One of the factors that limits performance in athletic contests such as the recently completed Tour de France is cardiac output -- how much blood the rider's heart can pump.
Fifty years from now will we have to have rules against riders with artificial hearts because they have an unfair advantage in cardiac capacity?
Or will we borrow the kinds of regulations that the various auto racing organizations impose on engines?
You can have an articial heart, but volume pumped must be constrained to be below X liters per minute, or the outlet into the aorta must be less than some number of square millimeters?
DES hardware exists, and is inexpensive and relatively secure. Using current hardware to impliment triple DES is easy.
Indeed.
It is one thing for NIST to recommend that everyone using software implementations of DES should change to something else
(although it appears that they are actually only recommending it to government users).
It is a very different thing to deal with the millions of consumer devices out there with hardware DES which would have to be replaced.
Scaled is a group of professional engineers, working on a budget with a benefactor.
Assume that SS1 can be extended in ways that make actual revenue-generating services possible -- and some serious extensions would seem to be necessary to make that happen.
For example, assume they can deliver, on extremely short notice, a person or similar amount of mass to anywhere on the planet via suborbital.
I, for one, would want the vessel on which I'm putting my company's chief engineer for delivery to Moscow to fix the broken whatever, to be designed, built, maintained and operated by paid professional engineers.
I'm sure that the insurance companies (losing my chief engineer is something I want to be insured against) will feel the same way.
Unless he/she has really strong feelings about amateur space (or near space) travel, I suspect that the chief engineer will also feel that way.
I think what Scaled Composites has done for $20M is really cool.
But producing a commercial vehicle will cost them at least one order of magnitude more money than SS1 has cost, very probably two orders of magnitude, and possibly three orders.
To which the MS guys, with a serious voice and without a moment's hesitation, only replied, "this we will never do"
Of course not.
Basic company policy is still set by Bill Gates, who wrote the
Open Letter to Hobbyists in 1976.
That letter set the direction for Microsoft -- sell lots of copies of proprietary software -- that made him the richest man in the world.
Any change in that direction will have to come from Bill himself, and the odds of that seem to range from slim to none.
When the Great Patent War commences next year, it won't be about getting checks... they want the products dead and customers scared off.
They don't even need to scare the customers in order to kill the product.
It's as simple as saying "You don't have permission to use my patent -- cease and desist immediately."
Unless the patent is trivially invalid, the courts will tend to side with the patent holder and issue the necessary orders to make the infringer halt while the case is decided.
There are no requirements that patents be licensed in any way that makes sense.
There's no equivalent to fair use.
Reverse engineering is illegal (and unnecessary in the case of a reasonable patent application, since the applicant must reveal all the secret stuff in order to get a patent).
To choose an example, Microsoft holds at least one patent that covers reading.wmv files.
ANY software without a license from MS that reads such a file on a computer in the US infringes.
And yes, MS has threatened at least one non-MS project with legal action for writing such a reader.
Using such software (even if you wrote your own implementation and are doing so in private) is infringing.
Downloading it from overseas (from somewhere that does not allow software patents, say) is almost certainly an illegal "importing" of the infringing product.
Those guys burn up to 12.000calories in one day (insert lame joke here)
A lot of people don't appreciate how big a role nutrition plays in Tour performance.
The typical rider consumes about 7,000 calories per day; almost all of them lose weight over the course of the race.
About 70% of the body's energy production uses carbohydrates; so they need about 5,000 carb calories per day; your muscles and liver can store about 2,000.
So lots of the calories have to be ingested while riding.
A typical rider will sweat about three gallons during a stage, and almost all of that needs to be replaced.
Lots of drinking while riding as well.
Small errors in the amount you eat and drink and what you eat and drink can make a very large difference at this level of performance.
Not to bring up the elephant standing in the room or anything, but we _are_ occupying a (now) hostile foreign country
Indeed.
NASA's budget is on the order of $18B per year.
In calendar 2003 and 2004, the US will spend on the near order of $180B on operations in Iraq.
In all liklihood, at least another $100B will be spent there before the troops leave.
