Well, while Tivo will most likely not muck with the ratio, they might be doing on-the-fly recompression and resampling at the same time to store the HD programming in a smaller amount of disk space and possibly using a different algorithm to encode... perhaps Mpeg4?
It makes sense if you plan on using essentially the same setup as other machines on the production line. What you would change would be the encoder/decoders, possibly upgrade the main CPU on the board, and up the drive to a 120GB single hard drive in the unit.
If they reduced the bitrate of the video stream when they save to disk, they would be able to allow for different levels of "quality" and thus be able to maximize the number of hours to store on the unit.
Definitely realistic, but I wonder whether or not Tivo will be decompressing on-the-fly and re-compressing and re-sampling to a format which will be slightly degraded, but take up less space?
For example, the current Tivos allow different grades of quality for storage, each one taking less space or more space. I wonder whether or not Tivo will be decompressing the HD stream/signal and then recompressing it at different rates/qualities.
Another possibility is that Tivo might start making use of Mpeg4 instead of Mpeg2 for their internal encoding. This would allow them to store a stream with HD dimensions without taking up the HD space.
Side Note: Wonder if they will be putting out software upgrades to those of us who own Tivo Series2's and older Tivos to make use of HD at somewhat degraded resolution? (Current Tivo2: 199HourCapacity)
So does this mean that my SoundCard will need Palladium as well?
Reasoning being that quite a few, if not all, expansion cards have their own BIOS which gets executed at startup to "mingle" on the pci/isa/etc bus.
And... what's to prevent someone from writing a wrapper for Palladium which allows arbitrary code to be executed from within Palladium authenticated code?
I would say that Final Fantasy on the Playstation would count as a good CD based game which was quite fun. FF7 incorporated video footage with live game scenes. The same was true of FF8 which was released during the time of N64. FF9 was released after GameCube came around, so I won't include it in the list.:)
I definitely agree with you that adding a CD drive to a game system increases the cost significantly. But at the same time, it also makes production of the copies of the games much cheaper, which is where they would reclaim their loses on the hardware, I would imagine.
As for the handheld systems, having seen the way my youngest brother tosses that GBA around like a sack of potatoes gives me incentive to agree with you.:) For handhelds, I definitely would go with chips as opposed to CDs.
Please re-read my original post. You missed some points.
Not all missles are the same. There are cruise missles, as you note. They are low-flight, radar evasion missles which uses a sophisticated computer with on-board terrain maps and visual images to compare against. They are designed to detect and evade key targets and strike effectively, being only detected by their explosion or when it is already too late.
That isn't the kind of missle I'm talking about.
Then, there are the smart bombs which are dropped from planes. They essentially course correct on their way down from the point of being dropped, being guided by laser to their targets.
Then, there are ballistic missles. Otherwise known as ICBM's(Inter-Continental Ballistic Missles) which uses an Apogee flight path to head up into the atmosphere and come back down after hitting their apogee. From that point on, they are then self-guided to their destinations from a programmed flight plan.
These are the kinds of missles I'm talking about.
A variation of which would be the Multiple warhead ICBMs.
So you are wrong in regards to the issue of apogee in missle flight paths.
All of the above are hard/nearly impossible to defend against.
The reason is not that one is more intelligent than the other. The reason is because of their relative speeds. In the case of the MW-ICBMs, the sheer number of warheads and their relative distances from each other adds to the difficulty.
We have problems creating a missle defense system because the incoming warheads are travelling towards their targets at speeds which by the time the engaging missle were to detonate, even near the missle, the explosion would have missed the targetted missle. The incoming missle would not even need to evade detection.
With multiple warheads, the problem is made worse by the fact that the warheads now cover a greater region of the sky and it is already unlikely to hit even one with multiple missles. To hit multiple warheads(dozens or hundreds) would be impossible in theory and in practice.
To borrow your cannon example, if a country could launch mortors into the air towards America from a sufficient distance, we would still not be able to do anything about it. On descent, the relative speeds would be high enough to evade our attempts to stop the projectile. Now, imagine hundreds of such mortors.
A "stupid" mortor shell with sufficient range would be able to decimate a city and we would have no way to stop the shells from landing.
It isn't because missles change speed. Planes change speed and we can hit those. It is because of the relative speeds of missles and their relatively small size and profile.
Missles benefit from guidance systems. I never implied or said that they don't. I suggested that a guidance system would be easy to make. Not that it would be "stupid" or that it would be defficient. My post noted that a guidance system can be made simply. Nothing more, nothing less.
For your information, the programming language doesn't matter as much as the response time from the guidance system. If your compiler for C or even Basic is efficient, then the resulting code for the chip(s) in question would be effective as a guidance system.
Btw, if you are in Information Systems, then you are not a CS major. You are headed for a management position. As a CS major, I started programming assembly in a 100 level course. The same was true for C and C++. Both were in my 2nd year of college. And that was only because I was taking plenty of electives. Most people I knew wrote their own programs to control small cars with a modified 8086 motherboard. Ie, self-contained guidance system. You seem to lack the understanding of the underlying hardware and software to really make an informed statement about this topic.
Since I already pointed out that I never implied or said a guidance system wasn't required, the GPS point is moot since a simple program with a serial interface is all that is needed to communicate with a GPS unit. Any micro controller can be wired to interface with such a standard off-the-shelf unit and to the flight control systems.
On the issue of flight control systems. Fins are only part of the equation. I would use minimal profile fins in conjunction with vector controlled thrusters. They would be more effective at higher altittudes to make the missle more effective against high flight spy planes and for long range flight through the atmosphere as an ICBM.
Your lack of understanding of how missles in general work along with your lack of knowledge about how computer systems work results in your flawed interpretation of my post.
