If the studios will save tons of money on distribution costs, then it would seem reasonable for them to contribute in some fashion (up front cash? lower fees to the theaters?) to the large costs of the digital projectors.
I agree with the basic sentiment-- like any other language, you learn to write code by writing. The missing piece in your loop, particularly for beginners, is the editor who marks up their attempts, who tells them "this part is good" and "that part sucks". I was a self-taught programmer over 30 years ago and in that first six unsupervised months, I picked up bad habits that took me years to break. If you're just starting, get someone (hopefully someone who knows what they're doing) to edit some of your code.
Once you reach a certain level of proficiency, you can also progress by reading other people's code. Almost every language has some standard "idioms" for expressing common ideas well, and you can learn a lot of that from other people's code.
An OC-12 (about 600 Mbps usable) per 60,000 subs gives each approx 3.24GB (gigabytes) each way per month, assuming absolutely fair sharing and no wasted bits, and that I'm doing muliplication and division properly. Unfortunately, at 4 AM the link is only about 20% occupied, so lots of bits are "wasted" each day.
Is the "national provider" a regulated monopoly of some sort in Canada? The economics for capital recovery are a lot different when you don't have to worry about a competitor entering the market...
Since you mention capping and volume limits, I'll assume you're talking about the last-mile provider. The cable and cable-modem business is the one I know the most about, so I'll do some back-of-the-envelope calculations for that. Assume for the moment that the data business has to bear all of the costs.
The first major component of the monthly bill is the cost to construct the network. Hybrid-fiber-coax, the architecture required to operate two-way services, costs about $40,000 per mile for materials and installation. A mile of plant passes surprisingly few houses on average-- call it 100 for ease of calculation. Assume 20% of houses passed subscribe to the data service, and the construction cost per subscriber is $2,000. If the company wants to get its construction costs back in five years (we'll ignore issues of interest and risk for the moment), the monthly revenue needed is $33.33. There's some head-end equipment (a cable-modem termination system costs about $30,000) but those are spread over a lot more subscribers. Call it $35/month to pay for the network itself.
Now consider other recurring monthly expenses. There's the billing system that generates monthly bills for millions of subscribers. There's the customer-care systems. There's the salaries for the people who maintain the equipment and answer the phones. There's the rent/electric/sewer/etc for the space where those people work. There are indirect costs associated with those employees-- if you have 5,000 employees who operate the network and take care of the customers, you need a personnel office, a finance office, etc. This type of cost can easily run to $20/month per subscriber.
You need connections to the larger Internet. An OC-12 (600 Mbps) connected to somebody's backbone costs about $120,000/month, and you need one of those for approximately every 60,000 subscribers. That's another $2/month per subscriber. There are a bunch of "little" costs like that, let's guess that they add up to $5/month per subscriber.
When you set out to do this on a large scale and in a short period of time, you have to borrow an enormous amount of money. The people who loan it to you want interest. Comcast is buying AT&T Broadband, with 16M subscribers (counting video and telephony subs as well as cable-modem subs) and assuming $20B in debt. With an average interest rate of 7.5%, the monthly interest payment works out to around $7.80/month per subscriber. Add all that up and you have a total of about $67.80/month.
To be honest, not all of the network construction costs should be charged to the data service, and the billing system and personnel are spread across other services as well. Even so, an allocated cost of $40/month for data service is not a bad estimate, and that doesn't include anything for profit.
According to some economic theorists, all profits beyond the long-term cost of money are due to some form of monopoly. For example, if you own the only factory capable of making widgets, you have a temporary monopoly (until someone else builds a factory). If you invent a cheaper way to make a widget and install that in your factory, you have a "monopoly" that results in higher profits than your competitors, at least until someone duplicates it or invents something better.
My own opinion is that the constitutional grants of a temporary monopoly on intellectual property recognize this situation as well as recognizing the ease with which IP can be copied.
At least in law, corporations are "persons" and enjoy many of the same Constitutional rights as "natural persons" like you or me. For example, a piece of land owned by a corporation may not be siezed without the same due process required if the land were owned by a single individual. Corporate "speech," such as an advertising brochure or an anonymous editorial in a newpaper, is provided the same protections as are guaranteed to you or me.
