> If I were a betting man, I'd bet that the web is still alive > and kicking 5 years from now - whether or not "new and better" Internet > technologies come along. The web is quite extendable > with browsers having the ability to run plug-ins and vbscript/javascript, > launch Active-X or Java applets, and so on.
I would second that - because of sheer inertia.
After giving her a twenty minute demo, I regularly get calls from my eighty year old aunt to look this or that up on the internet. So I am thinking about finding one of the early Imacs for her to get internet access herself.
Now, my main browser is NS 4.08, installed on Nov. 3rd, 1998 on my harddisk. Incidentally, this is almost exactly 5 years ago. If I turn javascript off, most pages do just fine. The rest will lose my business, or, if I really need the information, I can just telnet into somebody else's box to use a newer browser. However, especially when using google, a newer computer with a newer browser improves my searching experience by exactly 0%.
So, when I do not switch to a new browser for 5 years, why would you then expect my aunt to do so?
As the internet grows into maturity, expect the software used to access it to die with the computer it was bought with.
As computers grow into maturity, expect the hardware to last at least ten years.
So, I think, the internet year is going to have 365 earth days in the near future.
> If SCO has copyright material that has been infringed upon, > they have to go to the INFRINGER (whoever has access to their code and copied it, > meaning the code and not just a work-alike clean-room code, into the kernel) > for damages. > End users and unwitting publishers of infringing materials are not listed in USC-17 as > liable for infringement. > You can't get damages from a publisher if one author of a short story collection lied about the authorship, > nor can you collect from the bookstores and purchasers.
Isn't there a difference between selling and licensing software ?
If I buy a box of Red Hat X.Y at $49, then I bought this software from a distributor that is licensed (GPL) or acting in good faith (infringing software claimed to be GPL, but actually under somebody else's copyright, e.g. SCO's). In this case, according to the parents post, the end user may be in the clear.
However if I license an OS from IBM or SUN (i am not talking about the dubious MS shrink wrap "end user licensing" here), I only get the right to run this OS on box X for a year and get a certain level of maintenance. Next year, I have to pay again.
Theoretically, if the company licensing that OS finds out before next year that they do not have the right to parts of the OS, they could say "sorry, this year you can only license an OS without a filesystem from us".
In this case, wouldn't the user of the software be in a much better position legally with an open source OS than with the most expensive proprietary OS he is running under a licensing scheme and that comes with the most far reaching promises of "indemnification" ?
I wonder if it would help to execute all the stuff required for booting out of a Ram-disk ?
Since disks these days are pretty fast (>20 MByte sustained for linear reading) you should be able to load say 100Mbyte (uncompressed) into a ramdisk in about 5 secs and then execute from there.
This way, you do not have change anything about the init scripts, and the init_rd stuff is already in the kernel. Of course, in your 100 Mbyte, there will be a lot of pages that are never touched, but I think it would be still a winner as reading the full track is almost free as compared to reading just a few blocks and then wait for more than half a revolution on average for the next block on another track.
But rumours of its death are greatly exaggerated!
on
Ford To Move To Linux
·
· Score: 1
> Most new, non-game applications these days are written for the Web.
> This way, any platform can connect to the application and run it the same way as anyone else would > on a Palm Pilot or Pocket PC or Linux or Solaris or Windows or whatever.
Maybe, but that hardly affects the decision what OS to run today because of the long lifecycles for business apps.
VW, for example, is also planning to migrate to linux for the servers in their dealerships.
In the smaller ones, they will replace 286 systems with 1 Megabyte of Ram (Nixdorf Quattro 8870/25) running a proprietary Nixdorf OS with applications programmed in "Business Basic". I know one place where the server was installed in 1987 - initially with 8" floppies and two 20 Meg harddisks.
They have now upgraded to 5 1/4" floppies and a single 130 Meg harddisk.
Unlike MS-DOS 5.0, the OS is still supported !
The server is running the accounting and the programs for spare parts. You can either use Nixdorf Terminals or Terminal Emulators on PCs for the clients.
They now have an emulator ready (I am not sure if for the Basic-Interpreter or the whole NIROS-OS) so they are ready to move to linux !
I would bet that the by the time the last non-Web Application is rewritten, you will be commuting to work in a flying car;-)
> Anyone else like how the old BeBoxes had LED bars showing the CPU usage?
Remember your granddad telling you about how in the old days, people would have two-digit 7-segment indicators on their computers to tell the speed of the cpu they would have liked to have in their box ?
Combine that with your CPU-usage speed-bar into the spoil-o-meter (TM) !
When you buy your latest 357 BogoHZ Nonium IIXXX Cpu at $998, put the next smaller CPUs you considered buying into a list, along with their price and computing power relative to CPU you eventually bought.
That would be the 337 BogoHZ Nonium IIXXX at $698, the 250 BogoHZ Celeronium XXV at $123, and the surplus 66MHz 486DX you found in your scrap heap at $0.