Roughly 15 years of NASA budget spent on a project that will, in all liklihood, fail in all of its goals except removing a particular despot from power.
I will cheerfully bet a beer that ten years from now Iraq is once again being run by a despot, or a small group of them.
While I would love to rejoice as much as the next Slashdotter, isn't anyone worried about appeals?
To be perfectly honest, no.
You can't appeal just because you didn't like the outcome.
You have to argue that the judge made an error in procedure, or an error in law, or (much harder) an error in the facts.
You have to lay the groundwork for your appeal in the original case.
If SCO could show precedents in case law that making demands beyond the (apparently) clear language in the contract was supported by the law, they needed to show those precedents before now.
The mainstream press is buying into SCO's claims just (AFAICT) based on the weight of how often they repeat them and the fact that they have an easy contact point, whereas there is no general "Linux" contact person.
Not to mention the fact that the other parties in the various lawsuits, being grownups, don't make comments about ongoing court cases.
After spending almost 25 years inside giant corporations, I did learn that the two rules are (1) no one but the lawyers is allowed to talk about the case and (2) the lawyers don't say anything.
Most public companies take their profits and reinvest them in the company.
Historically, MOST large companies (eg, those in the S&P 500) regardless of industry pay a dividend on the order of 4% per year.
Earlier in the 20th century, dividends averaged as high as 6-7%.
Recently the average has dropped to 2-3%, which is a historical low.
The tech companies that have gotten so big so fast in the last 20 years and still pay a tiny dividend are an exception, not the rule.
Those companies have now reached a scale where it is unrealistic to expect them to be high-growth businesses (on a percentage basis), and are struggling to adjust to the fact that they are now "mature" firms.
Microsoft seems to be making the adjustment somewhat more quickly than the other new tech giants.
Realizing that rapid share price appreciation was probably gone for good, they quit issuing options to the employees and now give limited stock grants.
Realizing that they have pissed off a lot of shareholders by accumulating $60B in cash that they can't seem to use, they are issuing the one-time special dividend and increasing their annual dividend (although it will still be on the order of only 1%).
I suspect that the stock buy-back is aimed more at trying to increase the share price for those thousands of disgruntled employees holding worthless options than anything else -- buying shares is no better "investment" than using the money internally, something they don't seem to be able to do, but it does have financial effects.
I love programming, however I work in a network operations center... what do all of you programmers do to avoid workplace monotony?
Your question here seems to be very poorly structured, and the range of answers people are giving reflect that.
Many of the suggestions (read, watch DVDs) assume that the problem is that too much of the time the job is merely to be physically present.
If what you need is an answer to how to fill up the idle hours, and you love to program, I'd suggest that no NOC I've ever seen has all the tools that it really needs.
Consider what kinds of tools would make your job easier when problems occur (or tools that can analysis the available data and identify potential problems before they occur) and write them.
If your problem is that you're already programming and you've got plenty to do, but there's no "human contact" in the way the job is done, try to add some.
It might be as simple as adding informal design reviews -- "Hey, Bob, can we get together for 30 minutes on Tuesday so I can describe how I'm structuring this, and get your opinion about it?"
When I had programming tasks, I always found that having such reviews sometimes led to people pointing out better ways to do a task, and always clarified my own thinking about what I was doing.
None of these are sports, even if someone did attach some variation of "olympiad" to them.
Let me put it this way -- could you hold an annual competition for the "best" mathematician (or biologist, or physicist) in the world each year?
Not the best at taking a multiple choice biology exam, or at solving differential equations in their head, but the best at creating new knowledge in the selected field?
Various organizations give awards for the "best" paper in different fields each year, but any one of those may be the culmunation of years of work and experimentation.
But the thing that makes people great at such fields are those elusive "Aha!" moments in the middle of the night, or in the shower, or whenever, when a collection of pieces fall into place and form a pattern that no one has ever seen before.
I can just imagine the announcers at the typical annual world math competition in, say, the topological manifolds event.
"And there's the final buzzer, Bob, and as usual -- NOTHING HAPPENED!
The greatest topologists in the world went at it for 60 minutes and NONE OF THEM HAD A SINGLE INSIGHT!"
