be careful with supabeast's advice. Remember that once your employer has the suspicion that you are looking for other jobs, there's no turning back. What i mean is that you may or may not leave your current job, but you can't take back the fact that your employer is now suspicious. Personally, i think of it as an ace-in-hand. I think it's better just to keep it secret until after you've accepted an offer and are almost ready to resign. Remember, that you don't want to burn bridges... even if you hate your current employer. You would be surprised how small your industry is. Even if you change industries, you may run across your former employer again in a different life. I used to get into arguments with one of my bosses on a regular basis. He was ridiculous... but i left under good circumstances, to do something that was "better for my career", not because i hated my job/boss. That boss still calls me about once a month just to see what's up and we bounce ideas off of one another. He gave me a glowing review recently for a different job.
More importantly though is your current situation. You don't need time to interview for other jobs, you just need more time, period. Once you take a few 'sick afternoons' to go interview or leave for a long lunch interview, you may realize just how easy it is to leave you job at a reasonable hour. I had a job that used to make us stay late and work long hours. I was very unproductive. One day, because i hated that job so much, i left early (granted early=6:30, 8:30-6:30). I started doing it every day. One day, as i was leaving my manager stopped me and said 'i noticed you've been leaving at 6:30' (he dare not say 'early'!). I told him, matter of factly, that tonight i was meeting a friend and had to get ready. I made no excuses for the other nights. After that he learned to have more respect for the time i was in the office. He knew that i would be the 1st to leave the office and he would make extra efforts to make sure that i would have everything i needed to finish my projects before 6:30. He knew i wouldn't stay one second later, unless it was an emergency.
My coworkers were jealous that I got to leave 'early' and they had to stay until 9-10pm every night. I told them it was their own doing and they could leave when they wanted. When bonus time came around, my billing was on target and my bonus reflected that. I got more done during 8:30-6:30 when i felt good about my job, than i ever got done working 8:30AM-10PM and hating it.
This is one of my great revelations. It was beginning of restructuring my social life, getting physically fit, starting bike racing, and generally getting my life in order.
Due to scandals, consulting firms come cheap right now. This is a GREAT oppurtunity for IBM to merge their consulting arm with PWC Consulting.
Remember, the new corporate scrutiny will only increase revenue for consulting firms, since the consulting firms are hired to "straighten out the books".
As far as "service&consulting being the last refuge of the Damned" goes... I don't think it applies to financial consulting. Only to software/engineering/research consulting.
These figures are off the top of my head, please correct me if wrong.
Consulting made up ~25% of IBM's 18 Billion in revenue last year. It's been a significant part of IBM for a long time and many credit the consulting arm of IBM for pulling IBM through the '90s. IBM is *much* more than a harddrive manufacturer or computer hardware company.
If you have a hard time understanding this, put it this way... IBM was the Microsoft of the 60s, 70s, and 80s. Back then, Computers were mostly for *accounting* and various services that you would hire consultants for. IBM took advantage of this and became big. I question Microsoft, because after ~10 years as a substantial Monopoly, they haven't successfully branched into another area and their largest competition is given away for free.
Anyways, I figured a consulting explanation was in order for anybody who wants to post a *WHY WOULD IBM BUY A CONSULTING FIRM?* comment.
BTW, does anybody know what happened to the name "Monday"? I thought PWC consulting was changing it's name to "Monday". Am I wrong?
Not supporting Apple is okay. That's also saying *I support capitalism and a free market*. I just think that work-arounds and copyright infringement don't support anything except the idea that p2p file sharing is bad.
Yes, I know that setting the date ahead and then back isn't really copyright infringement, but you're still getting an improved version of a product that you aren't supposed to have, without paying for it.
Another 30 second workaround is: Support the software by purchasing it!! It's amazing how effective it is. Since I purchased quicktime pro, it hasn't nagged me once when I started. This super-secret method also enables some super-secret features. For example, now I can save Quicktime in Mpeg or DV formats! I can cut and crop movies and parts of movies or even combine movies.
Doing this with software such as Quicktime is also rebellious, as it sends the word to Apple that *hey your software is worthwhile to me*. It states that *I don't succumb to predatory bundle-it-with-the-OS techniques* and most importantly *I support capitalism and the free market*.