In the future, it might be beneficial to yourself if you actually understood what you were talking about or at least research a little before posting with such an derogatory tone.
While it might be true that Sony could be making money now on the PS1 hardware sales, it isn't true of the PS2. At least, not yet.
Console companies typically don't make money off of the hardware consoles. It is normally from the sale of game development licenses and from royalties of the sold games. (Charge for people to develope and charge when games are sold)
A typical development license+hardware/software for the PS2 will run you $10,000.00. To get additioanl development software like Metrowerks' IDE, you will pay anouther $5000.00.
The PS3 will most likely be at a big loss at first. But then again, Sony is intending to market and license the chips to other companies for additional income. The PS3 will most likely be about the same size as the PS2 since most of the chips will be integrated and not spread out into seperate modules. Part of the reason why Sony is going to Cell, I'm sure.
The Xbox was a modified PC motherboard and graphics hardware. The size was dictated by the motherboard. Sony has the option to shape their boards however they might. Same for Nintendo's GameCube.
As for cost of the hardware...
Development of the primary chips is usually in the hundreds of millions if not several billion dollars. Sony invests heavily into R&D.
The chips on the system cost, I would imagine, about $3-5 each to make in bulk amounts in the 100's of thousands. However, their first few thousand will be in small batches to do testing. This adds to the cost. Then there is the cost of the motherboard design and how all the systems would integrate. The mainboard will probably run about $50-$100 each after assembly in raw materials. The drive assembly about half that. Then the power supply and such.
After the labour involved is taken into account and when you consider the numer of units made, you are basically looking at a piece of hardware that would cost about $1000-$2000 if bought in bulk and $3000+ if bought in smaller quantities. You'll probably be looking at a PS3 shelf retail price of $499.99 - $699.99. Why would they sell at a loss? Because they make the bulk of the money in game sales.
A family will probably buy on PS3. But during the life of that PS3, they will buy 20-50 games.
Games are cheap to make. So let's take your average game of $49.99(round to $50).
Store/Retailer takes in $25 or so per game.
The retail price is normally double, at least, of the wholesale price => $25.
Sony takes in about $12.50 from the sale of the game to the stores.
The wholesale price is normally double+ of the profit price. => $12.50
Depending on the royalty arrangement, Sony will probably end up taking another 20% of this: $2.50 or so.
So, after all is said and done:
Store: $25
Sony: $12.50 + $2.50 => $15.00 / game
Sony also collects license fees and the such for the development of the game, but we can ignore that segment.
So, at 20 - 50 games per household: $300 - $750
(Note, my figures are extremely ball-park'ish, so the numbers would vary depending...)
But given that Sony sells millions of the stations on the opening week and will continue to sell millions for the next 6 months or so each month: 10mil + ( 6 * 1mil) => 16million systems in first half year
You take that same number and you get something like $4.8 billion - $12.0 billion gross from game sales. (Tax and overhead cuts into that...)
As more machines are produced, their cost goes down significantly. The first year will be a break-even year. The second year will be major net profits.
So, initially... for the first year or so, hardware is sold at a loss. But the game sales and licenses more than make up for the cost of the hardware.
Given the way Sony designs and builds their hardware, it is highly unlikely that the PS3 will be bigger than the Xbox. More likely, it will be round the size of the PS2.
One might think that the programming interface was part of the security system.;)
Seriously though, Cel computing would be kind of expensive for a one-use device. More likely, a cruise missle can be just as easily guided by say... a high speed DSP tied with a 32-bit RISC processor on a custom board to handle telemetry and image/position recognition.
Such a board would be smaller and cheaper.
One of the reasons why it's always been a mind boggling thing to me why people keep talking about "stealing missles" and/or needing to "steal technology" to build guided missles.
If you are launching a missle up into the air, then having it come back down in a arc, you basically want it to land somewhere within a 10' radius. That is basically a small Basic Stamp, GPS decoder, gyroscope, and digital compass system.
Ground based launch site for telemetry/guidance tracking during the launch and partway into the apogee. Internal guidance with image recognition on the descent.
It would be fair to say that most any 2nd or 3rd year computer science major who's done some image library coding, circuit board soldering, and Stamp programming can easily get a working missle guidance system.
The really hard part is the construction of the missle and the security protocols to lock down the missle. But if you ignore that, the "brain" is easy to build.
relatively speaking.
A PS2 or PS3 would be overkill and a waste of time and money.
IANA[EWdM]
(I Am Not A [Engineer/Weapons Developer/Mathematician])
Sony's Spec claims, like anyone else's spec claims, are just that. Raw Performance specs. How the game designer and game coder makes use of that potential afterwards is the limiting factor, not the machine.
PS2 didn't burn people who had realistic expectations. They promised a great system and month after month of games lined up and they delivered.
The dreamcast was slightly more advanced than the PS2. However, what does that matter if the company can't remain stable? In that regards, Sega dropped the ball and Sony delivered by remaining and deciding to support their customers who owned PS1 game titles. Currently, Sega doesn't support DC anymore. Currently, Sony still supports the PS1 with new games and new hardware.
The Xbox was MS attempting to get into the game market. The box is basically a PC put inside of a large console box. Is it better than the PS2? From a raw performance standpoint, it is. From a game standpoint, it is a good system. But, will they be around in the game market long enough to know whether they will pull a Sega? Or perhaps a MS with constant revisions and upgrades? Who knows. They are the new kid on the block and they have to prove themselves. So far, they are holding their own even though their console sales are suffering.
The point is, you point out that Sony's specs are BS because consoles put out after it have higher specs. That is flawed logic.
Sony delivered the system they said they would deliver. Ie, one which was many fold more powerful than their previous Playstation.