I appreciate that high bandwidth is becoming available at a pretty good clip; I enjoy a 3Mbps connection from Comcast right now. But I question whether they have what it takes to deliver a sustained 3Mbps for two hours now.
It will also depend on whether the streamed video source is on Comcast's local (typically metropolitan area) network or not.
For cable companies, connections to the backbone are usually sized at roughly 20-50 Kbps (yes, kilobits) per subscriber -- if they have 50,000 high-speed data subs, the backbone connection will be between 1 and 2.5 Gbps.
At 3 Mbps per stream, 1 Gbps is sufficient to support 333 simultaneous users -- less than 1% of the high-speed data subs.
Some forecasts for conventional video-on-demand have peak-load estimates (simultaneous use) as high as 25% of digital video subscribers.
If VOD over cable modem achieved even a fraction of that level, the data service would go to hell quickly.
Long before the problem reaches that point, expect to see the cable companies impose monthly limits on bit volumes received, with hefty fees added to the bill for repeated violations.
Depending on the local cable plant, that level of conventional VOD may or may not be a problem.
The company I worked for (before being bought out by AT&T Broadband, and in turn by Comcast) tried to design the network with 700 MHz of downstream bandwidth shared across 500 households.
Whether that was sufficient to support lots of VOD depended in part on how many analog channels we were required to deliver (federal regulations currently require local channels to be delivered in analog; many cable networks have long-term contracts requiring analog delivery).
Under reasonable assumptions about moving cable to digital, and a gradual increase in the VOD peak rate, it looked like enough bandwidth.
The most difficult part was the "middle" portion of the network, where the VOD streams for many groups of households were aggregated.
But he does have a point. Most of the effort that's gone into hardware and software development, has been aimed at doing the same things faster. Real innovation is very rare. Our desktops still are essentially the same as the 1984 Macintosh.
And the 1984 Macintosh desktop was a representation of the physical desktop that had been in common usage for the past 200 years or more.
Real innovation in what people do with their time is very rare.
Alan seems to believe that personal computers should have drastically changed how people spend their time, or what their daily activities are.
What do "typical" grown-up people do at home (ignoring business uses of personal computing)?
Cooking, cleaning, sex and other entertainment, interact with their kids, take care of the personal paperwork, take care of the house (inside and outside), etc.
For most, things like education are VERY far down on the list.
The average adult in the US DIDN'T READ A BOOK LAST YEAR.
Alan's not the first one to make this mistake -- many of the early advocates of television believed that people would make heavy use of it for educational purposes as well.
Indeed.
Even the casual rider will notice a large difference over the course of several kilometers between two essentially identical bikes, one properly fitted and the other not.
The current rules do "standardize" enough -- double triangle geometry, minimum weight -- to avoid the worst of the America's Cup fiascos.
The kinds of changes that are being made in wheels and such these days seem to be second- or third-order effects:
as another poster pointed out, wind tunnel tests show that one of the largest sources of drag in the current bike/rider configuration is the paper rider identification numbers each rider must wear.
I'm so sick of CEOs complaining that there are not enough engineers being educated in this country...There's a reason that nobody is getting these dgrees - no jobs!
Indeed.
The first follow-up question any reporter should ask when interviewing Andy Grove (for example), and he goes off on his rant that we're not educating enough scientists, should be "How many net PhDs did Intel add to their staff in the past year? How many Masters-level engineers? How do their salaries stack up?"
I have a friend who manages a research group at Intel Labs -- they have been forced to reduce their headcount for each of the last few years.
If companies like Intel are shrinking their research staff, then we've presumably already got too many scientists.
Or do you mean they're unable to navigate the absurd sociopolitical burearacracies found in any large company?
Speaking from some experience in both worlds (more the corporate, but some academic), there is at least as much vicious infighting in the typical university department as there is in any large corporation.
If you can (a) manage an advisor and dissertation committee well enough to graduate, (b) survive the publication battles, (c) successfully obtain grant monies, and (d) reach the holy grail of tenure, you have all the people management skills you need to for life in a big corporation.