Do you get to do that by playing with the calendar?
I am using it right now. My guess is that it's still IE 5.0 codebase. Not really much is changed. I clicked on one of my bookmarks at the top only to find, that it still pulls down a menu occasionally, rather then going to the site of the bookmark.
The quartz rendering is nice, but I havn't noticed many other improvements.
I agreed with Bladerunner being #1 on the list. I did miss, however, Total Recall. IMHO that should have been on the list. Although a Clockwork Orange is a good movie, does it really qualify as Sci-Fi? Perhaps Total Recall could have taken that spot.
I'm a little unsure about Jurassic Park being on the list. Sure it's another good movie, but these are supposed to be the best Sci-Fi movies. I would have put Aliens or T2-Judgement day on the list in order to have 2 from the same series.
Don't forget that T9 technology has nothing to do with the layout of the keys. You can use T9 in addition to having a nicely laid out keypad and that would speed you up VERY much. The layout of a number pad is bad for entering letters. Enough said.
In NO way does Despair's trademark spoof the USPTO. It is a valid trademark for a valid company.
The fact that they applied for it and were granted it was SOP (Standard Operating Procedure). They then went on to joke about *patenting* the emoticon and charging people for every time it is used. This might have been the *spoof* part, but not the trademark itself. I wish that people would get this straight.
Aagh! The irony nazi's summary of the current state of affairs:
1. Stealing MP3s is illegal. Somebody will post an intelligently worded comment that will be modded to +5 based upon this premise. Most of the other comments that point this out will be modded -1 Troll.
2. RIAA Screws over artists and consumers. Like we didn't know this already. </sarcasm>
3. Kazaa is a file sharing network, even if it is a *spyware* file sharing network. If the RIAA feels it can make more money without the Kazaa Spynet, then it won't negotiate with Kazaa. Plain and simple. If the RIAA geniuses have an idea that they think will make bigger $$$ then they will do it. They will then sue Kazaa who needs illegal MP3s to make their network popular, yet will be sued if they use them. (napster all over again).
4. RIAA will come out with their own draconian file format and continue to blame illegal file sharing whenever the consumer chooses getting more *value* for their money (via trading MP3s) as opposed to getting less *value* for more money (restricted and encrypted non-MP3 formats). Personally, I blame capitalism for this consumer decision, not file sharing. I feel that I have a right, as a consumer, to choose the file format, distribution technique, and amount that I pay. Although this argument shouldn't be used to justify illegal copying of music, in a sense, one can assign a $$ value for breaking the law. I would surely pay a small amount of money for music just to know that I wasn't doing anything illegal. I would place even more value (i.e. pay more) if I knew that the artist was getting a significant portion of what I was paying.
A long time ago (~1996), I said that the recording industry will never be able to do away with an open audio format (at least as open as MP3 is). The reason why is because they will have to offer something with much more *percieved* value than what I can download for free and play on any music device. No more $14 CDs (a new CD was $14 at the time).
cygnus IBM Red Hat (although it may be accounting tricks, they did claim profit last quater) Tivo (I'm not sure about them, but insert a hardware vender here, Axis cameras if not Tivo) O'reilly Publishers (some of their books fall under GPLish licenses)
I tried to cover all the bases too. Services, Services/Hardware, Services/OS, Consumer Hardware, and a book company.
Nobody is claiming that games/propriatary applications have to be GPL'd. I don't think they should be. The basics of the OS, and the tools of programming, OTOH, should be opened and free so that a maximum number of people can make them better and understand them.
I'm sorry, I'm browsing at 5 and don't know what this thread is about. Mod me as offtopic if appropriate. I'm writing this to clear up a few things in guacamolefoo's post and say a few things in general about software companies.
On the topic of dividends:
MS doesn't need to pay dividends. They will continue to NOT pay dividends on the stock and the stockholders will be happy. Here's why: 1. MS earns revenues, pays taxes on revenues (although MS pays little or no taxes, but that's a another tale). 2. after-tax income would go to investors in the form of dividends 3.investors pay taxes on dividends. What I desribe here is that the investors end up with income that has had taxes paid twice on it, which is rather watered down if you ask me. What I would rather have happen, and THIS is what MS is known for is the following: 1. MS earns revenues and pays taxes on it. 2. MS buys back its own stock 3. Stock price increases substantially 4. Investors have a capital gain, which they don't have to pay any taxes on until they sell their stock, and if they are smart, they will hardly pay any capital gains tax.
