... at how lazy American graduate students are. I returned to University after 7 years absence and that's been the most difficult adjustment. I've gotten used to those around me being willing to work 50 hour weeks every week. Graduate students complain if they have to work 20 hours. It's been great for me... all my professors think I'm much brighter than I really am. It's easy to fool them since I just treat my school work like a real job and try to spend at least 8 hours each day working at it.
There is one skill that I'd recommend you try to pick up immediately, though: RESEARCH. You need to learn two things. First, you need to learn how to find papers about what you're interested in. And, scholar.google.com isn't sufficient (yet). Second, you need to learn to actually read and understand those papers. Sometimes it can take a month just to get through a single paper. But, learn how to push through and you'll reap dividends. Some graduate departments do a good job of teaching research skills. Others are lousy. Your experience could differ alot. But, to be good in graduate school requires alot more than just attendance. So, learn to do good research before you're paying to do so and you'll be happier.
As far as admissions goes, I think you should call the department you want to study in and ask the graduate coordinator. Every department is different. Some Universities really cater to returning professional students -- George Mason University in Virginia is that way. Other Universities have different philosophies.
In my field (mathematics) most admissions decisions are ultimately made based upon academic recommendations. I think that GPA and test scores are important, too. My wife says that when she worked in the admissions office they used GPA and test scores just to sort the applications into "bother to read recommendations" and "don't bother" piles. YMMV.
I can't speak to the differences between US and UK schools. I've got a number of friends who went to school in Canada and I don't think our systems differ much. Except Canadian schools are better funded so they're cheaper.
If you're trying to decide where to apply I'd recommend that you start by reading conference papers and published research. Try to find some research topics that are of particular interest to you. Then, look at where there are academics who are currently working in that field. You may discover that the preeminent researcher in your interest is at a really cheap school! How cool would that be! If you have no idea what you want to do with your life, then try to find a department that's large enough to support lots of research interests. That way, you can used your Masters degree as an opportunity to sample everything and see what you like best.
If you're returning to school because you just want another degree (and eventually a better job) then you should consider trying to make sure you stay very well rounded. On the other hand, if you're trying to become an academic you need to worry more about when you're going to start publishing. Even small teaching colleges are starting to look for applicants to have multiple published papers. It's by far best if you get a couple of things published before you finish your PhD. Otherwise, finding a job could be tough. So, if this is your goal you may want to try specializing soon enough that you can start working on papers. Again, YMMV.
Finally, I just wanted to say that both my wife and I returned to graduate school after many years away and we've both had a blast. I hope you have a similar experience! Good luck!
"The large majority of America's nonfarm workers, about 85 percent, are employed in service-providing industries, construction, and government--sectors where import competition is minimal. To those workers, imports are an unambiguous blessing that spurs innovation, expands consumer choice, and raises real wages."Full Paper Here
Moreover, this breifing goes on to argue employment grows in proportion to imports. There's a fairly rational reason for this, if we can all stop foaming at the mouth long enough to actually think rationally: when employment grows we (consumers) have more cash to spend on goods and services. Since imports are a relatively fixed percentage of the overall economy, whenever the overall economy grows, so must imports. Why am I discussing imports if the argument is over services? Well, services are imported and exported just like goods. So, let's understand the real numbers, here:
The United States had a $64.8 billion trade (BEG ITAL) surplus in services in 2002, despite economic stagnation in Europe and Japan. Services accounted for 30 percent of all U.S. exports and 43 percent ($3.1 billion) of U.S. exports to India.Full Article Here
But, if half of our exports to India are in the form of services why are so many technical jobs going to India? Actually, there's no real evidence that's happening at all. There are two basic erroneous arguments made by the media today supporting the assumptions in this question. First, is the post hoc mistake: because the US economy is losing jobs and because after that happened India started gaining technology jobs, then India must be responsible for losses in American technology jobs. Actually, poor investments by venture capitalists and fund managers caused the loss in US jobs. The fact those losses occured coincidentally with India's technology boom is completely irrelevant.
Second, is the hasty generalization mistake: Bob Smith has just lost his job because his company opened a software development office in India, therefore all American technology jobs must be moving overseas. There just isn't enough evidence to support the generalization made by reporters. We may suspect that India is taking some portion of American jobs, but news reports by well-intentioned NPR and New York Times reporters aren't evidence that its hurting our economy.