I believe that the case law establishing that corporations are "persons" goes back to the 1860s when certain railroads were allowed to receive land grants which Congress had promised to "persons" satisfying certain requirements about using the property.
I do agree with your general sentiment that it is easy for these very rich, very long-lived "persons" to abuse laws intended for mere humans.
I recall reading somewhere that the cost of a state-of-the-art fab line was doubling roughly every 36 months (approximately half the Moore's Law rate). The article predicted that about 2010, the fixed costs for such a line would make chips produced on it so expensive that there would be very little market for them. Can anyone with actual experience in the field comment on how that cost prediction is holding up?
Working in a technical field, the half-life on knowledge is about five years-- that is, five years from now, half of your technical knowledge will be obsolete. You'll have to learn new stuff to replace it. One of the side effects of a decent college program is that you will learn a lot about the process of learning, which will serve you well in a field where you're doomed (or blessed, depending on your perspective) to be a part-time student forever. Few high-school graduates have acquired the needed skills for that kind of ongoing study.
I lived through the last major antitrust case in this country, the breakup of AT&T in the first half of the 80's. At Bellcore (now Telcordia), on the local service monopoly side of that split, annual training in antitrust was mandatory for all salaried employees. There were things you could do or say, and things that you couldn't do or say, and the differences were spelled out in detail. If nothing else, you learned to confine your discussions about things like "cutting off their air supply" to verbal comments when no outsiders were present. Everyone had to sign the legal documents certifying they had attended and understood the material. Failure to attend was a serious offense and could get you fired.
As a convicted monopolist, MS would be well served by instituting the same type of training for their staff. It will save them an enormous amount of grief in the long run.
I saw a demo system involving a mechanical work stand that held and moved a large assembly under voice control. The human attaching various components and wiring controlled the positioning with simple commands: higher, lower, rotate right, etc. Since they didn't have to put down the components or their tools to adjust the position, the work went much faster.
Lots of the new LED replacement bulbs for traffic lights have been going in here in Colorado. One unanticipated (I hope) side effect: when it snows and the wind is blowing hard, the sunscreens over the lights that face into the wind eventually accumulate enough snow to obscure the light. Fortunately, the LED bulbs are bright enough you can still tell which light is on. This doesn't seem to happen with the old incandescent lamps, I assume because they produce enough waste heat to melt the snow and keep it from building up.
The page includes the statement that a golf-ball-sized device could have the light output equivalent to 10,000 watts of incandescent bulbs (100 100-watt bulbs). Granted that they are more efficient than incandescent, how hot is such a device likely to get during operation? Even at 90% efficiency, a device that size converting 1,000 watts to heat is going to be damned hot.
I feel badly for teachers, who seem to be caught in the economic trap of a job whose productivity cannot be significantly increased. Compared to thirty years ago, (1) individual classes are pretty much the same size and (2) graduating students have learned about the same amount. In so many other fields, the addition of mechanical or computational support has made the people who are performing/guiding the process much more productive. Are there any forms of automation that would be effective in increasing teacher productivity?
I'm not sure that it's within reach of the casual experimenter, but many years ago the double-slit-with-electrons experimental results caused me to change my major from physics on the grounds of "If this is the way the universe works, I don't want to know any more."
I have to disagree about the spec being "supplemental" to the code, at least depending on what the OV goal is. If the goal is to have OV code included in any app that uses the format (much like any proprietary format, except the OV people give you source instead of binary), then it's okay to short-change the spec since there's only going to be the one goup of people doing implementations. If the goal is to push the format and have many implementations, then the OV code is a useful tool, but the spec is the actual product and getting that out should take precedence at some point in time, as the spec will be important in establishing who's right/wrong when different implementations fail to interoperate properly.
I feel bad about defending MS, but I don't think the situation is as bad as we all initially think when Bill says that they can't easily pull IE out. I suspect that, at least in the NT-derived systems, stuff is probably quite modular. Speculating:
There's some chunk of HTTP transfer and HTML rendering code in the OS. It provides an API that a program can use to access the functionality. The API is probably pretty extensive, since it has to provide a means for setting security policies, etc. Given this API, writing a browser is pretty easy: you just need to write the code for things like history and options and such. Writing an HTML-based help system is also easy: just write a different and simpler UI around the rendering code and use local files for the content.