Will Sony's next machine, the PS3, be just as impressive? I certainly hope so. But until it is delivered to the customers, it is a moving target. One which Sony will most likely update with new technology before putting the unit on the production lines.
Regarding Nintendo, my beef with them is their resistance of letting go the "game cartridge" mindset. Which is unfortunate since the cartridge design doesn't slow down pirating or hacking of their systems in the slightest, but it does increase their overhead and thin out their margins.
Regarding PS2 and Xbox, I prefer to think of it this way: If you are comparing the graphics quality of a system, The Xbox, released 2-3 years after another system, The PS2, that speaks pretty highly of the power of the hardware of the PS2 and that of the PS2 game designers and coders.
More reason why in my mind, Sony delivered on the PS2 and will continue to deliver in their future platforms. It is also the reason why I feel secure that my PS1 and PS2 games will still be playable when I'm running on a PS5.:)
Hardly an accident I would imagine. Considering that the other game consoles weren't, at the time, marketing backwards compatibility and considering that Sony's money comes not from the consoles, but from game sales. They would have been stupid to not continue to milk their existing base of games and of future games to be made for the PS1 platform.
The PS1 cpu was also a RISC processor. This makes it an efficient IO controller as well as being a programmable one to boot.
Hardly an accident. Most likely planned strategy to garner continued income from the current base of software for the PS1.
As for Cell computing, it is essentially like a PVM or MPI system. Except with the kind of technology they are talking about for interconnects, it would be like comparing a homebuilt Beowulf cluster with 10mpbs connections to one which uses gigabit connections.
With Cell computing, they are just taking the PS2 scheme of using a PS1 chip as IO to the next level. Only more generalized. With on-chip integration, they would save on circuit board real estate as well as power consumption. The system would also be blazingly fast.
The only problem with multiple processors on the same chip die is the cost.:|
But basically, with Cell computing, they would be able to easily emulate the PS2 with special enhancements. Heck, they would be able to emulate mulitple PS2s so multiple people can play on multiple screens, all from the same box.
From their press releases, you would also be able to link up the various PS3's into one large computing unit.
That in my mind makes buying the PS3 a very desirable proposition. Owning more than one might actually be rather cool. >:)
Watched the show(thanks, Tivo!) and quite frankly, it was a waste of time.
The show didn't show much in the way of explaining why the animals would evolve the way they do and even why humans would leave the planet entirely.
Some problems I had with the program's assumptions and ideas:
Thinking Squid On Land and In Trees: I have problems with the show's idea of squid leaving the waters and going onland. In the case of the "small squid" who swing from branch to branch, that makes no sense. Unless the oceans began to dry up or unless the squids began to explore land and develop a skeleton, they would be immobile on land. The idea that "because they need to navigate complex jungles, their brains will increase in size" is complete nonsense. Brain size doesn't increase from sheer need to navigate. Brain performance can certainly be maintained or improved, but it is the addition of proteins and additional nutrients which can allow a brain to grow. Through this and selective death/life, they might evolve larger brains, but they would need to evolve lungs and some way of moving outside of water first. The giant squids going on land with "rings of muscles" like earthworms is pure BS. Without a skeletal structure, that thing would be the size of a small snake at best and certainly not the behemoth it is made out to be. There are no basis for the kinds of animals shown on the purely fictional and fantasy-based show.
Another funny and incorrect one was the algae harvesting termites. We already have insects which farm. They are ants and they grow mold/fungi. But this isn't harvested from the outside or from other animals. It is cultivated and carried from site to site. It would also make no sense for them to attempt to grow a sunlight driven crop. Sunlight requires openings in the mounds which would let rain and predators in. It would also restrict the efficiency of their colony. Mold and Fungi grow in the dark. They would be more efficient and be the correct evolutionary choice.
Sharks in the future: They seem to believe that sharks would start communicating by light patterns, hunt in packs, and swim slower(25knots). What is this? Current sharks can already swim between 21-35 knots per hour(25-40 miles per hour). Current sharks have a keen sense of smell. Current sharks hunt alone. It is only after a kill has occured that the blood draws in other sharks. The shape of the current sharks are streamlined and sleek. They are already efficient killing machines. The designs have horns and other oddities on the head would would serve no purpose in a hunt. The additional drag wold lower their chances for survival and their slower speed would most likely render them unable to catch prey. The only likely evolutionary change I see would be an increse in the size of the liver to allow them to go for longer periods of time without food. This would allow them to go through dead parts of the ocean without fear of starving part-way through.
In general, I just found the show to be horribly inaccurate and lacking in entertainment value. They people who created the show should go back and have a good look at biology, physics, and concepts of evolution in general.
If it was a for-pay movie, I'd ask for my money back.
It would seem that DNA would serve better as a compression algorithm as opposed to actual raw data storage.
Take any living thing as an example. The entire physical being of the creature is derived from the processing of the DNA or RNA. One can say that the compression ratios are quite spectacular.
Instead of attempting to encode raw data into the genome or unused portions of the genome, we should look for ways of having the genes produce the data as an output of a living thing.
For example:
Book Trees: The genetic structure of the tree is encoded with complete entries for a book. Each leaf produced by the tree would have a text pattern emerge from it to form readable text. The tree and a forest can serve as a living library.
Genetic Tattoos: Imagine having tattoos change and emerge on your skin as a genetic pattern. You would be able to have images scanned and then imprinted as an encoded program to be inserted into your genes to produce the image as output on your skin.
At Birth Knowledge: Birds or other animals born with genetically encoded knowledge and can be passed down with each generation. People can also be given such knowledge/instincts genetically.