Then write a utility that displays the difference in price between your current CPU, and the cheapest one in the list whose computing power would just exceed your current cpu load, on a three-digit 7-segment display.
If you begin to feel depressed, install "SETI AT HOME" - or Windows XP.
Since the area of the membranes used in fuel cells scales with peak power, they do not get significantly cheaper per watt in bigger units - at least not by two orders of magnitude.
> In some cities like San Francisco, it's cheaper > to pay somebody to keep driving your car around > in circles than it is to pay for a parking > garage. Why pay thousands for an automated > system when you could pay a teenager $6/hour?
Maybe because new cars only start at $10,000/apiece ?
> To be fair, I said "this is what we need." Not, > "this is what's going to happen."
This seems too pessimistic to me.
Look at what premiums laptop users pay for a Pentium M (Banias / Centrino) over a plain P4 or Celeron.
These chips have roughly the same computing power, but the Pentium M uses much less than half the power at more than twice the price.
Of course, you can attribute that mostly to clever markeing, but I think it really did appeal to great customer demand.
And I do not want to know how many million PCs get replaced every year, inspite of the fact that they are still perfectly adequate for their tasks - just because the fans started getting noisy !
I think that as soon as the innovation speed in LCD displays cools down a bit (say, in 2..3 years), another form factor besides "desktop", "tower" and "notebook" will be the "clip-on", a flat 12 x 8 inch enclosure holding a motherboard, a powersupply and one or two drives.
The clip-on will be about one inch thick and clip to the back of your LCD display, using standard connectors for mains (in from LCD), video and USB-like busses (memory-sticks, digicams, etc) that will have passive connectors in the LCDs front bezel or stand.
This will allow you to upgrade PC and Display seperately, as you do today with desktop-PC and crt - but without the annoying cabling, and at a smaller footprint.
I think there will be a huge market for this (especially corporate), and it will take off, if intel gets the mechanical specs standardised like the ATX-form-factor.
> Electronics are becoming commodities as they become efficient and > cost-effective a few basic tasks that people find entertaining and useful.
I think that is correct.
I am pretty sure that in 2007, the decision to buy a new computer will look like this:
-------------
Imagine a small office, say at a dentist's. There are three chairs, three computers, a couple of potted plants and a coffeemaker.
NURSE: Doctor, i think my computer is dead. There is smoke coming out of it.
DOCTOR: Oh shit. I just send Jane to Staples to bring the printer paper we ran out of. Quickly call her on the mobile so she will pick up a new one for you. In the meantime, use mine.
NURSE (on the phone): Hello Jane, are you still at Staples? My Computer broke down, so please bring me a new one. Try to get one with the integrated scotch-tape and post-it dispenser on the front, you know I like my desk real neat and tidy.
JANE (on the phone): Now i am in the computer department. Well, it looks as if the original scotch-brand computers cost at least 5 dollars more than a no-name. Do you insist on a "scotch" ?
NURSE: No, but at least make sure the color will go nicely with the new coffemaker.
-------------
This is how you will buy a desktop calculator today, and a desktop computer tomorrow.
I remember my mother, who is the mom part in my parents mom-and-pop business, having her electro-mechanical desktop calculator repaired every two of years between 1970 and 1990, for a price you could eventually buy three new electronic ones for.
Since then, makers of desktop calculators have rarely made it to the front page of the Wall-Street-Journal.
What I wonder about is why they use a sphere and not a cube or a tetrahedron.
If they have a perfect crystal, then it might be possible to use etching to align the sides of the cube or tetrahedron(with silicon having four bonds ?) to the lattice of the crystal and then count the layers either directly (e.g with these raster tunnel microscopes) or by using interference effects that would group a fixed number of layers.
Is a silicon crystal pure enough so that the basic orientation of the crystal lattice is undisturbed over distances of 10cm ?
I think the number of atoms along the sides of a 1kg cube of silicon (app. 8cm) would be low enough ( 10^9) so you could hope you could actually count the atoms in reasonable time (eg. within 20minutes ~= 1000secs at 1MHz counting frequency)
Of course, a sphere would be easier to handle and less prone to chipping bits off.
Another idea about the silicon cube would be to apply "wear and smear detection circuitry" on the six sides using standard silicon semiconductor processing, e.g by covering the surface with a mesh of little custom cells that would detect the presence of their neighbors and report failed cells when they do not respond.
But I think this would introduce too much error, as 50ug error on the old 1Kg iridium slab (1:2*10^7) would translate to a dimensional error of just about 1nm, less than 10 atomic layers - and these guys are aiming for a noticeable improvement on this !
> NASA has a program in development. The prototype has already flown. > It's called the AERCam Sprint.
Its a little bit bigger than I would have envisioned it, but then it was 1997 when it flew, probably a lot earlier when design began, and it looks cheap (by space standards, that is).
However, I do not usually consider me particularly paranoid;)
I think I still remember images of the shuttle's belly shot from a high-altitude reconnaissance jet in one of the first flights to check for missing tiles.
So I think, as a diagnostic, something like this thing should have been really nice to have.