But if you pull out that nice modular HTTP/HTML component, lots of things get broken. In fact, everything that uses the API provided by the component is automatically broken. Putting in some other browser (say, Opera) won't fix those other apps unless the browser implements the same API for use by other programs. That may be hard, depending on how the underlying communication mechanism for the API is implemented.
In newer versions of the OS, MS has repeated this trick with media playback. The intended result is that it's easy for a developer to add audio and video playback to their application. The side effect is that stand-alone media players won't effectively replace the MS component unless they exactly duplicate that API.
At some level, this is not much different than the UNIX/X situation. If you remove X, all of the apps that depend on it are broken. If you put in a different windowing system that doesn't implement the X "API" correctly, the X apps are still broken. Of course, the X API is documented in detail, so anyone who wants to can take a shot at implementing it.
There have been a variety of economics articles published over the last few years (sorry, no links readily at hand) that attempt to make the point that the undeveloped portions of the world are becoming more and more angry at the developed world as global satellite television distribution makes it easy for them to see directly, on a regular basis, just how much richer the developed nations are. And not just richer in a direct material sense, but the richness of opportunity that's available to individuals.
And, of course, there is serious doubt as to whether the world has sufficient cheap energy and other raw materials to ever raise the undeveloped countries up to the current level of the United States, Canada, or western Europe.
The content people are in a technology war that they cannot win. They depend on consumer electronics that have to be cheap, require stable standards and have relatively long lifetimes. The technology people have general-purpose expensive devices that double in processing power every 18 months. Even when the protection depends on secret knowledge, it eventually leaks out (eg, Xing's player key stored in the clear) or is brute-force broken.
They are also fighting a legal war (eg, DMCA) that I also think they will eventually lose. The high-tech industry spans most of the world, and it seems unlikely that the content people can get reverse-engineering banned everywhere. As long as software capable of making a copy is available, everyone who really wants it will be able to get a copy. At some point, when the studios try to put a million college kids in jail over copying (some fair use, some not), they'll lose the legal war.
But they could win an economic war almost without trying. If DVDs and CDs were properly priced (and there's some evidence that even now they are priced too high to maximize profits), it's not worth the time and hassle to make a copy instead of buying an original. One exception might be those college students -- cash flow is a problem, and many have computer equipment completely out of line with their income. It's been a long time, but once I got out of school, the hassles of making a copy of an album were just not worth it.
Does it bother anyone but me that the electronics appears to be powered off a wall transformer? So that when commercial power fails, the phone service goes out as well? Providing network-based power for this type of device on a large scale that could allow phone service to continue when commercial power fails is EXPENSIVE! Not to mention that you would have to run copper of some sort in parallel with the fiber in order to carry that power.
Yes, it's expensive, but I don't think it's quite that expensive. Installed plant costs for hybrid-fiber-coax plant by a cable company currently run about $40,000 per mile including power, amplifiers, taps, etc.
Back in the late 70's, when AT&T was the One Bell System, an internal estimate of the cost to install just the fiber for fiber-to-the-home for the entire system was about $300B. Fiber is cheaper now, but labor is more expensive, and the process is labor-intensive. And a dollar doesn't go as far as it used to.
Many of the "households" served are going to be in high-density housing, ie, apartments and condos. The costs for doing wiring (fibering?) in those can range from almost nothing (vertical risers with lots of empty space) to a ridiculous amount per foot (eg, ripping open firewalls to run the fiber and then restoring things to meet city code).
Still, I'd bet that the job could be done for less than $10T if universal fiber was mandated:^) For me, a more realistic concern is that the number of people qualified to install fiber is rather limited, and getting the job done would take a LONG time even if you had the money.
That excluding the occasional mystic references that didn't have a whole lot to do with the overall plot, the original Star Wars was a Western. Young man on the frontier, parents slaughtered by the bad guy's gang, learns to be a gun-fighter from the grizzled, philisophical old man, teams up with a callous drifter (with a heart of gold in the long term), uses new-found skills to beat the bad guys, gang leader escapes just in case there's enough money to make a sequel, townspeople have big celebration and make him the marshall.
I agree, the XBox is another MS attempt to get a platform for MS applications and services into the family room. Previous attempts have included the purchase of WebTV, the development of UltimateTV, and their ongoing forays into cable set top box software. Sales of the first two were pretty pathetic and at least so far, they have not shown much ability to actually deliver working STB software.