It just seems that the DNA should be used for what it is used best for. Ie, as a compressed form of information/program to produce the readable output. Like Postscript.
I'm sorry, but does this sound alot like what happened during the "Dot Com" craze?
Ie, people were offered relatively low salaries and were paid in stock options. Which, should the company grow and become profitable... would be worth millions... and if the company should fail or never recieve funding, would be worth less than the notice they are credited on.
Granted, to stay networked and to keep one's skills up are good things if you want to get back into the job market. But to work for essentially what amounts to potential tips and no salary, no health benefits... that is essentially a what I would call a fool's gamble. A desperate fool, but a fool nonetheless.
These people obviously have job skills, why work for virtual stock when they can put that same energy into working for real dollars? The real dollars would be less than what they are used to, but it would be better than zero.
I dunno.
I'm sure the IRS and SEC would be interested in people who are collecting unemployment and yet, at the same time are working/gambling for virutal pay.
This can result in some pretty horrible fines for those who are collecting and working "off the record" for these companies...
Slave labour and a new form of accounting: Zero overhead in terms of personel, but pure profit when it comes to work generated. Since these people aren't employees or even consultants and what they are being offered are stock options, they may not even have a leg to stand on if the company decides to shaft them.
It sets a bad precedent and it sets up a whole group of hard working people for yet another harder fall.
True, so why not get a bunch of those 5400 RPM drives... say... 3 or 4 and RAID stripe them? That way, you would be able to dump your fragmented disk data to those drives and spool them from there onto the 16x DVD burner.
You'd suffer initial spool time to the array, but that way, you won't lose anything should the striped array get munged since it's only being used as a spool space for the burner.
It would also allow you to spool multiple disk images in preparation for lots of burns @ 16x.
Failing that, you could always just pickup a few of those SCSI-IDE adapters for IDE drives and hook them up to a scsi chain. You'd be able to software stripe them with LVM/lvm-tools and get similar performance.
All in all, it is not impossible or even difficult to achieve solid sustained throughput at those rates. It isn't with single modern/recent ATA66/100/133 disks and downright easy with a RAID setup.
Another possibility is to have an external 7200RPM firewire drive setup dedicated to the spooling task.
LED's and Fiber in/out connections would consume more power than say direct serial links. The serial links would also be less bulky since you would not need to have conversion modules.
There would be no speed advantage since you would still have the serial links hooked up to the LED's or have 100TX network connections routed through a fiber module. Your MAX available throughput/bandwidth would still be limited.
A serial-to-serial setup would allow you to basically make a metal/copper matrix grid to plug in each module into to provide power/signal lines/network hub/control lines/remote terminal matrix.
That way, you would be able to have the modules almost side/by/side on a tray. The tray itself being a power/data bus matrix for the modules.
Well, pure diamond won't be a problem since the doping process for diamond won't be the same for Si.
If I remember correctly, Si is doped by overheating it and exposing it to a gaseos environment of the required doping agent. This causes a layer to be effectively doped. Subsequent UV/mask/doping layers is what builds up the Si based chips.
With a diamond, I don't be believe that that would be feasible unless they incorporate the gaseous doping agents during the diamond layer process.
All in all, the cost would be astronomical. What did the article say? It would costs thousands of dollars of a few square mm of a diamond wafer whereas the same for Si wafers is virtually free.
It will be an interesting period... with what would be, I'm sure, equally spectactular plan "accidents". >:)
Having said all of that though, since the source code is GPL, it shouldn't take much for some smart and document friendly coder to _document_ the code itself. Ie, make the documentation part of the code.
Adds to software code bloat, but since it can be configured and ifdef'd out of the final compile, it won't result in actual binary code bloat.
ymmv.
I agree with OpenSource. But when it comes right down to it, a business is going to do what it can to survive. And that often times means holding the line that the agreement draws in the sand.
I would have to definitely agree with the charging of the fee for support.
That company has made the source freely available to those who would use it. They work on it and improve it, fixing bugs as necessary. But the support itself costs money. If they were just another open source coder, then I'm sure they'd more than happily help you for free and maybe a thanks. But they are a company and they are charging support fees.
This is definitely within their right to do so in both the spirit and letter of open source. Though whether or not it agrees with different peoples' versions and understandings of open source is another matter.
As for documentation, what kind of documentation is being referred to? A help file? A Howto? Or a custom tailored document to help the user?
As for the having people basically sign a NDA to not disclose how they were shown to perform the install, that is something which is beyond the scope of "open source".
The reasoning is that open source covers the accessibility of the source code by the masses in a way which the masses can understand. If the code is beyond the means of the masses to understand, assuming it has not been obfuscated, then they require support to assist them with getting the code/app to work with their system. This help is billable and could very well be restricted information. Not from a security standpoint, but from a commodity standpoint. Ie, it is the model upon which their business is based.
One can think of it as buying software which comes with basic instructions which works for some, but doesn't for others. You can always pay more to obtain support and/or documents to better assist you, but you are not allowed to copy that document since it is copyrighted and is essentially the incentive for people to purchase support.
So I would agree with your assessment with the contractor example.
Some might point out that RedHat/etc are charging for support as well in a similar manner. Though I do not know if they are having people NDA'd.
Diamond IS an insulator. An insulator to electrical current, but it is a very efficient heat conductor/radiator.
Silicon is also an insulator. It is doped with Oxygen to improve the insulation factor and doped with other compounds(arsenic,gallium,phosphor,etc.) to improve conductivity and to create P and N regions.
The question would be how one would get the doping to work for Diamonds.
Personally, I think they will have to revert to particle accelerators. Since bombarding the carbon matrix of a diamon would be equivelent to doping... albeit at a much lower concentration level. Works well enough to turn a clear diamond blue...