But if you have ever flown in a commercial airliner, you might have noticed the conspicuos absence of any "crash due in xx seconds, pray now!" signs, as compared to "stop smoking" or "fasten seat belts"
Paranoids will suspect the reason is that, although prototypes of such systems were tested successfully by Boeing in 1997, marketing VPs at the airlines vetoed the introduction because it might disturb some of the feebler minds among customers and make them defect to AMTRAK.
But then, the agnostic in me just suggests that, statistically, the effect of stopping smoking for 30 minutes during every flight has a more beneficial impact on the average lifetime of passengers than praying in the rare case of an imminent crash, so I think airlines are acting in the best interest of their customers (in this case, anyway).
For NASA, it probably has been a fair decision in the best interest of their crews not to put them under mental stress (and probably even induce serious human error this way) by telling them "you might burn up on reentry" if
a) damage could be assessed after landing
and
b) they knew there was nothing they could do
during flight that would be significantly more
effective than praying
Still, I am pretty sure that such a system will be used on future flights.
- three 1.8inch disks used as flywheels for stabilisation - C02 cartridge from soda dispenser as gas-tank - six exhaust nozzles coupled to the CO2-cartridge
with magnetic valves - tilt-and swivel camera like the one guarding the ATM at your bank - hf-video link - six-channel RC-control - NiCd-battery pack
Put this assembly into a clear plastic bag and leak some of the CO2 into it to get a few mbar of pressure to avoid stiction (of course, the exhaust nozzles would stick out). You might also want to replace some cheap capacitors.
At radio shack, this would set you back less than $1000 and could be assembled by any radio ham.
For a few million dollars, you could get a version of this assembled without scotchtape, weighing 1..2 kg.
The only reason I can imagine that the shuttle, threatened by tiles falling off since day one, does not carry such an inspection pod on each mission, is that NASA did not want the crew or the public to know there is a problem either on a mission doomed to failure or in advance on the missions that came back lucky.
Re:physicists studied flying - mechanics flew !
on
AI Going Nowhere?
·
· Score: 1
I heard a bit about the Wright brothers on the
radio the other day. A historian was giving an account of how it happened, and emphasized the fact that they were *not* tinkering; they were in fact taking a highly scientific approach -- developing a theory, or a refinement to a theory, designing a test setup, and then implementing and executing it.
Maybe tinkering was not the correct term - I did not want to denounce the effort of the Wright brothers by any means!
What I was trying to point out was that both the Wright brothers and Lilienthal were not members of the established scientific community at that time.
Both received none or little public funding, and were outsiders that each used their bicycle factory to fund their systematic engineering approach that made them test prototype after prototype for years on end.
For example, Lilienthal had a large hill put up near his home town Berlin, where the landscape was essentialy flat, to test his glider design. He first studied birds wings and the seeds of some plants to understand how they kept afloat for a long time. He then build and tested and improved on his gliders for something like 10 years.
The Wright brothers spent a few years after their first flight in 1903 perfecting their design before they showed it off to larger audiences.
Both did not disregard scientific method, but put at least an equal amount of effort into hands-on prototype building and, when in doubt, rather relied on their experience with this than on established academic knowledge that later turned out wrong or irrelevant.
The history of technological development is full of examples where phenomena that went against established scientific wisdom were utilised long before a sufficient theory could be worked up, such as intercontinental radio communication despite the fact that line of sight would be required by the contemporary theory.
It is quite possible that AI may turn out to be a field where this hands-on, evolutionary approach does not work and a breakthrough in theory is required first for the development of AI, but I would rather put my bets on something like the race between Microsoft and Sony for the 100 billion $ pet market after animal pets get outlawed by the FDA following the reappearance of canine SARS in 2011;)
physicists studied flying - mechanics flew !
on
AI Going Nowhere?
·
· Score: 1
Did it ever occur to you that some of the main contributors to the fledling field of flying were bicycle manufacturers, such as the Wright brothers, or Otto Lilienthal, who constructed, tested and improved on gliders for years at the end of the 19th century and eventuallied died in a crash testing his last model.
In these days, bicycle manufacturers usually were not MBAs or engineers, but mechanics.
Contrary to that, physicists and philosophers had theoreticised about flying for thousends of years, but none of them actually did fly. But then again, none of them crashed either !
How can you be sure that the Wright brothers of AI will not be two tinkerers trying to mimic their aunt's dog, or building a robotic football team ?
As the discussion in both this thread and in AI shows, there is not yet a common understanding of what creating an artificial intelligence requires.
How can people then be sure that playing with robotic toys will never provide valuable insights into this field?
price fixing in dram is not a major problem
on
DRAM Price Fixing
·
· Score: 1
I think price-fixing in dram is not a major problem compared to other areas like CPUs, because there is a strong counterforce: extreme elasticity of demand.
Most people would find their computing experience degrade quite heftily by leaving this CPU out of their box, while using just 8 256 Mbit chips instead of 16 will often be hardly noticeable.