All this panic and paranoia about jobs moving overseas doesn't even make sense when we consider the real economics of it. The "entire employment of the US" can't possibly be outsourced. Even if your argument wasn't a textbook example of the slippery slope fallacy, you'd still be wrong on an economic basis. If the USA loses a sufficient number of jobs, i.e. unemployment rises, the consumers will have less capital with which to buy foreign-made products. Domestic workers who are out of work will be willing to work for less, thus driving down the cost of locally made goods. When the cost of local goods and services drops below the cost of foreign made goods and services, then jobs will start to flow back into the USA. Adam Smith's invisible hand at work.
During the Clinton Administration monetary policy for the dollar kept our currency strong, which helped keep prices for foreign made consumer goods low. This was a good thing during that time because Asia and Europe were both in the midst of deep recessions and American consumer spending helped to bolster those economies through that trying time. The Bush Administration has since let the US Dollar sag in relation to other currencies. This has helped decrease the price of American goods and services abroad
You may want to consider XEP, as well. Its a commercial XSL-FO processor and is much less buggy than FOP. FOP is a great idea, but as of several months ago it wouldn't produce documents I would call "professional". XEP has most of the formatting objects requirements already implemented and seems to work quite well.
"Impressive" to your uncle must mean something along the lines of "capable of destroying all humanity." Which is, I admit, one possible definition. A bit sick, though.
Ebola is so scary because of how little would have to change for it to become "impressive". Ebola is an incredibly efficient killer, way more than smallpox's 30%-50% fatality rates. The Ebola that's around right now would be nothing compared to that incubated in an (infectious) victim for 6 months before the victim bled out. You could see epidemics wipe out entire countries in just a few years, if such a virus existed.
There are some interesting models for Ebola infection. They're all pretty scary. Here's one, in an Excel spreadsheet. Your uncle may have higher standards, but I get a bit freaked out when mathematical models start predicting 80% population losses.
FWIW, there are alot of interesting papers out there, if you want some hair-raising science...
P.S. AIDS is not a virus, its a syndrome caused by the HIV.
Its definately an FCC problem. The newest a/b/g chipset drivers are what's called "software defined radios". SDR is a major regulartory nightmare for the FCC because they can be reprogrammed by the user (more or less). Hence, the FCC wants to see infrastructure on the card to authenticate the code that runs there [kernel discussion]. Annoyingly.
The madwifi project is developing drivers for the Atheros a/b/g chipset. I've been using them and they appear to be reasonably good, for the moment. But, the distribute with a uuencoded binary blob that unpacks into a kernel module... I hear there are access points on the market that are basically Atheros a/b/g mini-PCI cards inside a plastic casing.
On the flip side, at least Intel sees the need to convince the FCC to open spectrums [quote]:
A third major challenge facing SDR technology is convincing the FCC to open the radio spectrum. In the past, the FCC has regulated specific radio bands for different types of communications. A radio device is then licensed for use in only a specific frequency range. Intel and other industry leaders would like to see devices licensed for multiple radio spectra, rather than for only one communications band. This would allow manufacturers to make a single device that could broadcast and/or receive at any appropriate frequency. The frequency used for a specific type of communication could then depend on the device or user identification, such as for National Guard, police, fire, Air-Sea Rescue, animal control, border patrol, road construction, clean-water works, and so on.
They should give more than the minimum to give this very valuable employee solid motivation to continue his fine work. Nikola Tesla came to the U.S. to work for Edison, who made life unpleasant by failing to reward Tesla for his excellent work. Tesla eventually left and invented A/C dynamos for
Westinghouse, which helped him defeat Edison in the electricty market. Eventually
making the Westinghouse corporation became so powerful J.P. Morgan and G.E. eventually gave up competing (and licensed the patents from Westinghouse).
Sadly, Edison was too arrogant to see the value
in Tesla's future efforts and, rather than reaping
the benefits of a lifetime's creative inventions and a potentially brilliant partnership, he was massively defeated in the electrical market. There's actually a great new book out called Empires of Light, in case you want to learn from
history by reading about it.
Anyway, filing a patent is normally worth hundreds of thousands of dollars in equity to a corporation. Even, frankly, if the patent isn't really worth enforcing or developing, it still increases the value of the corporation. An employee who is creative enough to continue developing patents is an employee I'd be
interested in keeping!