True quartz has piezoelectric factors... but Diamonds can burn! Just like coal... except they explode due to the pressures inherit in their creation...
Overclocker1: Shit, wonder how many more degrees I can clock this baby...
Overclocker2: Did you remove the oxygen?
Overclocker1: What? Why?
*BOOM*
Okay, maybe not that drastic... but it's pretty nasty.
Good point about the portmaster size. And about the cell computing as well.
Along that vein, what about the serial lines handling message passing and using the built-in 100TX as a communication bus. Since each device can have an arbitrary number of ethernet address identities, but only one can be active at any one time, one cell can take over for another cell by taking on its identity when the cell "dies".
If these are assembled on "pluggable" modules, then a module which is "dead" can be pulled out and replaced while the system is running.
Slashdot Mirror
Well, while Tivo will most likely not muck with the ratio, they might be doing on-the-fly recompression and resampling at the same time to store the HD programming in a smaller amount of disk space and possibly using a different algorithm to encode... perhaps Mpeg4?
It makes sense if you plan on using essentially the same setup as other machines on the production line. What you would change would be the encoder/decoders, possibly upgrade the main CPU on the board, and up the drive to a 120GB single hard drive in the unit.
If they reduced the bitrate of the video stream when they save to disk, they would be able to allow for different levels of "quality" and thus be able to maximize the number of hours to store on the unit.
That'd be my guess, at any rate.
Definitely realistic, but I wonder whether or not Tivo will be decompressing on-the-fly and re-compressing and re-sampling to a format which will be slightly degraded, but take up less space?
For example, the current Tivos allow different grades of quality for storage, each one taking less space or more space. I wonder whether or not Tivo will be decompressing the HD stream/signal and then recompressing it at different rates/qualities.
Another possibility is that Tivo might start making use of Mpeg4 instead of Mpeg2 for their internal encoding. This would allow them to store a stream with HD dimensions without taking up the HD space.
Side Note: Wonder if they will be putting out software upgrades to those of us who own Tivo Series2's and older Tivos to make use of HD at somewhat degraded resolution? (Current Tivo2: 199HourCapacity)
So does this mean that my SoundCard will need Palladium as well?
Reasoning being that quite a few, if not all, expansion cards have their own BIOS which gets executed at startup to "mingle" on the pci/isa/etc bus.
And... what's to prevent someone from writing a wrapper for Palladium which allows arbitrary code to be executed from within Palladium authenticated code?
*LOL*
So I guess the laser printers left burn marks...
The dye subs left you with a tattoo...
And the melted wax printers left you with 3rd degree burns? ;)
I would say that Final Fantasy on the Playstation would count as a good CD based game which was quite fun. FF7 incorporated video footage with live game scenes. The same was true of FF8 which was released during the time of N64. FF9 was released after GameCube came around, so I won't include it in the list. :)
I definitely agree with you that adding a CD drive to a game system increases the cost significantly. But at the same time, it also makes production of the copies of the games much cheaper, which is where they would reclaim their loses on the hardware, I would imagine.
As for the handheld systems, having seen the way my youngest brother tosses that GBA around like a sack of potatoes gives me incentive to agree with you. :) For handhelds, I definitely would go with chips as opposed to CDs.
And yes, night trap sucked pretty badly.
Please re-read my original post. You missed some points.
Not all missles are the same. There are cruise missles, as you note. They are low-flight, radar evasion missles which uses a sophisticated computer with on-board terrain maps and visual images to compare against. They are designed to detect and evade key targets and strike effectively, being only detected by their explosion or when it is already too late.
That isn't the kind of missle I'm talking about.
Then, there are the smart bombs which are dropped from planes. They essentially course correct on their way down from the point of being dropped, being guided by laser to their targets.
Then, there are ballistic missles. Otherwise known as ICBM's(Inter-Continental Ballistic Missles) which uses an Apogee flight path to head up into the atmosphere and come back down after hitting their apogee. From that point on, they are then self-guided to their destinations from a programmed flight plan.
These are the kinds of missles I'm talking about.
A variation of which would be the Multiple warhead ICBMs.
So you are wrong in regards to the issue of apogee in missle flight paths.
All of the above are hard/nearly impossible to defend against.
The reason is not that one is more intelligent than the other. The reason is because of their relative speeds. In the case of the MW-ICBMs, the sheer number of warheads and their relative distances from each other adds to the difficulty.
We have problems creating a missle defense system because the incoming warheads are travelling towards their targets at speeds which by the time the engaging missle were to detonate, even near the missle, the explosion would have missed the targetted missle. The incoming missle would not even need to evade detection.
With multiple warheads, the problem is made worse by the fact that the warheads now cover a greater region of the sky and it is already unlikely to hit even one with multiple missles. To hit multiple warheads(dozens or hundreds) would be impossible in theory and in practice.
To borrow your cannon example, if a country could launch mortors into the air towards America from a sufficient distance, we would still not be able to do anything about it. On descent, the relative speeds would be high enough to evade our attempts to stop the projectile. Now, imagine hundreds of such mortors.
A "stupid" mortor shell with sufficient range would be able to decimate a city and we would have no way to stop the shells from landing.
It isn't because missles change speed. Planes change speed and we can hit those. It is because of the relative speeds of missles and their relatively small size and profile.
Missles benefit from guidance systems. I never implied or said that they don't. I suggested that a guidance system would be easy to make. Not that it would be "stupid" or that it would be defficient. My post noted that a guidance system can be made simply. Nothing more, nothing less.
For your information, the programming language doesn't matter as much as the response time from the guidance system. If your compiler for C or even Basic is efficient, then the resulting code for the chip(s) in question would be effective as a guidance system.
