if I really need to concentrate I pull the power plug on the broadband modem or take a non-WiFi laptop out on the deck.
Why bitching about Google maps is bad
on
Google Terror Threat
·
· Score: 4, Insightful
Before governments publicized their worries about terrorists looking at maps of sensitive places, the government could probably make use (subpoenaed/secret/coerced, etc.) access to Google's logs to see who was trying to peek at these places. Cross-referencing anyone who tried to look at super secret "nuclear installation Q-345" with other data might help the government find terrorist cells. Now, after the government complains, 100,000 normal folks immediately go to Google Earth to try to find these sensitive locations and pollute the access log files.
I don't really know what you are talking about. Doctors go to work every day. They see patients and evaluate the effectiveness of their treatments and meticulously document their opinions. Where does the additional cost of new treatments come from besides collecting those opinions? I have yet to run into them at the cancer treatment research facility where I work.
I see your point. I do know that Phase 1 clinical trials, for pharma products in the U.S., involve testing on healthy volunteers (usually on the order of 100 to 1000 people). These are people seen by doctors paid to do just the clinical trial (not other treatments) so the test subject (or their insurance company) is not the payor. So 100% of the Phase I costs are paid by the company with the new product because the doctors aren't doing it in the course of routine treatment.
Later phases involve patients that do need treatment. I suspect this creates some added cost due to added documentation that goes above and beyond the usual treatment regime and medical records keeping. But the bigger issue is that I doubt the patient's insurance company (in the U.S.) would reimburse for experimental treatments (I don't know how Medicare or other countries handle this). Thus the full cost of treatment probably gets paid by the company with the experimental drug. I'm sure cancer treatment isn't cheap (you may know this better than I), so covering the treatment costs on thousands of patients is quite costly.
Does your facility do clinical trials? I'd be interested in what the doctors say about who pays for what when a clinical trial is involved.
A cheap gadget like this will quickly be tested in every conceivable way by hungry graduate students at every University in existence like TLDs were.
I'm sure that grad students can do so valuable preclinical work -- showing that exposure to this devices doesn't cause cancer in human cell cultures, showing that the device kills X% of type Y bacteria with Z seconds of exposure, etc. But that won't get the FDA's approval. Where do the grad students get the money to test the device on 1,000, 10,000, or 100,000 patients? Clinical testing is expensive. It requires real doctors on real salaries and if far beyond the scope of the labors of a grad student.
There's enough red tape as it is. Please don't make me go Federal for everything.
I agree 100%. Personally, I think the current system of patents on medical products is broken, but that a government-funded approach would be far worse for the reasons that you state.
But if the government does not pay for clinical testing, who does? And if some private company pays for it, how do they get paid? If large scale clinical testing is a prerequisite for ensuring the safety and efficiacy of medical products, then we either need a system of government funding (bad) or patent/government protection for those that spend the money on clinical testing (bad). I'll not suggest the option of dispensing with clinical testing and letting anyone sell anything in the medical arena.
The greater issue is that medical products aren't like TLDs or OSS. Medical devices have potentially lethal effects and are highly regulated. Gaining approval is much more costly than getting a few people to try your software or arguing a case before a standards body or OSS development team.
A device such as this will require clinical testing to prove that it is both safe and effective. Those tests will take on the order of 2 to 5 years and cost on the order of $25 to $200 million for each proposed use.
Who is going to spend that kind of money if the minute they get approval, some other company can sell these devices without the clinical testing costs? The company that performed the tests will need to add $25 to $200 to the price of the device (in addition to manufacturing costs), assuming they sell a million of them. And the competitor will be able to undercut the first company on price.
The math is even worse on a risk-adjusted basis because so many promising products fail during testing. Thus, the costs of developing several failed devices must be paid for by each successful device.
Until governments foot the bill for all medical R&D and clinical testing, patents are a crucial part of the medical device & pharma industry.
The point is that without a patent, nobody will pay for testing, the device will sit on a shelf, and it will do no one any good. This is why pharma and medical devices will never be like OSS -- the invention of the first instance is an extremely minor part of the cost of development. Building a better medical mousetrap is nothing. Proving it is safe and effective and gaining govt approval is everything.
And if you are lugging a carry-on and a laptop bag on the way to your flight, what stops the phone from deciding you are not you because the added weight changes your gait? TFA said the false alarm (accidental lockout) rate was 4%. I'd bet the rate is much worse if you are carrying something (suitcase, kid, groceries, etc.)