As a technologist and a manager, I'm often saddened at the short-sightedness of my peers. These kinds of decisions are legal. But, they're cultural suicide for a corporation. Imagine the bright young engineer with several good ideas who hears this guy come back to his cube and tell this story? Why would he let out a peep about his own patentable ideas in that office?
If he holds on to them, he can negotiate better terms when he changes jobs. Which he may begin looking to do very shortly.
If you have to create a new project to deal with upgrading code in a distributed environment, you incurred the bug way back in the requirements / design stage of the system. Implementing a fix then would have been pretty expensive, too. Does the cost of that project really change all that much between the beginning and the end of the lifecycle?
Posit: It costs more because where one massive design problem has been built into the system, there are likely to be several (dozens?) more just as bad or worse. Finding that one bug is like turning the lights on in the kitchen--all of a sudden you spot twenty roaches, not just one. If your mobile phone infrastructure included a mechanism to test, deploy, and undeploy code from the very beginning then each incremental update would be cheaper.
Getting rid of a single roach can be a hassle but its rarely expensive in a well-designed system. Getting rid of a colony of roaches that are living inside a roccoco infrastructure is a massive undertaking-- after all, you wouldn't want to damage the gold leaf plate, right?:-)
Design simpler systems, with fewer inherent problems, and you end up with fewer colonies. There will always be roaches, but at least you won't have an infestation on your hands. When you do find a small colony, you'll have less to rip apart and fix. Knowing good design when you see it seems to be an art, rather than a science. There appear to be lots of "software architects" who think complexity is the hallmark of good design.
A friend and I were driving through a small Vermont town a few weeks ago and figured that would be a great place for this kind of experiment. Downtown areas in the rural northeast seem to have pretty dense concentrations of buildings that could easily be hooked together this way. Moreover, using the right kind of antennae you can likely connect buildings together that are a mile or more apart with very satisfying results. Another likely application is high-speed Internet access in third-world countries where wired infrastructure like broadband is unavailable.
/. has not been saying this for years. I rarely see cogent slashdot posts on economics, much less posts that include a mathematical model. Slashdotters usually limit themselves to the type of comments you just provided, "See!! We're right! Woo!!" But, we mustn't confuse instinct with academic analysis. Moreover, we ought not confuse the article, originally
posted here, with the actual paper. The staff research report by Boldrin and Levine here
is 40 pages of economic theory. The summary is mostly fluff and sound bytes. Yeah, its appealing
to think it may be correct, but the arguments on both sides are very strong. more
Applying, say, a contrast filter to a digital image to bring out details is no different from the subjective treatment that a conventional photograph gets when developed in a darkroom.
At least with fingerprint analysis, the goal is to identify and match artifacts of two images -- one taken under controlled circumstances, the other recovered. My experience with signal enhancement algorithms is limited, but spurious artifacts are a problem with many of them. "Applying a contrast filter" sounds simple and straightforward. But, what if there's a bug in the algorithm? What if a new algorithm is found that's faster, but produces more artifacts?
So, perhaps a good question is: What constitutes a "forensic" algorithm? Which algorithms have forensic utility and which do not? What methods do we use to test and evaluate them? With physical processes, these questions seem easier to answer. The algorithms are natural, e.g, luminosity of blood under a certain light source. But, if they weren't how would we know they're accurate? If a computer "enhances" a picture of a blood stain, how do we know it can tell the difference between katsup and blood? We wouldn't know until we'd tested it heavily.
I'm no M$ fan, but I deal with it at work so I make a point at figuring out how to deal with the problems. Frankly, this isn't a suprise. The most well secured enterprises I've seen allow only internal ActiveX publishers -- ActiveX is just too hard to make safe.
Looking forward, I recently picked up.NET Framework Security -- anyway, it seems like Microsoft is at least attempting to solve this particular problem. And, their approach isn't completely idiotic. Really.
Mobile code that runs in the.NET common language runtime (read: M$ JVM) is controlled by a fairly sophisticated access control system. The default policy in XPsp1 from M$ allows no code from the Internet to execute, at all. Not exactly what I want as a user, but its what I want as an admin...
Frustratingly, you can't run.NET mobile code without also enabling ActiveX controls. Not sure what the issue there is, but I suspect the CLR loader is some sort of ActiveX control. Anyone know about that?