OTOH, if you can't manage b, c, and d, and are determined to stay in academia, you're probably doomed to a life of poorly-paid post-doc and instructor positions.
Those generalities allow you to recognize certain patterns and also reject other patterns
Don't the top-notch computer programs already do this in one form or another?
I mean, it seems like such an obvious line of research for pattern matching and pattern recognition people to explore.
Or perhaps not -- that may be one of the reasons that computers aren't as good at Go as they are at chess, Go appearing (to a rank beginner) to depend more on pattern recognition and less on straight-forward deductive analysis.
Congress has talked about doing away with bundling, letting subscribers pick and choose channels. If that happens, watch the crud channels die away as no one subscribes to them - accentuating this apparent trend of fewer channels... Something else that I find ironic is such a scheme would promote a free market in cable channels
If Congress is going to do this, they must also address the issue of bundling when channels are sold to the cable/satellite operator.
So long as ESPN (just to use as an example) is allowed to require the operators to purchase the entire catalog of ESPN channels in order to get ESPN proper, it is silly to require that the operator unbundle that group.
Major changes in the structure of "wholesale" pricing will also be needed.
ESPN currently charges the operators on the basis of the total number of subscribers; if that charge is $2.00 per sub per month, but only 50% of subs take ESPN under a la carte, the operator must charge $4.00 to those subs in order to break even.
And it seems likely that ESPN would raise its rates to the operators even faster than it has in the past, since it will have to offset losses in advertising revenue: current rates are set on the assumption that ESPN is available in all households that get more than the minimal service, an assumption that is not true in an a la carte world.
In an ideal world, the cable/satellite companies would be enablers, who charge both the subscribers and the networks for access to delivery pipe, and subscribers would buy content from those networks that they were interested in.
We are so far from that model today that I worry that tinkering with any one part of the system runs the risk of making things worse than they are.
Considering that nearly all GPL/Free/OpenSource software says that "THERE IS NO WARRANTY" (etc, etc), your claim is without merit.
Many states have implied warranty laws;
basically these say that buyers of the things that you sell can expect some level of functionality based on your representation of the thing, and that simply plastering "No Warranty" on it doesn't get you off the hook.
IIRC, mostly these laws entitle you to get your money back;
I'm less sure about whether you can make claims for damages.
In at least some cases, if the product is given away, there's no recourse;
money has to change hands for the implied warranty to come into play.
It took a long time before implied warranty laws came into existance, and not every place has them today.
It will take a long time for legislatures to come to conclusions that software is a product for which there should be implied warranties, and how extensive those warranties should be.
If I sell you a C compiler, you have a reasonable expectation that it will take a source file and produce an executable image;
would an implied warranty law require that it produce correct code under all circumstances?
I'm not sure that there are any C compilers around that don't generate incorrect code under some really odd conditions.
Also, consider that many of these options will go unused, either because the employee leaves the company before they vest or because the options are underwater when they expire.
In the next few years enormous numbers of options are going to expire unexercised.
The enormous runup in share prices for tech and telecom companies in the late 1990s created the situation where companies that were far too large to be considered growth companies (two companies have revenues of $50M and $50B respectively; which one will be able to realistically double its revenues?) to behave like "little" companies.
They could pitch options to employees as "almost" guaranteed compensation.
Semi-crooked corporate officers could grant themselves enormous paychecks that didn't show up except in footnotes.
Outside of small companies, where options are much more legitimate (take a piece of the company INSTEAD of a paycheck, not in ADDITION to one) and probably shouldn't be expensed, the bubble deflation has made the whole option issue much less important.
I find it telling that Microsoft no longer grants options; I expect other large companies to eventually follow that lead.
The most likely reason for a lack of free software for this application is it's not common for someone to have experience programming and this sort of thing.
I would modify this statement to say that it is not common for someone to be (a) an experienced programmer, (b) working in the investment field, and (c) not tied up with a company that insists on keeping the development work private.
There are lots of firms doing lots of work to try and identify profitable patterns in stock, bond and index prices (ETFs effectively falling into the last category).