Remember when dram prices went through the roof some 4 years ago ? I went to my computer dealer, told I would need 256Mbyte ram with the new box, heard what he had to charge for it, and went home to take an 128Mbyte Dimm out of another box.
Remember all the machines selling with 1GHz Cpus and 64Mbyte of ram at the discounters during that time, that would swap like a VAX750 with a hundred users?
Remember all the utility programs that would reclaim unused Ram portions of some processes in the Mac when it didn't use a mpu ?
Therefore, I think that currently, price fixing is not a problem. If it raises prices in the short run, it may even be beneficial to the consumer, because it may keep some of the smaller vendors in the market and allow the industry to speed development of yet bigger and cheaper memories.
However, if it is used to kill off the competition, then we may come to a situation like that in the CPU market before AMD came back as a competitive player with the Athlon.
I think that historically, the Ram in a computer has always cost the same or even more than the CPU, and the disk-drive was much more expensive still. In the last years, this trend has changed, and I would think because the disk-drive and the DRAM-Market are much more competitive and commoditised as compared to the CPU market.
However, I think in the last paragraph, there is a little error, you have probably forgotten to multiply by 360.
An 18m diameter object at a roughly 18.000m distance is equivalent in apparent diameter to an object of 1m diameter at 1000m or 100mm at 100m.
This is the equivalent of a fist over the length of a soccer field, which should be visible.
Another comparison is the thickness of a pencil (mine measures app. 7mm = 0.3inch) across a room (7m = 20feet).
But I imagine this Balloon would be hard to find in the day, because it would probably be highly reflective.
Since most of the athmosphere (app. 9/10th?) would be below it, most of the light would be scattered by the part of the atmosphere below it, so I would think that it would have just a few percent of contrast aginst the background and would therefore be virtually invisible to the naked eye.
On a normal car *I* have to drive in, I would not trust tires that last 70.000 miles, because under adverse conditions (cold, wet) the rubber would be too hard to get good grip on the road.
If you want to reduce the cost of driving, buy a smaller car or drive more carefully to save on gas and tire wear, but do not risk your and your families health by compromising the single most important active safety component of your car.
Well, if you live in Europe, I think you could make the biggest impact in this way:
- order a machine from direct sales
- when it arrives, say you do not accept the eula for windows and demand a refund
- If they refuse you the refund, you can ship the whole machine back at no charge (free shipping) within I think 10 days without even giving a reason. This is your right according to an EU-directive (in german I think it is the "Fernabsatzgesetz").
They may have a right to deny you a refund for windows alone, but you have right to demand a full refund for the full machine.
Its as easy as that, all it'll cost you is a phone call.
IANACM - but this looks highly questionable to me
on
A Rock Moves In Space
·
· Score: 1
Its acronym time again !
IANACM (I Am Not A Celestial Mechanic), but this:
> 1.Feb 1, 2019 (distance: ~28500 km) > 2.Feb 1, 2044 (distance: ~91100 km) > 3.Feb 1, 2053 (distance: ~53500 km) > 4.Feb 1, 2060 (distance: ~3570 km) > 5.Feb 1, 2078 (distance: ~15900 km) > 6.with interesting passes every 7 to 14 years after that > (your asteroid's mileage may vary:-))
looks highly questionable to me.
This suggests that the asteroid, which according to the bbc article has an orbit that does not lie in the ecliptic plane, comes within less than 8 earth diameters on 5 occasions in less than ten 7-year periods.
This would require two things:
a) The period of the asteroid and the earth must be synchronized to a ratio of 3/7 to within less than app. 1 hour in 60 years - an accuracy of approximately 1:500000
b) All those near flybys must not significantly alter the course of the asteroid. (comparison: geostationary satellite orbit is app. 35000km, and the satellite is deflected by 360 degrees in one day)
I would expect the first flyby in 2019, according to the poster well within geosynchronous orbit, to change the course of the asteroid by something like 0.1 degree (the asteroid is app 10 times faster relative to the earth as the satellite).
After that, the course has changed randomly and the speed has changed significantly (the swingby effect), so any synchronization that may have been there is lost. If we then have 4 more close flybys, the chances for that are more like 5 lottery wins in a row.
Maybe there was just a decimal error in those calculations.
And when these teenagers grow up, they join the army and get killed in the middle east while trying to the secure the supply of gas for their parents SUV.
Time to think again about your energy consumption ?
Or are you going to solve the problem by only sending kids from black ghettoes ?
>I am not a physicist, but I think that the speed of light (app. 300m/s) > should give an ballpark estimate on the > average speed of a molecule at room temperature.
Sorry, off by a trifling matter of six magnitudes.
Slashdot Mirror
> If I were a betting man, I'd bet that the web is still alive
> and kicking 5 years from now - whether or not "new and better" Internet
> technologies come along. The web is quite extendable
> with browsers having the ability to run plug-ins and vbscript/javascript,
> launch Active-X or Java applets, and so on.
I would second that - because of sheer inertia.