Anyway... here's some additional links to M$ references on mobile code:
FWIW, the Best Buy near me has a demo of XP MC, right now. Its running on some HP widget that looks okay with a decent flat screen. Price point was higher $2,300 for everything. Insanity.
A few years ago, one of my freinds suggested that spam be defeated by requiring everyone to do a peice of hard computation before sending you an e-mail. Make the computations something interesting -- SETI@Home blocks? -- or valuable to have done, maybe. So, only spammers care to circumvent the system by (a) refusing or (b) cheating, both of which can be detected at some point.
Everyone else will figure, "what the heck, why not?" and just do it...
Always seemed like an elegent solution to me, really.
It's the conventional wisdom that we hear so much and that we'd all like to believe--mainstream, big-label music sucks, and all the interesting stuff is being done by small, independent artists--but the fact is that it just isn't true.
What a great point. The flipside of this, which I haven't yet heard, is that the Internet doesn't help good, unknown music get heard more. The Internet just makes it cheaper to move music around.
In light of this, the record labels are fighting to protect a business model that doesn't make sense any more. Since they used to have control over distribution, and distribution is now super cheap (Napster, Gnutella, WinMX), it makes no sense to support the RIAA in their efforts to charge for the distribution alone.
But, music will still be a business after its all shaken out. One way or another, people have to make a living. The recording studio costs money, producers who mix down good sounding tracks cost money, studio musicians to back-fill cost money.
So, what we REALLY need is a way for artists to charge fees for what they do without charging unreasonable fees for distribution. The current business models don't really deal with this very well, frankly. Which isn't suprising because everyone charges for everything at distribution time -- toothpaste, cars, houses, programming services, etc. All paid for when you recieve the thing.
Figure out an open and efficient marketplace where charges are incurred reasonably but not at distribution time. That would be cool.
I'm not sure why everyone is complaining. My Intel Pentium III with Speedstep runs perfectly well under linux and winders. And, I have complete control over the power management settings. If you're curious about the ongoing development of the linux kernel drivers to support speedstep and similar technologies from AMD and others look here. These drivers are available in the 2.4.19 kernel.
Now, caveat emptor, the linux drivers (at least) are still very much in development. Intel has (god bless them) chosen to withhold specifciations for SpeedStep and the cpufreq project is, therefore, having to reverse engineer them.
But, I find this feature to be very cool -- it helps my laptop run longer on battery power, which road-warriors like myself really appreciate. Especially during cross country flights. Ugh. Nothing like 6 hours to kill and no computer to code on, right?;-)
If you are finding you can't run (say) XP with all those fading menus and visual candy canes, then you have my sympathy. Honestly. Go buy yourself a new laptop, its sure to fix whatever problems you're having with the current one.
Your little credit union probably bought the service it provides you from a larger company like Fiserv.
Having worked for several credit unions as a consultant, I doubt that any little ones write their own code. Their attitude is generally, "I want to do whatever is necessary." But, they're complete tightwads. I guess you have to be to be a banker.
Anyway, I suspect companies don't do this deliberately, but out of a sense of minimizing the work-load. They don't want to do QA, at all, but they definately don't want to spend for QA on 10 different browser platforms.
In the end, I wonder why people get so up in arms about 1 browser versus another. I have 3 on my desktop. Is there something wrong with me that I don't really get emotional over which one I need to use to best view the site?
There are several such initiatives under way in the on-line library community -- librarians collect so much cruft and since they tend never to throw things out, they feel an even stronger need than you for good metadata. Dewey Decimal System is one such (very simple) metadata standard (sortof).
Anyway, SunSITE.UNC.edu -- now iBiblio.org -- has required Linux developers uploading software to/incoming to include a inux Software Map (lsm) file for quite a long time now. The.lsm file is a basic metadata file in a fairly simple format. So, you might look at that: http://www.ibiblio.org/pub/Linux/LSM-TEMPLATE
The Dublin Core initiative is a more generalized attempt to answer the question, "How do we standardize on a metadata format?" Dublin Core is using XML and XML DTDs as the basis for their work. It applies to not only software but also to other online resources. So, as one might guess its arcane and difficult to understand at best and completely impenetrable most of the time. You can find more about Dublin Core at http://www.purl.org/dc
Sadly, most search engine companies focus on searching a specific kind of document type -- like HTML -- for arbitrary content. Interestingly, searching metadata is both an easiser computational problem to solve and more productive for the user. Unfortunately, its also a far more difficult social problem. Getting everyone to write common metadata is very, very difficult. Going back and writing metadata for any sizeable archive (say, iBiblio, for example) is a Herculean task. I think most of the coders who write search engines are more interested in the actual mathematics behind searching than they are in actual Document Retrieval. You might also check out http://www.cnidr.org, who were the authors of Isearch and some other good searching tools.