On the topic of a MS having STAGGERING Cash on hand:
This is acceptable and even expected. Look at balance sheets. MS has more cash-on-hand than Walmart, Exxon, and Ford combined. Is this a surprise? No. Walmart has thousands of chain stores accross the country. Exxon has billions of dollars worth of fuel pumping, refining, and distribution equipment. Ford has billions of dollars of factories, raw materials, and vehicles. If any of these companies had excess cash-on-hand, that means they wouldn't be making money off of it by building stores, or pipelines, or Vehicles. When you look at it this way, what does MS really have? They don't have much. They have one campus in Redmond Washington and a bunch of Intellectual property, which doesn't translate well to a financial statement.
The only thing that gives MS *any* investment value is this $40B. The same holds for any software company and this is the reason that so many of the "mind-share" dot-com companies went bankrupt. If a software company doesn't have cash, then it doesn't have much of anything, and nobody wants to invest in it. You can't add *nix guru's and Aeron Chairs to your end-of-year report. *nix gurus can always get jobs elsewhere and Aeron chairs are, well, a waste of money.
About catastrophe Hedging software:
A lot of people on Slashdot have been writing that this software is to protect MS against a DOJ "disaster", a marketing flop (X-box?) or other lawsuits. This couldn't be further from the truth. Catastrophe Hedging is a very active part of finance and I can assure you that they hedging against large interest rate changes, stock-market disasters, foreign currency melt-downs, and Enrons... just like any other Risk Analysis software does. This is basic quantitative finance.
To end this rant session, I would like to state that Quantitative finance is a very interesting and geeky field. It's the 100% math and computer field that is often overlooked by the geek community. Did you know that a large number of hedge funds use Linux primarily for their analysis computers? Did you know that PDE's and Stochastic Calculus runs rampant all throughout finance?
Thanks for the tip about Dreamweaver and XHTML. I've always liked that program, but find it hard to justify the purchase. Maybe now, I'll look into picking up a copy.
One nitpick, however...
I feel a since of vindication that my stuff already works
Do you feel a 'since' or a 'sense'?
It's been a while since I felt a sense of 'since', but to be honest, I haven't had a sense for 'since' since. or I'll give my 2 cents, since you seem to need sense.
I just wanted to give you a sense for the grammar nazi since he is not around.
Sorry. I've lost my sense since I haven't had much sleep lately. Please forgive me after you read this.
People who don't understand why you would/wouldn't log in as root are *extremely* unlikely to be playing around with 'rm', 'chmod', and 'mv'.
You would have a better argument saying something to the effect of "dragging an important system file into the trash" or moving/renaming an important file/folder.
I find it amazing how many people don't want to *login* to their computers.
They tell me, "I know that it's safer to log into my computer, but it's such a pain." --to which my usual reply is "You don't know that it's safer to log in."
Well Mr. pq. I can assure you that IANAPA and I would like to *inform* you of a few things.
First of all, have you ever heard of Occum's razor? I would have assumed that you knew of such things, beinging a professional astronomer and all. It states that all things being equal, the simplest explanation holds more weight than the complicated explanations. It is simpiler to go with the current theories rather than trying to figure out the unnecessary complications involved with a cyclic universe. The cyclic universe that you describe in the abstract is similar to something that I read in Brief History of Time in 11th grade. Again, I am suspect to calling you a professional astronomer if you have never read this book. Although much of it has been disproved (and, as your abstract states, re-hypothesized), I would still assume that every astronomer would have this book as second knowledge.
Anyways, to get to the point of my post, the website that you link to has PDF and PS and DVI included.
Working at a hobby shop, selling telescopes, does NOT qualify you as a professional astronomer.
Please make a note of that in your future/. comments.
Don't worry about it. I got a supergeek job in NYC doing quantitative finance. Programming computers, running simulations, solving PDEs. In addition to the cool work projects, my work week is officially 35 hours (full-time), I get 4 weeks of paid vacation per year (starting on my first day), unlimited sick time.