After giving her a twenty minute demo, I regularly get calls from my eighty year old aunt to look this or that up on the internet. So I am thinking about finding one of the early Imacs for her to get internet access herself.
Now, my main browser is NS 4.08, installed on Nov. 3rd, 1998 on my harddisk. Incidentally, this is almost exactly 5 years ago. If I turn javascript off, most pages do just fine. The rest will lose my business, or, if I really need the information, I can just telnet into somebody else's box to use a newer browser. However, especially when using google, a newer computer with a newer browser improves my searching experience by exactly 0%.
So, when I do not switch to a new browser for 5 years, why would you then expect my aunt to do so?
As the internet grows into maturity, expect the software used to access it to die with the computer it was bought with.
As computers grow into maturity, expect the hardware to last at least ten years.
So, I think, the internet year is going to have 365 earth days in the near future.
> If SCO has copyright material that has been infringed upon,
> they have to go to the INFRINGER (whoever has access to their code and copied it,
> meaning the code and not just a work-alike clean-room code, into the kernel)
> for damages.
> End users and unwitting publishers of infringing materials are not listed in USC-17 as
> liable for infringement.
> You can't get damages from a publisher if one author of a short story collection lied about the authorship,
> nor can you collect from the bookstores and purchasers.
Isn't there a difference between selling and licensing software ?
If I buy a box of Red Hat X.Y at $49, then I bought this software from a distributor that is licensed (GPL) or acting in good faith (infringing software claimed to be GPL, but actually under somebody else's copyright, e.g. SCO's). In this case, according to the parents post, the end user may be in the clear.
However if I license an OS from IBM or SUN (i am not talking about the dubious MS shrink wrap "end user licensing" here), I only get the right to run this OS on box X for a year and get a certain level of maintenance. Next year, I have to pay again.
Theoretically, if the company licensing that OS finds out before next year that they do not have the right to parts of the OS, they could say "sorry, this year you can only license an OS without a filesystem from us".
In this case, wouldn't the user of the software be in a much better position legally with an open source OS than with the most expensive proprietary OS he is running under a licensing scheme and that comes with the most far reaching promises of "indemnification" ?
I wonder if it would help to execute all the stuff required for booting out of a Ram-disk ?
Since disks these days are pretty fast (>20 MByte sustained for linear reading) you should be able to load say 100Mbyte (uncompressed) into a ramdisk in about 5 secs and then execute from there.
This way, you do not have change anything about the init scripts, and the init_rd stuff is already in the kernel. Of course, in your 100 Mbyte, there will be a lot of pages that are never touched, but I think it would be still a winner as reading the full track is almost free as compared to reading just a few blocks and then wait for more than half a revolution on average for the next block on another track.
> Most new, non-game applications these days are written for the Web.
;-)
> This way, any platform can connect to the application and run it the same way as anyone else would
> on a Palm Pilot or Pocket PC or Linux or Solaris or Windows or whatever.
Maybe, but that hardly affects the decision what OS to run today because of the long lifecycles for business apps.
VW, for example, is also planning to migrate to linux for the servers in their dealerships.
In the smaller ones, they will replace 286 systems with 1 Megabyte of Ram (Nixdorf Quattro 8870/25) running a proprietary Nixdorf OS with applications programmed in "Business Basic". I know one place where the server was installed in 1987 - initially with 8" floppies and two 20 Meg harddisks.
They have now upgraded to 5 1/4" floppies and a single 130 Meg harddisk.
Unlike MS-DOS 5.0, the OS is still supported !
The server is running the accounting and the programs for spare parts. You can either use Nixdorf Terminals or Terminal Emulators on PCs for the clients.
They now have an emulator ready (I am not sure if for the Basic-Interpreter or the whole NIROS-OS) so they are ready to move to linux !
I would bet that the by the time the last non-Web Application is rewritten, you will be commuting to work in a flying car
> Anyone else like how the old BeBoxes had LED bars showing the CPU usage?
Remember your granddad telling you about how in the old days, people would have two-digit 7-segment indicators on their computers to tell the speed of the cpu they would have liked to have in their box ?
Combine that with your CPU-usage speed-bar into the spoil-o-meter (TM) !
When you buy your latest 357 BogoHZ Nonium IIXXX Cpu at $998, put the next smaller CPUs you considered buying into a list, along with their price and computing power relative to CPU you eventually bought.
That would be the 337 BogoHZ Nonium IIXXX at $698, the 250 BogoHZ Celeronium XXV at $123, and the surplus 66MHz 486DX you found in your scrap heap at $0.
Then write a utility that displays the difference in price between your current CPU, and the cheapest one in the list whose computing power would just exceed your current cpu load, on a three-digit 7-segment display.
If you begin to feel depressed, install "SETI AT HOME" - or Windows XP.
> It's one of my favorite words.
Then you should check out "The defenestration of Ermintrude Inch", a short story by A.C. Clarke.
A classic on social interaction for geeks.
It is included in his book "Tales from the White Hart".
Have fun !