Slashdot Mirror
There is one skill that I'd recommend you try to pick up immediately, though: RESEARCH. You need to learn two things. First, you need to learn how to find papers about what you're interested in. And, scholar.google.com isn't sufficient (yet). Second, you need to learn to actually read and understand those papers. Sometimes it can take a month just to get through a single paper. But, learn how to push through and you'll reap dividends. Some graduate departments do a good job of teaching research skills. Others are lousy. Your experience could differ alot. But, to be good in graduate school requires alot more than just attendance. So, learn to do good research before you're paying to do so and you'll be happier.
As far as admissions goes, I think you should call the department you want to study in and ask the graduate coordinator. Every department is different. Some Universities really cater to returning professional students -- George Mason University in Virginia is that way. Other Universities have different philosophies.
In my field (mathematics) most admissions decisions are ultimately made based upon academic recommendations. I think that GPA and test scores are important, too. My wife says that when she worked in the admissions office they used GPA and test scores just to sort the applications into "bother to read recommendations" and "don't bother" piles. YMMV.
I can't speak to the differences between US and UK schools. I've got a number of friends who went to school in Canada and I don't think our systems differ much. Except Canadian schools are better funded so they're cheaper.
If you're trying to decide where to apply I'd recommend that you start by reading conference papers and published research. Try to find some research topics that are of particular interest to you. Then, look at where there are academics who are currently working in that field. You may discover that the preeminent researcher in your interest is at a really cheap school! How cool would that be! If you have no idea what you want to do with your life, then try to find a department that's large enough to support lots of research interests. That way, you can used your Masters degree as an opportunity to sample everything and see what you like best.
If you're returning to school because you just want another degree (and eventually a better job) then you should consider trying to make sure you stay very well rounded. On the other hand, if you're trying to become an academic you need to worry more about when you're going to start publishing. Even small teaching colleges are starting to look for applicants to have multiple published papers. It's by far best if you get a couple of things published before you finish your PhD. Otherwise, finding a job could be tough. So, if this is your goal you may want to try specializing soon enough that you can start working on papers. Again, YMMV.
Finally, I just wanted to say that both my wife and I returned to graduate school after many years away and we've both had a blast. I hope you have a similar experience! Good luck!
The Cato Institute thinks the following:
"The large majority of America's nonfarm workers, about 85 percent, are employed in service-providing industries, construction, and government--sectors where import competition is minimal. To those workers, imports are an unambiguous blessing that spurs innovation, expands consumer choice, and raises real wages." Full Paper Here
Moreover, this breifing goes on to argue employment grows in proportion to imports . There's a fairly rational reason for this, if we can all stop foaming at the mouth long enough to actually think rationally: when employment grows we (consumers) have more cash to spend on goods and services. Since imports are a relatively fixed percentage of the overall economy, whenever the overall economy grows, so must imports. Why am I discussing imports if the argument is over services? Well, services are imported and exported just like goods. So, let's understand the real numbers, here:
The United States had a $64.8 billion trade (BEG ITAL) surplus in services in 2002, despite economic stagnation in Europe and Japan. Services accounted for 30 percent of all U.S. exports and 43 percent ($3.1 billion) of U.S. exports to India. Full Article Here
But, if half of our exports to India are in the form of services why are so many technical jobs going to India? Actually, there's no real evidence that's happening at all. There are two basic erroneous arguments made by the media today supporting the assumptions in this question. First, is the post hoc mistake: because the US economy is losing jobs and because after that happened India started gaining technology jobs, then India must be responsible for losses in American technology jobs. Actually, poor investments by venture capitalists and fund managers caused the loss in US jobs. The fact those losses occured coincidentally with India's technology boom is completely irrelevant.
Second, is the hasty generalization mistake: Bob Smith has just lost his job because his company opened a software development office in India, therefore all American technology jobs must be moving overseas. There just isn't enough evidence to support the generalization made by reporters. We may suspect that India is taking some portion of American jobs, but news reports by well-intentioned NPR and New York Times reporters aren't evidence that its hurting our economy.