Btw, if you are in Information Systems, then you are not a CS major. You are headed for a management position. As a CS major, I started programming assembly in a 100 level course. The same was true for C and C++. Both were in my 2nd year of college. And that was only because I was taking plenty of electives. Most people I knew wrote their own programs to control small cars with a modified 8086 motherboard. Ie, self-contained guidance system. You seem to lack the understanding of the underlying hardware and software to really make an informed statement about this topic.
Since I already pointed out that I never implied or said a guidance system wasn't required, the GPS point is moot since a simple program with a serial interface is all that is needed to communicate with a GPS unit. Any micro controller can be wired to interface with such a standard off-the-shelf unit and to the flight control systems.
On the issue of flight control systems. Fins are only part of the equation. I would use minimal profile fins in conjunction with vector controlled thrusters. They would be more effective at higher altittudes to make the missle more effective against high flight spy planes and for long range flight through the atmosphere as an ICBM.
Your lack of understanding of how missles in general work along with your lack of knowledge about how computer systems work results in your flawed interpretation of my post.
In the future, it might be beneficial to yourself if you actually understood what you were talking about or at least research a little before posting with such an derogatory tone.
It is obvious what is wrong with you.
While it might be true that Sony could be making money now on the PS1 hardware sales, it isn't true of the PS2. At least, not yet.
Console companies typically don't make money off of the hardware consoles. It is normally from the sale of game development licenses and from royalties of the sold games. (Charge for people to develope and charge when games are sold)
A typical development license+hardware/software for the PS2 will run you $10,000.00. To get additioanl development software like Metrowerks' IDE, you will pay anouther $5000.00.
The PS3 will most likely be at a big loss at first. But then again, Sony is intending to market and license the chips to other companies for additional income. The PS3 will most likely be about the same size as the PS2 since most of the chips will be integrated and not spread out into seperate modules. Part of the reason why Sony is going to Cell, I'm sure.
The Xbox was a modified PC motherboard and graphics hardware. The size was dictated by the motherboard. Sony has the option to shape their boards however they might. Same for Nintendo's GameCube.
As for cost of the hardware...
Development of the primary chips is usually in the hundreds of millions if not several billion dollars. Sony invests heavily into R&D.
The chips on the system cost, I would imagine, about $3-5 each to make in bulk amounts in the 100's of thousands. However, their first few thousand will be in small batches to do testing. This adds to the cost. Then there is the cost of the motherboard design and how all the systems would integrate. The mainboard will probably run about $50-$100 each after assembly in raw materials. The drive assembly about half that. Then the power supply and such.
After the labour involved is taken into account and when you consider the numer of units made, you are basically looking at a piece of hardware that would cost about $1000-$2000 if bought in bulk and $3000+ if bought in smaller quantities. You'll probably be looking at a PS3 shelf retail price of $499.99 - $699.99. Why would they sell at a loss? Because they make the bulk of the money in game sales.
A family will probably buy on PS3. But during the life of that PS3, they will buy 20-50 games.
Games are cheap to make. So let's take your average game of $49.99(round to $50).
Store/Retailer takes in $25 or so per game.
The retail price is normally double, at least, of the wholesale price => $25.
Sony takes in about $12.50 from the sale of the game to the stores.
The wholesale price is normally double+ of the profit price. => $12.50
Depending on the royalty arrangement, Sony will probably end up taking another 20% of this: $2.50 or so.
So, after all is said and done:
Store: $25
Sony: $12.50 + $2.50 => $15.00 / game
Sony also collects license fees and the such for the development of the game, but we can ignore that segment.
So, at 20 - 50 games per household: $300 - $750
(Note, my figures are extremely ball-park'ish, so the numbers would vary depending...)
But given that Sony sells millions of the stations on the opening week and will continue to sell millions for the next 6 months or so each month: 10mil + ( 6 * 1mil) => 16million systems in first half year
You take that same number and you get something like $4.8 billion - $12.0 billion gross from game sales. (Tax and overhead cuts into that...)
As more machines are produced, their cost goes down significantly. The first year will be a break-even year. The second year will be major net profits.
So, initially... for the first year or so, hardware is sold at a loss. But the game sales and licenses more than make up for the cost of the hardware.
Given the way Sony designs and builds their hardware, it is highly unlikely that the PS3 will be bigger than the Xbox. More likely, it will be round the size of the PS2.
One might think that the programming interface was part of the security system. ;)
Seriously though, Cel computing would be kind of expensive for a one-use device. More likely, a cruise missle can be just as easily guided by say... a high speed DSP tied with a 32-bit RISC processor on a custom board to handle telemetry and image/position recognition.
Such a board would be smaller and cheaper.
One of the reasons why it's always been a mind boggling thing to me why people keep talking about "stealing missles" and/or needing to "steal technology" to build guided missles.
If you are launching a missle up into the air, then having it come back down in a arc, you basically want it to land somewhere within a 10' radius. That is basically a small Basic Stamp, GPS decoder, gyroscope, and digital compass system.
Ground based launch site for telemetry/guidance tracking during the launch and partway into the apogee. Internal guidance with image recognition on the descent.
It would be fair to say that most any 2nd or 3rd year computer science major who's done some image library coding, circuit board soldering, and Stamp programming can easily get a working missle guidance system.
The really hard part is the construction of the missle and the security protocols to lock down the missle. But if you ignore that, the "brain" is easy to build.
relatively speaking.
A PS2 or PS3 would be overkill and a waste of time and money.
IANA[EWdM]
(I Am Not A [Engineer/Weapons Developer/Mathematician])
Sony's Spec claims, like anyone else's spec claims, are just that. Raw Performance specs. How the game designer and game coder makes use of that potential afterwards is the limiting factor, not the machine.