The reason why fuel cells appear in laptops first and only later in cars is the price/power ratio:
Laptop : $2,000 / 100W = $20.00/W
Car : $20,000 / 100KW = $ 0.20/W
Since the area of the membranes used in fuel cells scales with peak power, they do not get significantly cheaper per watt in bigger units - at least not by two orders of magnitude.
> In some cities like San Francisco, it's cheaper
> to pay somebody to keep driving your car around
> in circles than it is to pay for a parking
> garage. Why pay thousands for an automated
> system when you could pay a teenager $6/hour?
Maybe because new cars only start at $10,000/apiece ?
> To be fair, I said "this is what we need." Not, > "this is what's going to happen."
This seems too pessimistic to me.
Look at what premiums laptop users pay for a Pentium M (Banias / Centrino) over a plain P4 or Celeron.
These chips have roughly the same computing power, but the Pentium M uses much less than half the power at more than twice the price.
Of course, you can attribute that mostly to clever markeing, but I think it really did appeal to great customer demand.
And I do not want to know how many million PCs get replaced every year, inspite of the fact that they are still perfectly adequate for their tasks - just because the fans started getting noisy !
I think that as soon as the innovation speed in LCD displays cools down a bit (say, in 2..3 years), another form factor besides "desktop", "tower" and "notebook" will be the "clip-on", a flat 12 x 8 inch enclosure holding a motherboard, a powersupply and one or two drives.
The clip-on will be about one inch thick and clip to the back of your LCD display, using standard connectors for mains (in from LCD), video and USB-like busses (memory-sticks, digicams, etc) that will have passive connectors in the LCDs front bezel or stand.
This will allow you to upgrade PC and Display seperately, as you do today with desktop-PC and crt - but without the annoying cabling, and at a smaller footprint.
I think there will be a huge market for this (especially corporate), and it will take off, if intel gets the mechanical specs standardised like the ATX-form-factor.
> Electronics are becoming commodities as they become efficient and
> cost-effective a few basic tasks that people find entertaining and useful.
I think that is correct.
I am pretty sure that in 2007, the decision to buy a new computer will look like this:
-------------
Imagine a small office, say at a dentist's. There are three chairs, three computers, a couple of potted plants and a coffeemaker.
NURSE: Doctor, i think my computer is dead. There is smoke coming out of it.
DOCTOR: Oh shit. I just send Jane to Staples to bring the printer paper we ran out of. Quickly call her on the mobile so she will pick up a new one for you. In the meantime, use mine.
NURSE (on the phone): Hello Jane, are you still at Staples? My Computer broke down, so please bring me a new one. Try to get one with the integrated scotch-tape and post-it dispenser on the front, you know I like my desk real neat and tidy.
JANE (on the phone): Now i am in the computer department. Well, it looks as if the original scotch-brand computers cost at least 5 dollars more than a no-name. Do you insist on a "scotch" ?
NURSE: No, but at least make sure the color will go nicely with the new coffemaker.
-------------
This is how you will buy a desktop calculator today, and a desktop computer tomorrow.
I remember my mother, who is the mom part in my parents mom-and-pop business, having her electro-mechanical desktop calculator repaired every two of years between 1970 and 1990, for a price you could eventually buy three new electronic ones for.
Since then, makers of desktop calculators have rarely made it to the front page of the Wall-Street-Journal.
What I wonder about is why they use a sphere and not a cube or a tetrahedron.
If they have a perfect crystal, then it might be possible to use etching to align the sides of the cube or tetrahedron(with silicon having four bonds ?) to the lattice of the crystal and then count the layers either directly (e.g with these raster tunnel microscopes) or by using interference effects that would group a fixed number of layers.
Is a silicon crystal pure enough so that the basic orientation of the crystal lattice is undisturbed over distances of 10cm ?
I think the number of atoms along the sides of a 1kg cube of silicon (app. 8cm) would be low enough ( 10^9) so you could hope you could actually count the atoms in reasonable time (eg. within 20minutes ~= 1000secs at 1MHz counting frequency)
Of course, a sphere would be easier to handle and less prone to chipping bits off.
Another idea about the silicon cube would be to apply "wear and smear detection circuitry" on the six sides using standard silicon semiconductor processing, e.g by covering the surface with a mesh of little custom cells that would detect the presence of their neighbors and report failed cells when they do not respond.
But I think this would introduce too much error,
as 50ug error on the old 1Kg iridium slab (1:2*10^7) would translate to a dimensional error of just about 1nm, less than 10 atomic layers - and these guys are aiming for a noticeable improvement on this !
> Your paranoia is amusing.
;)
> NASA has a program in development. The prototype has already flown.
> It's called the AERCam Sprint.
Its a little bit bigger than I would have envisioned it, but then it was 1997 when it flew, probably a lot earlier when design began, and it looks cheap (by space standards, that is).
However, I do not usually consider me particularly paranoid
I think I still remember images of the shuttle's belly shot from a high-altitude reconnaissance jet in one of the first flights to check for missing tiles.