All this panic and paranoia about jobs moving overseas doesn't even make sense when we consider the real economics of it. The "entire employment of the US" can't possibly be outsourced. Even if your argument wasn't a textbook example of the slippery slope fallacy, you'd still be wrong on an economic basis. If the USA loses a sufficient number of jobs, i.e. unemployment rises, the consumers will have less capital with which to buy foreign-made products. Domestic workers who are out of work will be willing to work for less, thus driving down the cost of locally made goods. When the cost of local goods and services drops below the cost of foreign made goods and services, then jobs will start to flow back into the USA. Adam Smith's invisible hand at work.
During the Clinton Administration monetary policy for the dollar kept our currency strong, which helped keep prices for foreign made consumer goods low. This was a good thing during that time because Asia and Europe were both in the midst of deep recessions and American consumer spending helped to bolster those economies through that trying time. The Bush Administration has since let the US Dollar sag in relation to other currencies. This has helped decrease the price of American goods and services abroad
You may want to consider XEP, as well. Its a commercial XSL-FO processor and is much less buggy than FOP. FOP is a great idea, but as of several months ago it wouldn't produce documents I would call "professional". XEP has most of the formatting objects requirements already implemented and seems to work quite well.
"Impressive" to your uncle must mean something along the lines of "capable of destroying all humanity." Which is, I admit, one possible definition. A bit sick, though.
Ebola is so scary because of how little would have to change for it to become "impressive". Ebola is an incredibly efficient killer, way more than smallpox's 30%-50% fatality rates. The Ebola that's around right now would be nothing compared to that incubated in an (infectious) victim for 6 months before the victim bled out. You could see epidemics wipe out entire countries in just a few years, if such a virus existed.
There are some interesting models for Ebola infection. They're all pretty scary.
Here's one, in an Excel spreadsheet. Your uncle may have higher standards, but I get a bit freaked out when mathematical models start predicting 80% population losses.
FWIW, there are alot of interesting papers out there, if you want some hair-raising science...
P.S. AIDS is not a virus, its a syndrome caused by the HIV.
Its definately an FCC problem. The newest a/b/g chipset drivers are what's called "software defined radios". SDR is a major regulartory nightmare for the FCC because they can be reprogrammed by the user (more or less). Hence, the FCC wants to see infrastructure on the card to authenticate the code that runs there [kernel discussion]. Annoyingly.
The madwifi project is developing drivers for the Atheros a/b/g chipset. I've been using them and they appear to be reasonably good, for the moment. But, the distribute with a uuencoded binary blob that unpacks into a kernel module... I hear there are access points on the market that are basically Atheros a/b/g mini-PCI cards inside a plastic casing.
On the flip side, at least Intel sees the need to convince the FCC to open spectrums [quote]:
A third major challenge facing SDR technology is convincing the FCC to open the radio spectrum. In the past, the FCC has regulated specific radio bands for different types of communications. A radio device is then licensed for use in only a specific frequency range. Intel and other industry leaders would like to see devices licensed for multiple radio spectra, rather than for only one communications band. This would allow manufacturers to make a single device that could broadcast and/or receive at any appropriate frequency. The frequency used for a specific type of communication could then depend on the device or user identification, such as for National Guard, police, fire, Air-Sea Rescue, animal control, border patrol, road construction, clean-water works, and so on.
Have you looked at OpenSTA? Our development team in NH uses it and they enjoy it quite well.
They should give more than the minimum to give this very valuable employee solid motivation to continue his fine work. Nikola Tesla came to the U.S. to work for Edison, who made life unpleasant by failing to reward Tesla for his excellent work. Tesla eventually left and invented A/C dynamos for Westinghouse, which helped him defeat Edison in the electricty market. Eventually making the Westinghouse corporation became so powerful J.P. Morgan and G.E. eventually gave up competing (and licensed the patents from Westinghouse).
Sadly, Edison was too arrogant to see the value in Tesla's future efforts and, rather than reaping the benefits of a lifetime's creative inventions and a potentially brilliant partnership, he was massively defeated in the electrical market. There's actually a great new book out called Empires of Light, in case you want to learn from history by reading about it.
Anyway, filing a patent is normally worth hundreds of thousands of dollars in equity to a corporation. Even, frankly, if the patent isn't really worth enforcing or developing, it still increases the value of the corporation. An employee who is creative enough to continue developing patents is an employee I'd be interested in keeping!