Next Up: Zoning, Public Right of Way
on
PCs Posted No Trespass
·
· Score: 1, Interesting
If PCs are like private real estate, then we could be opening a pandora's box of real estate, zoning, and property law issues. If a PC is like a piece of land with built structures, then there is both the issue of trespass and civic responsibility. Zoning laws might limit how much bandwidth a person take from a shared network infrastructure. Subnets might create their own PC-owners associations (fashioned on Home-Owners Associations) that restrict activity to avoid inconveniencing others in the virtual neighborhood (good and bad).
This may not be all bad. Perhaps some zombie PCs could be condemned under such an interpretation. Or, for example, if people create municipal mesh networks, then anyone joining the network may be required to provide public access to their part of the network for routing purposes (no leaching).
My point is that real estate/private property is heavily regulated and that entrenching the PC = private property analogy in case law could have some interesting legal consequences.
Does anyone know if these patents cover genes (a particular location on the DNA) or alleles (a specific variant that this found for some gene)? If the patent covers a specific DNA sequence, then it is an allele. If it covers an allele, then the number of possible patents is much larger than the number of genes.
Some cellphone service provider is going to make a killing on a voice-activated old-school RPG. The player speaks their commands ("move forward" "pick up crystal") and a pleasant synth or sampled voice tells the player what they see ("You are in a dark forest and the sun is setting. In front of you lies a shining sword and a bulging bag. A trail leads forward toward a crumbling castle").
Just think of the minutes burned as the "caller" explores some world/dungeon for hours.
Just turn on the broadcast flag so the visual data can't be copied.
That's only slightly tongue-in-cheek. (Yes, I know that between all the holes in the OS and all the holes in user's heads that screen-loggers will get installed with admin privileges.)
As much as I hate DRM ("lets assume 100% of computer users are illegal content distributors" and inconvenience everyone), it seems that it could be useful as part of locking down a machine from copying selected types of data to unauthorized external locations.
Why money is enough
on
Space Tourism?
·
· Score: 4, Interesting
Anything that gets more people excited by space and space travel is good. Space needs better marketing -- yet another scientific experiment in space doesn't capture the imagination. But marketing is expensive unless you can get free publicity -- I can't see Congress giving NASA the OK to put on a $100 million ad campaign.
If letting a space tourist go up can attract media attention, then that's great. Its even nice that the customer pays the organization to create good marketing for the organization.
Besides, I'd bet the economics of space flight are such that the cost of filling an empty seat aren't that high. The average cost of putting a pound into orbit may be extremely high, but the cost of adding another pound of person and supplies is probably not bad. It's like the airlines -- if you're going to fly anyway, why not fill every seat.
Creating the idea that space is accessible to an increasing number of people -- not just a few astronauts that spent their life in the program -- is the key to the future of space funding.
I would argue that mutual respect one key to a long-term relationship and that tests like this could help determine
what qualities a person has that are respectable
what qualities a person considers in bestowing respect.
It could be intelligence, knowledge on any of a number of dimensions, social grace, physical strength, affection, aggressiveness, niceness, humor, ambition, earning-power, etc.
Disclaimer: I've been married nearly 22 years so that means I either know what I'm talking about or have an insufficient sample size to comment on this.
The Sun's magnetic field may be very weak (about 5 Gauss at the surface, about 0.00005 Gauss in solar wind), but it's very big. Creating a field with a compact object (say 100 meters in diameter -- quite a large space craft!) that creates a 0.00005 Gauss field at a distance of 160 million kilometers would require a field strength on the order of about 1.28 x 10^18 Gauss. This is NOT compatible with living things. Fields stronger than 100,000 Gauss can levitate living things. I suspect that the needed deflector field would strip the electrons off the spacecraft's atoms (even a 200,000 gauss magnets have a tendency to explode).
Even if I'm off by many orders of magnitude (IANAP), the required field strength will be unattainably high.
Slashdot Mirror
I wonder if Roman ships may have been innately more flammable than that 80 year old boat. The use of tar or pitch to seal rough-hewn planks on the sides of the Roman ships would have made them more susceptible to fire. Any oiled cloth would also have made these ancient boats more flammable.