Read Thomas Bass' The Predictors for an example of the struggles to build a business around such work, and the level of high-powered mathematicians and programmers that have been involved in the past.
I would go even further and state that there are lots of hobbyist efforts in this field, but the successful hobbyists are also motivated to keep their work secret -- they did not do the work in order to demonstrate that it could be done, they did it to make money.
If you give away your code, others can use it to discover the same patterns;
if enough people do that, the pattern will eventually be traded out of profitability.
An afternoon with Google will turn up an enormous number of software products employing a wide range of techniques to search for profitable patterns: genetic algorithms, neural networks, etc.
Ditto for underlying structures whose parameter spaces can be searched: Elliot waves, Bollinger bands, every kind of combination of moving averages you can imagine, fractals, etc.
Not to be discouraging, but going over the same search grounds that have been worked to death is not the way to succeed in the original goal (find profitable patterns).
Productive results are much more likely to come from identifying new types of patterns, which almost guarantees that they are not discoverable by the existing set of tools.
The widget is not fully working.
It has bugs.
But, in relation to milestone n, it has zero defects, i.e., it passes all of the tests for milestone n.
Of course, we all know how difficult it is to write a complete set of tests.
Particularly if all error cases must be included.
The last time I had that particular pleasure was for a relatively simple (compared to an HTML widget) piece of real-time software.
The documentation ended up going on for dozens of pages about the conditions that the internal data structures and a received message must satisfy in order to be consistent (inconsistent messages were discarded and reported).
The section specifying what the component actually did when it received a consistent message was much smaller.
Slashdot Mirror
One of the factors that limits performance in athletic contests such as the recently completed Tour de France is cardiac output -- how much blood the rider's heart can pump. Fifty years from now will we have to have rules against riders with artificial hearts because they have an unfair advantage in cardiac capacity? Or will we borrow the kinds of regulations that the various auto racing organizations impose on engines? You can have an articial heart, but volume pumped must be constrained to be below X liters per minute, or the outlet into the aorta must be less than some number of square millimeters?
Indeed. It is one thing for NIST to recommend that everyone using software implementations of DES should change to something else (although it appears that they are actually only recommending it to government users). It is a very different thing to deal with the millions of consumer devices out there with hardware DES which would have to be replaced.
Assume that SS1 can be extended in ways that make actual revenue-generating services possible -- and some serious extensions would seem to be necessary to make that happen. For example, assume they can deliver, on extremely short notice, a person or similar amount of mass to anywhere on the planet via suborbital. I, for one, would want the vessel on which I'm putting my company's chief engineer for delivery to Moscow to fix the broken whatever, to be designed, built, maintained and operated by paid professional engineers. I'm sure that the insurance companies (losing my chief engineer is something I want to be insured against) will feel the same way. Unless he/she has really strong feelings about amateur space (or near space) travel, I suspect that the chief engineer will also feel that way.
I think what Scaled Composites has done for $20M is really cool. But producing a commercial vehicle will cost them at least one order of magnitude more money than SS1 has cost, very probably two orders of magnitude, and possibly three orders.
Of course not. Basic company policy is still set by Bill Gates, who wrote the Open Letter to Hobbyists in 1976. That letter set the direction for Microsoft -- sell lots of copies of proprietary software -- that made him the richest man in the world. Any change in that direction will have to come from Bill himself, and the odds of that seem to range from slim to none.
They don't even need to scare the customers in order to kill the product. It's as simple as saying "You don't have permission to use my patent -- cease and desist immediately." Unless the patent is trivially invalid, the courts will tend to side with the patent holder and issue the necessary orders to make the infringer halt while the case is decided. There are no requirements that patents be licensed in any way that makes sense. There's no equivalent to fair use. Reverse engineering is illegal (and unnecessary in the case of a reasonable patent application, since the applicant must reveal all the secret stuff in order to get a patent).
To choose an example, Microsoft holds at least one patent that covers reading .wmv files.
ANY software without a license from MS that reads such a file on a computer in the US infringes.
And yes, MS has threatened at least one non-MS project with legal action for writing such a reader.
Using such software (even if you wrote your own implementation and are doing so in private) is infringing.