As a technologist and a manager, I'm often saddened at the short-sightedness of my peers. These kinds of decisions are legal. But, they're cultural suicide for a corporation. Imagine the bright young engineer with several good ideas who hears this guy come back to his cube and tell this story? Why would he let out a peep about his own patentable ideas in that office? If he holds on to them, he can negotiate better terms when he changes jobs. Which he may begin looking to do very shortly.
If you have to create a new project to deal with upgrading code in a distributed environment, you incurred the bug way back in the requirements / design stage of the system. Implementing a fix then would have been pretty expensive, too. Does the cost of that project really change all that much between the beginning and the end of the lifecycle?
:-)
Posit: It costs more because where one massive design problem has been built into the system, there are likely to be several (dozens?) more just as bad or worse. Finding that one bug is like turning the lights on in the kitchen--all of a sudden you spot twenty roaches, not just one. If your mobile phone infrastructure included a mechanism to test, deploy, and undeploy code from the very beginning then each incremental update would be cheaper.
Getting rid of a single roach can be a hassle but its rarely expensive in a well-designed system. Getting rid of a colony of roaches that are living inside a roccoco infrastructure is a massive undertaking-- after all, you wouldn't want to damage the gold leaf plate, right?
Design simpler systems, with fewer inherent problems, and you end up with fewer colonies. There will always be roaches, but at least you won't have an infestation on your hands. When you do find a small colony, you'll have less to rip apart and fix. Knowing good design when you see it seems to be an art, rather than a science. There appear to be lots of "software architects" who think complexity is the hallmark of good design.
A friend and I were driving through a small Vermont town a few weeks ago and figured that would be a great place for this kind of experiment. Downtown areas in the rural northeast seem to have pretty dense concentrations of buildings that could easily be hooked together this way. Moreover, using the right kind of antennae you can likely connect buildings together that are a mile or more apart with very satisfying results. Another likely application is high-speed Internet access in third-world countries where wired infrastructure like broadband is unavailable.
FWIW, you can find more of Levin's work at various places. Prof. Danny Quah also has some thoughts on the subject.
At least with fingerprint analysis, the goal is to identify and match artifacts of two images -- one taken under controlled circumstances, the other recovered. My experience with signal enhancement algorithms is limited, but spurious artifacts are a problem with many of them. "Applying a contrast filter" sounds simple and straightforward. But, what if there's a bug in the algorithm? What if a new algorithm is found that's faster, but produces more artifacts?
So, perhaps a good question is: What constitutes a "forensic" algorithm? Which algorithms have forensic utility and which do not? What methods do we use to test and evaluate them? With physical processes, these questions seem easier to answer. The algorithms are natural, e.g, luminosity of blood under a certain light source. But, if they weren't how would we know they're accurate? If a computer "enhances" a picture of a blood stain, how do we know it can tell the difference between katsup and blood? We wouldn't know until we'd tested it heavily.
I'm no M$ fan, but I deal with it at work so I make a point at figuring out how to deal with the problems. Frankly, this isn't a suprise. The most well secured enterprises I've seen allow only internal ActiveX publishers -- ActiveX is just too hard to make safe.
.NET Framework Security -- anyway, it seems like Microsoft is at least attempting to solve this particular problem. And, their approach isn't completely idiotic. Really.
.NET common language runtime (read: M$ JVM) is controlled by a fairly sophisticated access control system. The default policy in XPsp1 from M$ allows no code from the Internet to execute, at all. Not exactly what I want as a user, but its what I want as an admin...
.NET mobile code without also enabling ActiveX controls. Not sure what the issue there is, but I suspect the CLR loader is some sort of ActiveX control. Anyone know about that?
.NET: Enforce Code Access Rights... .NET Framework
Looking forward, I recently picked up
Mobile code that runs in the
Frustratingly, you can't run
Anyway... here's some additional links to M$ references on mobile code:
Security in
Security in the
FWIW, the Best Buy near me has a demo of XP MC, right now. Its running on some HP widget that looks okay with a decent flat screen. Price point was higher $2,300 for everything. Insanity.
A few years ago, one of my freinds suggested that spam be defeated by requiring everyone to do a peice of hard computation before sending you an e-mail. Make the computations something interesting -- SETI@Home blocks? -- or valuable to have done, maybe. So, only spammers care to circumvent the system by (a) refusing or (b) cheating, both of which can be detected at some point.