So I think, as a diagnostic, something like this thing should have been really nice to have.
But if you have ever flown in a commercial airliner, you might have noticed the conspicuos absence of any "crash due in xx seconds, pray now!" signs, as compared to "stop smoking" or "fasten seat belts"
Paranoids will suspect the reason is that, although prototypes of such systems were tested successfully by Boeing in 1997, marketing VPs at the airlines vetoed the introduction because it might disturb some of the feebler minds among customers and make them defect to AMTRAK.
But then, the agnostic in me just suggests that, statistically, the effect of stopping smoking for 30 minutes during every flight has a more beneficial impact on the average lifetime of passengers than praying in the rare case of an imminent crash, so I think airlines are acting in the best interest of their customers (in this case, anyway).
For NASA, it probably has been a fair decision in the best interest of their crews not to put them under mental stress (and probably even induce serious human error this way) by telling them "you might burn up on reentry" if
a) damage could be assessed after landing
and
b) they knew there was nothing they could do
during flight that would be significantly more
effective than praying
Still, I am pretty sure that such a system will be used on future flights.
- three 1.8inch disks used as flywheels for stabilisation
- C02 cartridge from soda dispenser as gas-tank
- six exhaust nozzles coupled to the CO2-cartridge
with magnetic valves
- tilt-and swivel camera like the one guarding the ATM at your bank
- hf-video link
- six-channel RC-control
- NiCd-battery pack
Put this assembly into a clear plastic bag and leak some of the CO2 into it to get a few mbar of pressure to avoid stiction (of course, the exhaust nozzles would stick out). You might also want to replace some cheap capacitors.
At radio shack, this would set you back less than $1000 and could be assembled by any radio ham.
For a few million dollars, you could get a version of this assembled without scotchtape, weighing 1..2 kg.
The only reason I can imagine that the shuttle, threatened by tiles falling off since day one, does not carry such an inspection pod on each mission, is that NASA did not want the crew or the public to know there is a problem either on a mission doomed to failure or in advance on the missions that came back lucky.
I heard a bit about the Wright brothers on the radio the other day. A historian was giving an account of how it happened, and emphasized the fact that they were *not* tinkering; they were in fact taking a highly scientific approach -- developing a theory, or a refinement to a theory, designing a test setup, and then implementing and executing it.
;)
Maybe tinkering was not the correct term - I did not want to denounce the effort of the Wright brothers by any means!
What I was trying to point out was that both the Wright brothers and Lilienthal were not members of the established scientific community at that time. Both received none or little public funding, and were outsiders that each used their bicycle factory to fund their systematic engineering approach that made them test prototype after prototype for years on end.
For example, Lilienthal had a large hill put up near his home town Berlin, where the landscape was essentialy flat, to test his glider design. He first studied birds wings and the seeds of some plants to understand how they kept afloat for a long time. He then build and tested and improved on his gliders for something like 10 years. The Wright brothers spent a few years after their first flight in 1903 perfecting their design before they showed it off to larger audiences. Both did not disregard scientific method, but put at least an equal amount of effort into hands-on prototype building and, when in doubt, rather relied on their experience with this than on established academic knowledge that later turned out wrong or irrelevant. The history of technological development is full of examples where phenomena that went against established scientific wisdom were utilised long before a sufficient theory could be worked up, such as intercontinental radio communication despite the fact that line of sight would be required by the contemporary theory.
It is quite possible that AI may turn out to be a field where this hands-on, evolutionary approach does not work and a breakthrough in theory is required first for the development of AI, but I would rather put my bets on something like the race between Microsoft and Sony for the 100 billion $ pet market after animal pets get outlawed by the FDA following the reappearance of canine SARS in 2011
Did it ever occur to you that some of the main contributors to the fledling field of flying were bicycle manufacturers, such as the Wright brothers, or Otto Lilienthal, who constructed, tested and improved on gliders for years at the end of the 19th century and eventuallied died in a crash testing his last model.
In these days, bicycle manufacturers usually were not MBAs or engineers, but mechanics.
Contrary to that, physicists and philosophers had theoreticised about flying for thousends of years, but none of them actually did fly. But then again, none of them crashed either !
How can you be sure that the Wright brothers of AI will not be two tinkerers trying to mimic their aunt's dog, or building a robotic football team ?
As the discussion in both this thread and in AI shows, there is not yet a common understanding of what creating an artificial intelligence requires.
How can people then be sure that playing with robotic toys will never provide valuable insights into this field?
I think price-fixing in dram is not a major problem compared to other areas like CPUs, because there is a strong counterforce: extreme elasticity of demand.
Most people would find their computing experience degrade quite heftily by leaving this CPU out of their box, while using just 8 256 Mbit chips instead of 16 will often be hardly noticeable.
Remember when dram prices went through the roof some 4 years ago ? I went to my computer dealer, told I would need 256Mbyte ram with the new box, heard what he had to charge for it, and went home to take an 128Mbyte Dimm out of another box.