PS2 didn't burn people who had realistic expectations. They promised a great system and month after month of games lined up and they delivered.
The dreamcast was slightly more advanced than the PS2. However, what does that matter if the company can't remain stable? In that regards, Sega dropped the ball and Sony delivered by remaining and deciding to support their customers who owned PS1 game titles. Currently, Sega doesn't support DC anymore. Currently, Sony still supports the PS1 with new games and new hardware.
The Xbox was MS attempting to get into the game market. The box is basically a PC put inside of a large console box. Is it better than the PS2? From a raw performance standpoint, it is. From a game standpoint, it is a good system. But, will they be around in the game market long enough to know whether they will pull a Sega? Or perhaps a MS with constant revisions and upgrades? Who knows. They are the new kid on the block and they have to prove themselves. So far, they are holding their own even though their console sales are suffering.
The point is, you point out that Sony's specs are BS because consoles put out after it have higher specs. That is flawed logic.
Sony delivered the system they said they would deliver. Ie, one which was many fold more powerful than their previous Playstation.
Will Sony's next machine, the PS3, be just as impressive? I certainly hope so. But until it is delivered to the customers, it is a moving target. One which Sony will most likely update with new technology before putting the unit on the production lines.
Regarding Nintendo, my beef with them is their resistance of letting go the "game cartridge" mindset. Which is unfortunate since the cartridge design doesn't slow down pirating or hacking of their systems in the slightest, but it does increase their overhead and thin out their margins.
Regarding PS2 and Xbox, I prefer to think of it this way: If you are comparing the graphics quality of a system, The Xbox, released 2-3 years after another system, The PS2, that speaks pretty highly of the power of the hardware of the PS2 and that of the PS2 game designers and coders.
More reason why in my mind, Sony delivered on the PS2 and will continue to deliver in their future platforms. It is also the reason why I feel secure that my PS1 and PS2 games will still be playable when I'm running on a PS5. :)
Hardly an accident I would imagine. Considering that the other game consoles weren't, at the time, marketing backwards compatibility and considering that Sony's money comes not from the consoles, but from game sales. They would have been stupid to not continue to milk their existing base of games and of future games to be made for the PS1 platform.
The PS1 cpu was also a RISC processor. This makes it an efficient IO controller as well as being a programmable one to boot.
Hardly an accident. Most likely planned strategy to garner continued income from the current base of software for the PS1.
As for Cell computing, it is essentially like a PVM or MPI system. Except with the kind of technology they are talking about for interconnects, it would be like comparing a homebuilt Beowulf cluster with 10mpbs connections to one which uses gigabit connections.
With Cell computing, they are just taking the PS2 scheme of using a PS1 chip as IO to the next level. Only more generalized. With on-chip integration, they would save on circuit board real estate as well as power consumption. The system would also be blazingly fast.
The only problem with multiple processors on the same chip die is the cost. :|
But basically, with Cell computing, they would be able to easily emulate the PS2 with special enhancements. Heck, they would be able to emulate mulitple PS2s so multiple people can play on multiple screens, all from the same box.
From their press releases, you would also be able to link up the various PS3's into one large computing unit.
That in my mind makes buying the PS3 a very desirable proposition. Owning more than one might actually be rather cool. >:)
Watched the show(thanks, Tivo!) and quite frankly, it was a waste of time.
The show didn't show much in the way of explaining why the animals would evolve the way they do and even why humans would leave the planet entirely.
Some problems I had with the program's assumptions and ideas:
In general, I just found the show to be horribly inaccurate and lacking in entertainment value. They people who created the show should go back and have a good look at biology, physics, and concepts of evolution in general.
If it was a for-pay movie, I'd ask for my money back.
It would seem that DNA would serve better as a compression algorithm as opposed to actual raw data storage.
Take any living thing as an example. The entire physical being of the creature is derived from the processing of the DNA or RNA. One can say that the compression ratios are quite spectacular.
Instead of attempting to encode raw data into the genome or unused portions of the genome, we should look for ways of having the genes produce the data as an output of a living thing.
For example:
It just seems that the DNA should be used for what it is used best for. Ie, as a compressed form of information/program to produce the readable output. Like Postscript.
They were neurons or rather, they were nueral tissue suspended in a gel mixture. They are used as logic circuits and AI for the ship.
I'm sorry, but does this sound alot like what happened during the "Dot Com" craze?
Ie, people were offered relatively low salaries and were paid in stock options. Which, should the company grow and become profitable... would be worth millions... and if the company should fail or never recieve funding, would be worth less than the notice they are credited on.
Granted, to stay networked and to keep one's skills up are good things if you want to get back into the job market. But to work for essentially what amounts to potential tips and no salary, no health benefits... that is essentially a what I would call a fool's gamble. A desperate fool, but a fool nonetheless.
These people obviously have job skills, why work for virtual stock when they can put that same energy into working for real dollars? The real dollars would be less than what they are used to, but it would be better than zero.
I dunno.
I'm sure the IRS and SEC would be interested in people who are collecting unemployment and yet, at the same time are working/gambling for virutal pay.
This can result in some pretty horrible fines for those who are collecting and working "off the record" for these companies...
Slave labour and a new form of accounting: Zero overhead in terms of personel, but pure profit when it comes to work generated. Since these people aren't employees or even consultants and what they are being offered are stock options, they may not even have a leg to stand on if the company decides to shaft them.
It sets a bad precedent and it sets up a whole group of hard working people for yet another harder fall.
Sounds like a bad deal to me.
True, so why not get a bunch of those 5400 RPM drives ... say... 3 or 4 and RAID stripe them? That way, you would be able to dump your fragmented disk data to those drives and spool them from there onto the 16x DVD burner.