It's all about sales growth for MS. PC sales are flattening out, a trend that is forecast to continue, since most homes that are interested in having a PC (and can afford one) have one and businesses are slowing their replacement cycles, and that slows MS's software sales. In order to grow, MS has to find a new platform. The home TV is a fairly obvious target if they could just get consumers to attach a "computer" to it. Unsurprisingly, they are also targeting most anything with a screen: hand-helds, cell-phones, etc.
... but if the government has a warrant there might not be much that can be done.
The ruling specifically states that the issues in this case are serious enough that before such a warrant is granted, the innocent third-party bookstore is entitled to an adversarial hearing, where they are allowed to oppose granting the warrant in order to protect their first amendment rights, and those of their customers. Given such a clear requirement by the state supreme court, there are probably not many judges here willing to ignore it. Obtaining a warrant like the one in this case just got a lot harder, at least in Colorado.
Not in our case, we threw out the database schema,...
It has long been my belief that writing correct code on a large scale is more about data than about the code itself. New features and such generally involve adding new data to the existing structure, or adding new meaning to existing items within the overall structure. As this happens, the rules that the data have to satisfy to be "consistent" become more and more complex. Formerly simple code like
if (cond1) {}
gets turned into
if (cond1 or cond2 or cond3 and !cond4) {}
At some point the right choice is the one you made-- start over with the basic questions of what are the data and how should they be organized?
Slashdot Mirror
If the studios will save tons of money on distribution costs, then it would seem reasonable for them to contribute in some fashion (up front cash? lower fees to the theaters?) to the large costs of the digital projectors.
I agree with the basic sentiment-- like any other language, you learn to write code by writing. The missing piece in your loop, particularly for beginners, is the editor who marks up their attempts, who tells them "this part is good" and "that part sucks". I was a self-taught programmer over 30 years ago and in that first six unsupervised months, I picked up bad habits that took me years to break. If you're just starting, get someone (hopefully someone who knows what they're doing) to edit some of your code.
Once you reach a certain level of proficiency, you can also progress by reading other people's code. Almost every language has some standard "idioms" for expressing common ideas well, and you can learn a lot of that from other people's code.
Is the "national provider" a regulated monopoly of some sort in Canada? The economics for capital recovery are a lot different when you don't have to worry about a competitor entering the market...
Since you mention capping and volume limits, I'll assume you're talking about the last-mile provider. The cable and cable-modem business is the one I know the most about, so I'll do some back-of-the-envelope calculations for that. Assume for the moment that the data business has to bear all of the costs.
The first major component of the monthly bill is the cost to construct the network. Hybrid-fiber-coax, the architecture required to operate two-way services, costs about $40,000 per mile for materials and installation. A mile of plant passes surprisingly few houses on average-- call it 100 for ease of calculation. Assume 20% of houses passed subscribe to the data service, and the construction cost per subscriber is $2,000. If the company wants to get its construction costs back in five years (we'll ignore issues of interest and risk for the moment), the monthly revenue needed is $33.33. There's some head-end equipment (a cable-modem termination system costs about $30,000) but those are spread over a lot more subscribers. Call it $35/month to pay for the network itself.
Now consider other recurring monthly expenses. There's the billing system that generates monthly bills for millions of subscribers. There's the customer-care systems. There's the salaries for the people who maintain the equipment and answer the phones. There's the rent/electric/sewer/etc for the space where those people work. There are indirect costs associated with those employees-- if you have 5,000 employees who operate the network and take care of the customers, you need a personnel office, a finance office, etc. This type of cost can easily run to $20/month per subscriber.
You need connections to the larger Internet. An OC-12 (600 Mbps) connected to somebody's backbone costs about $120,000/month, and you need one of those for approximately every 60,000 subscribers. That's another $2/month per subscriber. There are a bunch of "little" costs like that, let's guess that they add up to $5/month per subscriber.
When you set out to do this on a large scale and in a short period of time, you have to borrow an enormous amount of money. The people who loan it to you want interest. Comcast is buying AT&T Broadband, with 16M subscribers (counting video and telephony subs as well as cable-modem subs) and assuming $20B in debt. With an average interest rate of 7.5%, the monthly interest payment works out to around $7.80/month per subscriber. Add all that up and you have a total of about $67.80/month.