I'l gladly give up knowledge of 100% of the internals in exchange for the power of OS X on a G5, but those old machines do provide a pleasant simplicity.
The moral dimensions to mass produced antibiotics and antivirals are more complex than just the issue of patents.
The modern stuff is not as easy to hack in this way because its hard to talk directly to the hardware -- too many drivers, libraries, and embedded smarts between the CPU and the printer's motors.
I wonder how CD player and disk sales are doing? Last I heard both were flat or declining. Once people realized that they wanted their music on an iPod, the CD became an added hassle. The same process will occur with DVDs.
But DVDs won't die for 10-20 years because some collectors will be willing to pay handsomely for the "Extended Platinum Director's Super Secret Cut Anniversary Re-release edition with matching book-ends." What will occur is that fewer B-list titles will appear on DVD because video-on-demand/pay-per-view/download services will offer a larger play list with lower distribution costs.
The iPod sells because people do want portable entertainment. PDAs are too much work and remind people too much of work. Personally, I love my Psion 5mx but its obvious that I am part of an economically unsustainable niche market.
These systems would also make ideal phishing grounds. Posting a fake "eBay" link ("look at this cool auction!!!") would take the target person to a faked eBay auction page (e.g with an IDN exploit) or just a scam domain (ebbay.com, etc.) that then asks for a eBay or Paypal password. Since many of the people that would follow a socially bookmarked eBay link are eBay/Paypal users the phisher would get a high hit rate.
Even if the system relies on some form of accumulated reputation or trust networks, its still possible for someone to cultivate a great reputation before abusing the system with spam or phishing.
What's interesting is that companies with extremely strong R&D foundations such as IBM and Lucent haven't done as well as low R&D companies such as Dell or Wal-Mart. Companies such as Dell and Wal-Mart show the power of very tightly managed business processes without a lot of the traditional science-based R&D.
I'm not saying that new materials aren't essential to the future, only that these new materials are useless without highly efficient business processes to commercialize them. I hope that space race R&D takes this fact into account.
Once the code is cracked, anyone can add a pattern of yellow dots that say anything. Assuming someone can tweeze the overlapping codes, they would discover that the document was printed 10/10/05 by printer 2721272 or 5/8/05 by printer 8798798 or 11/2/05 by printer 9813982, etc. If one can get the alignment right, one could even fill-in the printer's native dot pattern so that all pages are printed on FF/FF/FF by printer FFFFFFF.
I wonder if Intel has created a Flash memory architecture that has massive internal parallelism on the read circuits of the flash RAM cells to feed a high-speed interface to RAM? It might read on the order of a thousand bits in parallel at slow speed (say 1 Mb/sec on each line) and reassemble in to output at high speed (1000 Mb/sec = 125 MB/sec). Seems feasible to me.
And if I really need to concentrate I pull the power plug on the broadband modem or take a non-WiFi laptop out on the deck.
Before governments publicized their worries about terrorists looking at maps of sensitive places, the government could probably make use (subpoenaed/secret/coerced, etc.) access to Google's logs to see who was trying to peek at these places. Cross-referencing anyone who tried to look at super secret "nuclear installation Q-345" with other data might help the government find terrorist cells. Now, after the government complains, 100,000 normal folks immediately go to Google Earth to try to find these sensitive locations and pollute the access log files.
I see your point. I do know that Phase 1 clinical trials, for pharma products in the U.S., involve testing on healthy volunteers (usually on the order of 100 to 1000 people). These are people seen by doctors paid to do just the clinical trial (not other treatments) so the test subject (or their insurance company) is not the payor. So 100% of the Phase I costs are paid by the company with the new product because the doctors aren't doing it in the course of routine treatment.
Later phases involve patients that do need treatment. I suspect this creates some added cost due to added documentation that goes above and beyond the usual treatment regime and medical records keeping. But the bigger issue is that I doubt the patient's insurance company (in the U.S.) would reimburse for experimental treatments (I don't know how Medicare or other countries handle this). Thus the full cost of treatment probably gets paid by the company with the experimental drug. I'm sure cancer treatment isn't cheap (you may know this better than I), so covering the treatment costs on thousands of patients is quite costly.
Does your facility do clinical trials? I'd be interested in what the doctors say about who pays for what when a clinical trial is involved.