You'd suffer initial spool time to the array, but that way, you won't lose anything should the striped array get munged since it's only being used as a spool space for the burner.
It would also allow you to spool multiple disk images in preparation for lots of burns @ 16x.
Failing that, you could always just pickup a few of those SCSI-IDE adapters for IDE drives and hook them up to a scsi chain. You'd be able to software stripe them with LVM/lvm-tools and get similar performance.
All in all, it is not impossible or even difficult to achieve solid sustained throughput at those rates. It isn't with single modern/recent ATA66/100/133 disks and downright easy with a RAID setup.
Another possibility is to have an external 7200RPM firewire drive setup dedicated to the spooling task.
That's what a RAID is for. >:)
Finally, all that... er... data... can be backed up quickly!
LED's and Fiber in/out connections would consume more power than say direct serial links. The serial links would also be less bulky since you would not need to have conversion modules.
There would be no speed advantage since you would still have the serial links hooked up to the LED's or have 100TX network connections routed through a fiber module. Your MAX available throughput/bandwidth would still be limited.
A serial-to-serial setup would allow you to basically make a metal/copper matrix grid to plug in each module into to provide power/signal lines/network hub/control lines/remote terminal matrix.
That way, you would be able to have the modules almost side/by/side on a tray. The tray itself being a power/data bus matrix for the modules.
Well, pure diamond won't be a problem since the doping process for diamond won't be the same for Si.
If I remember correctly, Si is doped by overheating it and exposing it to a gaseos environment of the required doping agent. This causes a layer to be effectively doped. Subsequent UV/mask/doping layers is what builds up the Si based chips.
With a diamond, I don't be believe that that would be feasible unless they incorporate the gaseous doping agents during the diamond layer process.
All in all, the cost would be astronomical. What did the article say? It would costs thousands of dollars of a few square mm of a diamond wafer whereas the same for Si wafers is virtually free.
It will be an interesting period... with what would be, I'm sure, equally spectactular plan "accidents". >:)
*LOL*
Let me get this straight. You call the guy a biotch and he _only_ tries to skrew you out of 10 bucks? Should count yourself luck. >:)
Funny, but your example is wrong. Someone else who replied got it right: The car is free. The howto manual isn't.
Having said all of that though, since the source code is GPL, it shouldn't take much for some smart and document friendly coder to _document_ the code itself. Ie, make the documentation part of the code.
Adds to software code bloat, but since it can be configured and ifdef'd out of the final compile, it won't result in actual binary code bloat.
ymmv.
I agree with OpenSource. But when it comes right down to it, a business is going to do what it can to survive. And that often times means holding the line that the agreement draws in the sand.
What you need is a strong gust of wind. >:)
Once again, IANAL.
*grins* Abosolutely nothing.
Though at the point when someone goes through the code and understands it... they may feel inclined to start their own version of it.
I would have to definitely agree with the charging of the fee for support.
That company has made the source freely available to those who would use it. They work on it and improve it, fixing bugs as necessary. But the support itself costs money. If they were just another open source coder, then I'm sure they'd more than happily help you for free and maybe a thanks. But they are a company and they are charging support fees.
This is definitely within their right to do so in both the spirit and letter of open source. Though whether or not it agrees with different peoples' versions and understandings of open source is another matter.
As for documentation, what kind of documentation is being referred to? A help file? A Howto? Or a custom tailored document to help the user?
As for the having people basically sign a NDA to not disclose how they were shown to perform the install, that is something which is beyond the scope of "open source".
The reasoning is that open source covers the accessibility of the source code by the masses in a way which the masses can understand. If the code is beyond the means of the masses to understand, assuming it has not been obfuscated, then they require support to assist them with getting the code/app to work with their system. This help is billable and could very well be restricted information. Not from a security standpoint, but from a commodity standpoint. Ie, it is the model upon which their business is based.
One can think of it as buying software which comes with basic instructions which works for some, but doesn't for others. You can always pay more to obtain support and/or documents to better assist you, but you are not allowed to copy that document since it is copyrighted and is essentially the incentive for people to purchase support.
So I would agree with your assessment with the contractor example.
Some might point out that RedHat/etc are charging for support as well in a similar manner. Though I do not know if they are having people NDA'd.
Take with big whopping grains of salt for IANAL.
Diamond IS an insulator. An insulator to electrical current, but it is a very efficient heat conductor/radiator.
Silicon is also an insulator. It is doped with Oxygen to improve the insulation factor and doped with other compounds(arsenic,gallium,phosphor,etc.) to improve conductivity and to create P and N regions.
The question would be how one would get the doping to work for Diamonds.
Personally, I think they will have to revert to particle accelerators. Since bombarding the carbon matrix of a diamon would be equivelent to doping... albeit at a much lower concentration level. Works well enough to turn a clear diamond blue...
True quartz has piezoelectric factors... but Diamonds can burn! Just like coal... except they explode due to the pressures inherit in their creation...
Overclocker1: Shit, wonder how many more degrees I can clock this baby...
Overclocker2: Did you remove the oxygen?
Overclocker1: What? Why?
*BOOM*
Okay, maybe not that drastic... but it's pretty nasty.
Good point about the portmaster size. And about the cell computing as well.
Along that vein, what about the serial lines handling message passing and using the built-in 100TX as a communication bus. Since each device can have an arbitrary number of ethernet address identities, but only one can be active at any one time, one cell can take over for another cell by taking on its identity when the cell "dies".
If these are assembled on "pluggable" modules, then a module which is "dead" can be pulled out and replaced while the system is running.
Multi processing as well as redundancey?