Remember all the machines selling with 1GHz Cpus and 64Mbyte of ram at the discounters during that time, that would swap like a VAX750 with a hundred users?
Remember all the utility programs that would reclaim unused Ram portions of some processes in the Mac when it didn't use a mpu ?
Therefore, I think that currently, price fixing is not a problem. If it raises prices in the short run, it may even be beneficial to the consumer, because it may keep some of the smaller vendors in the market and allow the industry to speed development of yet bigger and cheaper memories.
However, if it is used to kill off the competition, then we may come to a situation like that in the CPU market before AMD came back as a competitive player with the Athlon.
I think that historically, the Ram in a computer has always cost the same or even more than the CPU, and the disk-drive was much more expensive still. In the last years, this trend has changed, and I would think because the disk-drive and the DRAM-Market are much more competitive and commoditised as compared to the CPU market.
... if you need 300,000 green suits to sell it ;)
If you take a look at the highway into town at 8 o'clock in the morning, 90% of the cars hold just one person.
Also, 90% of the cars have four or five seats.
Why restrict yourself to two seats if four can be had at a marginal extra expense, if any ?
Maybe some day you want to take a few extra passengers, even though the need never arouse during the last couple of years of commuting.
This is why consumers will buy 64-bit CPU's as soon as they are affordable.
Thanks for actually figuring this out.
However, I think in the last paragraph, there is a little error, you have probably forgotten to multiply by 360.
An 18m diameter object at a roughly 18.000m distance is equivalent in apparent diameter to an object of 1m diameter at 1000m or 100mm at 100m.
This is the equivalent of a fist over the length of a soccer field, which should be visible.
Another comparison is the thickness of a pencil (mine measures app. 7mm = 0.3inch) across a room (7m = 20feet).
But I imagine this Balloon would be hard to find in the day, because it would probably be highly reflective.
Since most of the athmosphere (app. 9/10th?) would be below it, most of the light would be scattered by the part of the atmosphere below it, so I would think that it would have just a few percent of contrast aginst the background and would therefore be virtually invisible to the naked eye.
... happen to come with your Ford Explorer ?
On a normal car *I* have to drive in, I would not trust tires that last 70.000 miles, because under adverse conditions (cold, wet) the rubber would be too hard to get good grip on the road.
If you want to reduce the cost of driving, buy a smaller car or drive more carefully to save on gas and tire wear, but do not risk your and your families health by compromising the single most important active safety component of your car.
... I would have expected tail fins measuring a little bit more than a measly 10nm !
Well, if you live in Europe, I think you could make the biggest impact in this way :
- order a machine from direct sales
- when it arrives, say you do not accept the eula for windows and demand a refund
- If they refuse you the refund, you can ship the whole machine back at no charge (free shipping) within I think 10 days without even giving a reason. This is your right according to an EU-directive (in german I think it is the "Fernabsatzgesetz").
They may have a right to deny you a refund for windows alone, but you have right to demand a full refund for the full machine.
Its as easy as that, all it'll cost you is a phone call.
Its acronym time again !
:-))
IANACM (I Am Not A Celestial Mechanic),
but this:
> 1.Feb 1, 2019 (distance: ~28500 km)
> 2.Feb 1, 2044 (distance: ~91100 km)
> 3.Feb 1, 2053 (distance: ~53500 km)
> 4.Feb 1, 2060 (distance: ~3570 km)
> 5.Feb 1, 2078 (distance: ~15900 km)
> 6.with interesting passes every 7 to 14 years after that
> (your asteroid's mileage may vary
looks highly questionable to me.
This suggests that the asteroid, which according to the bbc article has an orbit that does not lie in the ecliptic plane, comes within less than 8 earth diameters on 5 occasions in less than ten 7-year periods.
This would require two things:
a) The period of the asteroid and the earth must be synchronized to a ratio of 3/7 to within less than app. 1 hour in 60 years - an accuracy of approximately 1:500000
b) All those near flybys must not significantly alter the course of the asteroid. (comparison: geostationary satellite orbit is app. 35000km, and the satellite is deflected by 360 degrees in one day)
I would expect the first flyby in 2019, according to the poster well within geosynchronous orbit, to change the course of the asteroid by something like 0.1 degree (the asteroid is app 10 times faster relative to the earth as the satellite).
After that, the course has changed randomly and the speed has changed significantly (the swingby effect), so any synchronization that may have been there is lost. If we then have 4 more close flybys, the chances for that are more like 5 lottery wins in a row.
Maybe there was just a decimal error in those calculations.
And when these teenagers grow up, they join the army and get killed in the middle east while trying to the secure the supply of gas for their parents SUV.
Time to think again about your energy consumption ?
Or are you going to solve the problem by only sending kids from black ghettoes ?
>I am not a physicist, but I think that the speed of light (app. 300m/s)
;-)
> should give an ballpark estimate on the
> average speed of a molecule at room temperature.
Sorry, off by a trifling matter of six magnitudes.
That was to be the speed of sound, not light