Downloading it from overseas (from somewhere that does not allow software patents, say) is almost certainly an illegal "importing" of the infringing product.
A lot of people don't appreciate how big a role nutrition plays in Tour performance. The typical rider consumes about 7,000 calories per day; almost all of them lose weight over the course of the race. About 70% of the body's energy production uses carbohydrates; so they need about 5,000 carb calories per day; your muscles and liver can store about 2,000. So lots of the calories have to be ingested while riding. A typical rider will sweat about three gallons during a stage, and almost all of that needs to be replaced. Lots of drinking while riding as well. Small errors in the amount you eat and drink and what you eat and drink can make a very large difference at this level of performance.
Indeed. NASA's budget is on the order of $18B per year. In calendar 2003 and 2004, the US will spend on the near order of $180B on operations in Iraq. In all liklihood, at least another $100B will be spent there before the troops leave. Roughly 15 years of NASA budget spent on a project that will, in all liklihood, fail in all of its goals except removing a particular despot from power. I will cheerfully bet a beer that ten years from now Iraq is once again being run by a despot, or a small group of them.
To be perfectly honest, no. You can't appeal just because you didn't like the outcome. You have to argue that the judge made an error in procedure, or an error in law, or (much harder) an error in the facts. You have to lay the groundwork for your appeal in the original case. If SCO could show precedents in case law that making demands beyond the (apparently) clear language in the contract was supported by the law, they needed to show those precedents before now.
Not to mention the fact that the other parties in the various lawsuits, being grownups, don't make comments about ongoing court cases. After spending almost 25 years inside giant corporations, I did learn that the two rules are (1) no one but the lawyers is allowed to talk about the case and (2) the lawyers don't say anything.
Historically, MOST large companies (eg, those in the S&P 500) regardless of industry pay a dividend on the order of 4% per year. Earlier in the 20th century, dividends averaged as high as 6-7%. Recently the average has dropped to 2-3%, which is a historical low. The tech companies that have gotten so big so fast in the last 20 years and still pay a tiny dividend are an exception, not the rule. Those companies have now reached a scale where it is unrealistic to expect them to be high-growth businesses (on a percentage basis), and are struggling to adjust to the fact that they are now "mature" firms.
Microsoft seems to be making the adjustment somewhat more quickly than the other new tech giants. Realizing that rapid share price appreciation was probably gone for good, they quit issuing options to the employees and now give limited stock grants. Realizing that they have pissed off a lot of shareholders by accumulating $60B in cash that they can't seem to use, they are issuing the one-time special dividend and increasing their annual dividend (although it will still be on the order of only 1%). I suspect that the stock buy-back is aimed more at trying to increase the share price for those thousands of disgruntled employees holding worthless options than anything else -- buying shares is no better "investment" than using the money internally, something they don't seem to be able to do, but it does have financial effects.
Your question here seems to be very poorly structured, and the range of answers people are giving reflect that. Many of the suggestions (read, watch DVDs) assume that the problem is that too much of the time the job is merely to be physically present. If what you need is an answer to how to fill up the idle hours, and you love to program, I'd suggest that no NOC I've ever seen has all the tools that it really needs. Consider what kinds of tools would make your job easier when problems occur (or tools that can analysis the available data and identify potential problems before they occur) and write them.
If your problem is that you're already programming and you've got plenty to do, but there's no "human contact" in the way the job is done, try to add some. It might be as simple as adding informal design reviews -- "Hey, Bob, can we get together for 30 minutes on Tuesday so I can describe how I'm structuring this, and get your opinion about it?" When I had programming tasks, I always found that having such reviews sometimes led to people pointing out better ways to do a task, and always clarified my own thinking about what I was doing.
I can just imagine the announcers at the typical annual world math competition in, say, the topological manifolds event. "And there's the final buzzer, Bob, and as usual -- NOTHING HAPPENED! The greatest topologists in the world went at it for 60 minutes and NONE OF THEM HAD A SINGLE INSIGHT!"