To be honest, not all of the network construction costs should be charged to the data service, and the billing system and personnel are spread across other services as well. Even so, an allocated cost of $40/month for data service is not a bad estimate, and that doesn't include anything for profit.
My own opinion is that the constitutional grants of a temporary monopoly on intellectual property recognize this situation as well as recognizing the ease with which IP can be copied.
I believe that the case law establishing that corporations are "persons" goes back to the 1860s when certain railroads were allowed to receive land grants which Congress had promised to "persons" satisfying certain requirements about using the property.
I do agree with your general sentiment that it is easy for these very rich, very long-lived "persons" to abuse laws intended for mere humans.
I recall reading somewhere that the cost of a state-of-the-art fab line was doubling roughly every 36 months (approximately half the Moore's Law rate). The article predicted that about 2010, the fixed costs for such a line would make chips produced on it so expensive that there would be very little market for them. Can anyone with actual experience in the field comment on how that cost prediction is holding up?
Working in a technical field, the half-life on knowledge is about five years-- that is, five years from now, half of your technical knowledge will be obsolete. You'll have to learn new stuff to replace it. One of the side effects of a decent college program is that you will learn a lot about the process of learning, which will serve you well in a field where you're doomed (or blessed, depending on your perspective) to be a part-time student forever. Few high-school graduates have acquired the needed skills for that kind of ongoing study.
As a convicted monopolist, MS would be well served by instituting the same type of training for their staff. It will save them an enormous amount of grief in the long run.
I saw a demo system involving a mechanical work stand that held and moved a large assembly under voice control. The human attaching various components and wiring controlled the positioning with simple commands: higher, lower, rotate right, etc. Since they didn't have to put down the components or their tools to adjust the position, the work went much faster.
Lots of the new LED replacement bulbs for traffic lights have been going in here in Colorado. One unanticipated (I hope) side effect: when it snows and the wind is blowing hard, the sunscreens over the lights that face into the wind eventually accumulate enough snow to obscure the light. Fortunately, the LED bulbs are bright enough you can still tell which light is on. This doesn't seem to happen with the old incandescent lamps, I assume because they produce enough waste heat to melt the snow and keep it from building up.
The page includes the statement that a golf-ball-sized device could have the light output equivalent to 10,000 watts of incandescent bulbs (100 100-watt bulbs). Granted that they are more efficient than incandescent, how hot is such a device likely to get during operation? Even at 90% efficiency, a device that size converting 1,000 watts to heat is going to be damned hot.
I feel badly for teachers, who seem to be caught in the economic trap of a job whose productivity cannot be significantly increased. Compared to thirty years ago, (1) individual classes are pretty much the same size and (2) graduating students have learned about the same amount. In so many other fields, the addition of mechanical or computational support has made the people who are performing/guiding the process much more productive. Are there any forms of automation that would be effective in increasing teacher productivity?
I'm not sure that it's within reach of the casual experimenter, but many years ago the double-slit-with-electrons experimental results caused me to change my major from physics on the grounds of "If this is the way the universe works, I don't want to know any more."
I have to disagree about the spec being "supplemental" to the code, at least depending on what the OV goal is. If the goal is to have OV code included in any app that uses the format (much like any proprietary format, except the OV people give you source instead of binary), then it's okay to short-change the spec since there's only going to be the one goup of people doing implementations. If the goal is to push the format and have many implementations, then the OV code is a useful tool, but the spec is the actual product and getting that out should take precedence at some point in time, as the spec will be important in establishing who's right/wrong when different implementations fail to interoperate properly.
I feel bad about defending MS, but I don't think the situation is as bad as we all initially think when Bill says that they can't easily pull IE out. I suspect that, at least in the NT-derived systems, stuff is probably quite modular. Speculating:
There's some chunk of HTTP transfer and HTML rendering code in the OS. It provides an API that a program can use to access the functionality. The API is probably pretty extensive, since it has to provide a means for setting security policies, etc. Given this API, writing a browser is pretty easy: you just need to write the code for things like history and options and such. Writing an HTML-based help system is also easy: just write a different and simpler UI around the rendering code and use local files for the content.