I'm sure that grad students can do so valuable preclinical work -- showing that exposure to this devices doesn't cause cancer in human cell cultures, showing that the device kills X% of type Y bacteria with Z seconds of exposure, etc. But that won't get the FDA's approval. Where do the grad students get the money to test the device on 1,000, 10,000, or 100,000 patients? Clinical testing is expensive. It requires real doctors on real salaries and if far beyond the scope of the labors of a grad student.
There's enough red tape as it is. Please don't make me go Federal for everything.
I agree 100%. Personally, I think the current system of patents on medical products is broken, but that a government-funded approach would be far worse for the reasons that you state.
But if the government does not pay for clinical testing, who does? And if some private company pays for it, how do they get paid? If large scale clinical testing is a prerequisite for ensuring the safety and efficiacy of medical products, then we either need a system of government funding (bad) or patent/government protection for those that spend the money on clinical testing (bad). I'll not suggest the option of dispensing with clinical testing and letting anyone sell anything in the medical arena.
The greater issue is that medical products aren't like TLDs or OSS. Medical devices have potentially lethal effects and are highly regulated. Gaining approval is much more costly than getting a few people to try your software or arguing a case before a standards body or OSS development team.
Who is going to spend that kind of money if the minute they get approval, some other company can sell these devices without the clinical testing costs? The company that performed the tests will need to add $25 to $200 to the price of the device (in addition to manufacturing costs), assuming they sell a million of them. And the competitor will be able to undercut the first company on price.
The math is even worse on a risk-adjusted basis because so many promising products fail during testing. Thus, the costs of developing several failed devices must be paid for by each successful device.
Until governments foot the bill for all medical R&D and clinical testing, patents are a crucial part of the medical device & pharma industry.
The point is that without a patent, nobody will pay for testing, the device will sit on a shelf, and it will do no one any good. This is why pharma and medical devices will never be like OSS -- the invention of the first instance is an extremely minor part of the cost of development. Building a better medical mousetrap is nothing. Proving it is safe and effective and gaining govt approval is everything.
And if you are lugging a carry-on and a laptop bag on the way to your flight, what stops the phone from deciding you are not you because the added weight changes your gait? TFA said the false alarm (accidental lockout) rate was 4%. I'd bet the rate is much worse if you are carrying something (suitcase, kid, groceries, etc.)
This may not be all bad. Perhaps some zombie PCs could be condemned under such an interpretation. Or, for example, if people create municipal mesh networks, then anyone joining the network may be required to provide public access to their part of the network for routing purposes (no leaching).
My point is that real estate/private property is heavily regulated and that entrenching the PC = private property analogy in case law could have some interesting legal consequences.
Does anyone know if these patents cover genes (a particular location on the DNA) or alleles (a specific variant that this found for some gene)? If the patent covers a specific DNA sequence, then it is an allele. If it covers an allele, then the number of possible patents is much larger than the number of genes.
Just think of the minutes burned as the "caller" explores some world/dungeon for hours.
That's only slightly tongue-in-cheek. (Yes, I know that between all the holes in the OS and all the holes in user's heads that screen-loggers will get installed with admin privileges.)
As much as I hate DRM ("lets assume 100% of computer users are illegal content distributors" and inconvenience everyone), it seems that it could be useful as part of locking down a machine from copying selected types of data to unauthorized external locations.
If letting a space tourist go up can attract media attention, then that's great. Its even nice that the customer pays the organization to create good marketing for the organization.
Besides, I'd bet the economics of space flight are such that the cost of filling an empty seat aren't that high. The average cost of putting a pound into orbit may be extremely high, but the cost of adding another pound of person and supplies is probably not bad. It's like the airlines -- if you're going to fly anyway, why not fill every seat.
Creating the idea that space is accessible to an increasing number of people -- not just a few astronauts that spent their life in the program -- is the key to the future of space funding.
It could be intelligence, knowledge on any of a number of dimensions, social grace, physical strength, affection, aggressiveness, niceness, humor, ambition, earning-power, etc.
Disclaimer: I've been married nearly 22 years so that means I either know what I'm talking about or have an insufficient sample size to comment on this.
If I could transfer my mod points from my post to yours, I would. Thanks for the info.
Even if I'm off by many orders of magnitude (IANAP), the required field strength will be unattainably high.
2. Thou shalt not let thy buffers overflow.
I hope those are in the Rational Unified Process (perhaps the construction phase of RUP).