It will also depend on whether the streamed video source is on Comcast's local (typically metropolitan area) network or not. For cable companies, connections to the backbone are usually sized at roughly 20-50 Kbps (yes, kilobits) per subscriber -- if they have 50,000 high-speed data subs, the backbone connection will be between 1 and 2.5 Gbps. At 3 Mbps per stream, 1 Gbps is sufficient to support 333 simultaneous users -- less than 1% of the high-speed data subs. Some forecasts for conventional video-on-demand have peak-load estimates (simultaneous use) as high as 25% of digital video subscribers. If VOD over cable modem achieved even a fraction of that level, the data service would go to hell quickly. Long before the problem reaches that point, expect to see the cable companies impose monthly limits on bit volumes received, with hefty fees added to the bill for repeated violations.
Depending on the local cable plant, that level of conventional VOD may or may not be a problem. The company I worked for (before being bought out by AT&T Broadband, and in turn by Comcast) tried to design the network with 700 MHz of downstream bandwidth shared across 500 households. Whether that was sufficient to support lots of VOD depended in part on how many analog channels we were required to deliver (federal regulations currently require local channels to be delivered in analog; many cable networks have long-term contracts requiring analog delivery). Under reasonable assumptions about moving cable to digital, and a gradual increase in the VOD peak rate, it looked like enough bandwidth. The most difficult part was the "middle" portion of the network, where the VOD streams for many groups of households were aggregated.
And the 1984 Macintosh desktop was a representation of the physical desktop that had been in common usage for the past 200 years or more. Real innovation in what people do with their time is very rare. Alan seems to believe that personal computers should have drastically changed how people spend their time, or what their daily activities are. What do "typical" grown-up people do at home (ignoring business uses of personal computing)? Cooking, cleaning, sex and other entertainment, interact with their kids, take care of the personal paperwork, take care of the house (inside and outside), etc. For most, things like education are VERY far down on the list. The average adult in the US DIDN'T READ A BOOK LAST YEAR. Alan's not the first one to make this mistake -- many of the early advocates of television believed that people would make heavy use of it for educational purposes as well.
Indeed. Even the casual rider will notice a large difference over the course of several kilometers between two essentially identical bikes, one properly fitted and the other not. The current rules do "standardize" enough -- double triangle geometry, minimum weight -- to avoid the worst of the America's Cup fiascos. The kinds of changes that are being made in wheels and such these days seem to be second- or third-order effects: as another poster pointed out, wind tunnel tests show that one of the largest sources of drag in the current bike/rider configuration is the paper rider identification numbers each rider must wear.
When's the next America's Cup? Those guys spend insane sums on technology and testing...
Indeed. The first follow-up question any reporter should ask when interviewing Andy Grove (for example), and he goes off on his rant that we're not educating enough scientists, should be "How many net PhDs did Intel add to their staff in the past year? How many Masters-level engineers? How do their salaries stack up?" I have a friend who manages a research group at Intel Labs -- they have been forced to reduce their headcount for each of the last few years. If companies like Intel are shrinking their research staff, then we've presumably already got too many scientists.
Speaking from some experience in both worlds (more the corporate, but some academic), there is at least as much vicious infighting in the typical university department as there is in any large corporation. If you can (a) manage an advisor and dissertation committee well enough to graduate, (b) survive the publication battles, (c) successfully obtain grant monies, and (d) reach the holy grail of tenure, you have all the people management skills you need to for life in a big corporation. OTOH, if you can't manage b, c, and d, and are determined to stay in academia, you're probably doomed to a life of poorly-paid post-doc and instructor positions.
Don't the top-notch computer programs already do this in one form or another? I mean, it seems like such an obvious line of research for pattern matching and pattern recognition people to explore. Or perhaps not -- that may be one of the reasons that computers aren't as good at Go as they are at chess, Go appearing (to a rank beginner) to depend more on pattern recognition and less on straight-forward deductive analysis.