Everyone else will figure, "what the heck, why not?" and just do it...
Always seemed like an elegent solution to me, really.
It's the conventional wisdom that we hear so much and that we'd all like to believe--mainstream, big-label music sucks, and all the interesting stuff is being done by small, independent artists--but the fact is that it just isn't true.
What a great point. The flipside of this, which I haven't yet heard, is that the Internet doesn't help good, unknown music get heard more. The Internet just makes it cheaper to move music around.
In light of this, the record labels are fighting to protect a business model that doesn't make sense any more. Since they used to have control over distribution, and distribution is now super cheap (Napster, Gnutella, WinMX), it makes no sense to support the RIAA in their efforts to charge for the distribution alone.
But, music will still be a business after its all shaken out. One way or another, people have to make a living. The recording studio costs money, producers who mix down good sounding tracks cost money, studio musicians to back-fill cost money.
So, what we REALLY need is a way for artists to charge fees for what they do without charging unreasonable fees for distribution. The current business models don't really deal with this very well, frankly. Which isn't suprising because everyone charges for everything at distribution time -- toothpaste, cars, houses, programming services, etc. All paid for when you recieve the thing.
Figure out an open and efficient marketplace where charges are incurred reasonably but not at distribution time. That would be cool.
I'm not sure why everyone is complaining. My Intel Pentium III with Speedstep runs perfectly well under linux and winders. And, I have complete control over the power management settings. If you're curious about the ongoing development of the linux kernel drivers to support speedstep and similar technologies from AMD and others look here. These drivers are available in the 2.4.19 kernel.
;-)
Now, caveat emptor, the linux drivers (at least) are still very much in development. Intel has (god bless them) chosen to withhold specifciations for SpeedStep and the cpufreq project is, therefore, having to reverse engineer them.
But, I find this feature to be very cool -- it helps my laptop run longer on battery power, which road-warriors like myself really appreciate. Especially during cross country flights. Ugh. Nothing like 6 hours to kill and no computer to code on, right?
If you are finding you can't run (say) XP with all those fading menus and visual candy canes, then you have my sympathy. Honestly. Go buy yourself a new laptop, its sure to fix whatever problems you're having with the current one.
Your little credit union probably bought the service it provides you from a larger company like Fiserv.
Having worked for several credit unions as a consultant, I doubt that any little ones write their own code. Their attitude is generally, "I want to do whatever is necessary." But, they're complete tightwads. I guess you have to be to be a banker.
Anyway, I suspect companies don't do this deliberately, but out of a sense of minimizing the work-load. They don't want to do QA, at all, but they definately don't want to spend for QA on 10 different browser platforms.
In the end, I wonder why people get so up in arms about 1 browser versus another. I have 3 on my desktop. Is there something wrong with me that I don't really get emotional over which one I need to use to best view the site?
There are several such initiatives under way in the on-line library community -- librarians collect so much cruft and since they tend never to throw things out, they feel an even stronger need than you for good metadata. Dewey Decimal System is one such (very simple) metadata standard (sortof). Anyway, SunSITE.UNC.edu -- now iBiblio.org -- has required Linux developers uploading software to /incoming to include a inux Software Map (lsm) file for quite a long time now. The .lsm file is a basic metadata file in a fairly simple format. So, you might look at that: http://www.ibiblio.org/pub/Linux/LSM-TEMPLATE
The Dublin Core initiative is a more generalized attempt to answer the question, "How do we standardize on a metadata format?" Dublin Core is using XML and XML DTDs as the basis for their work. It applies to not only software but also to other online resources. So, as one might guess its arcane and difficult to understand at best and completely impenetrable most of the time. You can find more about Dublin Core at http://www.purl.org/dc
Sadly, most search engine companies focus on searching a specific kind of document type -- like HTML -- for arbitrary content. Interestingly, searching metadata is both an easiser computational problem to solve and more productive for the user. Unfortunately, its also a far more difficult social problem. Getting everyone to write common metadata is very, very difficult. Going back and writing metadata for any sizeable archive (say, iBiblio, for example) is a Herculean task. I think most of the coders who write search engines are more interested in the actual mathematics behind searching than they are in actual Document Retrieval. You might also check out http://www.cnidr.org, who were the authors of Isearch and some other good searching tools.