But if you pull out that nice modular HTTP/HTML component, lots of things get broken. In fact, everything that uses the API provided by the component is automatically broken. Putting in some other browser (say, Opera) won't fix those other apps unless the browser implements the same API for use by other programs. That may be hard, depending on how the underlying communication mechanism for the API is implemented.
In newer versions of the OS, MS has repeated this trick with media playback. The intended result is that it's easy for a developer to add audio and video playback to their application. The side effect is that stand-alone media players won't effectively replace the MS component unless they exactly duplicate that API.
At some level, this is not much different than the UNIX/X situation. If you remove X, all of the apps that depend on it are broken. If you put in a different windowing system that doesn't implement the X "API" correctly, the X apps are still broken. Of course, the X API is documented in detail, so anyone who wants to can take a shot at implementing it.
There have been a variety of economics articles published over the last few years (sorry, no links readily at hand) that attempt to make the point that the undeveloped portions of the world are becoming more and more angry at the developed world as global satellite television distribution makes it easy for them to see directly, on a regular basis, just how much richer the developed nations are. And not just richer in a direct material sense, but the richness of opportunity that's available to individuals.
And, of course, there is serious doubt as to whether the world has sufficient cheap energy and other raw materials to ever raise the undeveloped countries up to the current level of the United States, Canada, or western Europe.
The content people are in a technology war that they cannot win. They depend on consumer electronics that have to be cheap, require stable standards and have relatively long lifetimes. The technology people have general-purpose expensive devices that double in processing power every 18 months. Even when the protection depends on secret knowledge, it eventually leaks out (eg, Xing's player key stored in the clear) or is brute-force broken.
They are also fighting a legal war (eg, DMCA) that I also think they will eventually lose. The high-tech industry spans most of the world, and it seems unlikely that the content people can get reverse-engineering banned everywhere. As long as software capable of making a copy is available, everyone who really wants it will be able to get a copy. At some point, when the studios try to put a million college kids in jail over copying (some fair use, some not), they'll lose the legal war.
But they could win an economic war almost without trying. If DVDs and CDs were properly priced (and there's some evidence that even now they are priced too high to maximize profits), it's not worth the time and hassle to make a copy instead of buying an original. One exception might be those college students -- cash flow is a problem, and many have computer equipment completely out of line with their income. It's been a long time, but once I got out of school, the hassles of making a copy of an album were just not worth it.
Does it bother anyone but me that the electronics appears to be powered off a wall transformer? So that when commercial power fails, the phone service goes out as well? Providing network-based power for this type of device on a large scale that could allow phone service to continue when commercial power fails is EXPENSIVE! Not to mention that you would have to run copper of some sort in parallel with the fiber in order to carry that power.
Back in the late 70's, when AT&T was the One Bell System, an internal estimate of the cost to install just the fiber for fiber-to-the-home for the entire system was about $300B. Fiber is cheaper now, but labor is more expensive, and the process is labor-intensive. And a dollar doesn't go as far as it used to.
Many of the "households" served are going to be in high-density housing, ie, apartments and condos. The costs for doing wiring (fibering?) in those can range from almost nothing (vertical risers with lots of empty space) to a ridiculous amount per foot (eg, ripping open firewalls to run the fiber and then restoring things to meet city code).
Still, I'd bet that the job could be done for less than $10T if universal fiber was mandated :^) For me, a more realistic concern is that the number of people qualified to install fiber is rather limited, and getting the job done would take a LONG time even if you had the money.
That excluding the occasional mystic references that didn't have a whole lot to do with the overall plot, the original Star Wars was a Western. Young man on the frontier, parents slaughtered by the bad guy's gang, learns to be a gun-fighter from the grizzled, philisophical old man, teams up with a callous drifter (with a heart of gold in the long term), uses new-found skills to beat the bad guys, gang leader escapes just in case there's enough money to make a sequel, townspeople have big celebration and make him the marshall.
It's all about sales growth for MS. PC sales are flattening out, a trend that is forecast to continue, since most homes that are interested in having a PC (and can afford one) have one and businesses are slowing their replacement cycles, and that slows MS's software sales. In order to grow, MS has to find a new platform. The home TV is a fairly obvious target if they could just get consumers to attach a "computer" to it. Unsurprisingly, they are also targeting most anything with a screen: hand-helds, cell-phones, etc.