If Congress is going to do this, they must also address the issue of bundling when channels are sold to the cable/satellite operator. So long as ESPN (just to use as an example) is allowed to require the operators to purchase the entire catalog of ESPN channels in order to get ESPN proper, it is silly to require that the operator unbundle that group. Major changes in the structure of "wholesale" pricing will also be needed. ESPN currently charges the operators on the basis of the total number of subscribers; if that charge is $2.00 per sub per month, but only 50% of subs take ESPN under a la carte, the operator must charge $4.00 to those subs in order to break even. And it seems likely that ESPN would raise its rates to the operators even faster than it has in the past, since it will have to offset losses in advertising revenue: current rates are set on the assumption that ESPN is available in all households that get more than the minimal service, an assumption that is not true in an a la carte world.
In an ideal world, the cable/satellite companies would be enablers, who charge both the subscribers and the networks for access to delivery pipe, and subscribers would buy content from those networks that they were interested in. We are so far from that model today that I worry that tinkering with any one part of the system runs the risk of making things worse than they are.
Many states have implied warranty laws; basically these say that buyers of the things that you sell can expect some level of functionality based on your representation of the thing, and that simply plastering "No Warranty" on it doesn't get you off the hook. IIRC, mostly these laws entitle you to get your money back; I'm less sure about whether you can make claims for damages. In at least some cases, if the product is given away, there's no recourse; money has to change hands for the implied warranty to come into play.
It took a long time before implied warranty laws came into existance, and not every place has them today. It will take a long time for legislatures to come to conclusions that software is a product for which there should be implied warranties, and how extensive those warranties should be. If I sell you a C compiler, you have a reasonable expectation that it will take a source file and produce an executable image; would an implied warranty law require that it produce correct code under all circumstances? I'm not sure that there are any C compilers around that don't generate incorrect code under some really odd conditions.
In the next few years enormous numbers of options are going to expire unexercised. The enormous runup in share prices for tech and telecom companies in the late 1990s created the situation where companies that were far too large to be considered growth companies (two companies have revenues of $50M and $50B respectively; which one will be able to realistically double its revenues?) to behave like "little" companies. They could pitch options to employees as "almost" guaranteed compensation. Semi-crooked corporate officers could grant themselves enormous paychecks that didn't show up except in footnotes. Outside of small companies, where options are much more legitimate (take a piece of the company INSTEAD of a paycheck, not in ADDITION to one) and probably shouldn't be expensed, the bubble deflation has made the whole option issue much less important. I find it telling that Microsoft no longer grants options; I expect other large companies to eventually follow that lead.
I would modify this statement to say that it is not common for someone to be (a) an experienced programmer, (b) working in the investment field, and (c) not tied up with a company that insists on keeping the development work private. There are lots of firms doing lots of work to try and identify profitable patterns in stock, bond and index prices (ETFs effectively falling into the last category). Read Thomas Bass' The Predictors for an example of the struggles to build a business around such work, and the level of high-powered mathematicians and programmers that have been involved in the past. I would go even further and state that there are lots of hobbyist efforts in this field, but the successful hobbyists are also motivated to keep their work secret -- they did not do the work in order to demonstrate that it could be done, they did it to make money. If you give away your code, others can use it to discover the same patterns; if enough people do that, the pattern will eventually be traded out of profitability.
An afternoon with Google will turn up an enormous number of software products employing a wide range of techniques to search for profitable patterns: genetic algorithms, neural networks, etc. Ditto for underlying structures whose parameter spaces can be searched: Elliot waves, Bollinger bands, every kind of combination of moving averages you can imagine, fractals, etc. Not to be discouraging, but going over the same search grounds that have been worked to death is not the way to succeed in the original goal (find profitable patterns). Productive results are much more likely to come from identifying new types of patterns, which almost guarantees that they are not discoverable by the existing set of tools.
Of course, we all know how difficult it is to write a complete set of tests. Particularly if all error cases must be included. The last time I had that particular pleasure was for a relatively simple (compared to an HTML widget) piece of real-time software. The documentation ended up going on for dozens of pages about the conditions that the internal data structures and a received message must satisfy in order to be consistent (inconsistent messages were discarded and reported). The section specifying what the component actually did when it received a consistent message was much smaller.
No. Nor have I noticed an *actual* fuel cell in *any* laptop or other portable device that might be carried on an airplane :^)