However, like you, I can never actually take a a vacation. I *actually* put in ~60 hours per week and I'm too busy to get sick or take a vacation. The only thing that keeps me going is a shot at a hefty annual bonus. Some people in my department have earned bonuses of 100% of their salary.
Yes, that might comfort me since I recently got my *low* resolution model of the PowerBook.
However, 32MB of video ram isn't something that I can shake a stick at. It's the difference between a smooth Castle Wolfenstein experience and a not-quite-adequate Castle Wolfenstein experience.
Single hidden-layer linear feed forward Neural network. That's really all that we've ever needed for any recognition problem. The hard part is training the network and deciding which features to provide the network.
The number of input nodes depends upon the length of the feature vector, which varies from machine to machine (and with firmware upgrades). The output layer also varies from machine to machine, depending upon if we want a confidence measurement, and a few other factors. The hidden layer is dependent upon the training. for preprocessing, a PC uses a backprop/simulated annealing/GA combo. The genetic algorithm only changes the number of nodes in the hidden layer and picks from a small subset of the features, thus it isn't much of a GA. We figure, this way the GA can remove the temperature features without the marketing people bothering us about it. The simulated annealing is because the training is rerun on the PC whenever a new thumb is entered into the database, and a IMHO, every trainer should use simulated annealing to speed things up. Especially on low-end hardware.
Pretty much all biometric tools work this same way. We've reverse engineered some competitors, and found similar systems, so I'm not really giving away any trade secrets here.
I'm sorry to debunk you Ms. Angela of LANL, but your statement is nonsense.
I design software for biometric systems and although I don't know where they are installed at, the US Gov. is our largest client. *NO* current systems verify a third dimensional component. The neural network that IDs the print is fed many parameters. Amongst them is color (as you stated), thumbprint temperature,ambient and outdoor temperature (because the human extremity body-temperature is so dependent upon the environment), plus many more features from the actual 2-Dimensional image. There is no 3-D component.
You might argue that the angling of the scanning lasers adds a third dimensional component (a shadow) to the 2D image, but this is still something that could be duplicated given an image.
A very basic components analysis of the Neural Network will show that the thumb temperature is an ineffective means of classifieing the print, yet where I work, marketing insists that we continue to use this. That is why we have tried to increase the temperature importance by also including ambient temperatures, but mostly, the temperature is useless as a classification feature.
As far as taping a photocopy of somebody's fingerprint to the scanner this won;t work. Our scanners are color images, and the light from the photocopier has to come in at the same angle as the lasers. Using a pane of glass, a red light angled in the right direction, and a camera, we have been able to create photos that pass for fingerprints ~97% of the time. The percentage would be slightly increased if you kept the image in your pocket (body-heat) until placing it on the thumbprint scanner. This number approaches the number of false-negatives that you get with any thumbscanner.
Using biometric information creates a *real* problem for identity theft. Bruce Schneier points this out in his second book. If the advanced criminals can't reproduce your thumbprint, then they might as well intercept your biometric going from the scanner to the computer and reproduce that on all subsequent machines.
This is something that I will definitely opt out of in the future. Using a pseudo-random key generator on a cel-phone and having it transmit the key would be more accurate than a biometric.
Slashdot Mirror
More importantly though is your current situation. You don't need time to interview for other jobs, you just need more time, period. Once you take a few 'sick afternoons' to go interview or leave for a long lunch interview, you may realize just how easy it is to leave you job at a reasonable hour. I had a job that used to make us stay late and work long hours. I was very unproductive. One day, because i hated that job so much, i left early (granted early=6:30, 8:30-6:30). I started doing it every day. One day, as i was leaving my manager stopped me and said 'i noticed you've been leaving at 6:30' (he dare not say 'early'!). I told him, matter of factly, that tonight i was meeting a friend and had to get ready. I made no excuses for the other nights. After that he learned to have more respect for the time i was in the office. He knew that i would be the 1st to leave the office and he would make extra efforts to make sure that i would have everything i needed to finish my projects before 6:30. He knew i wouldn't stay one second later, unless it was an emergency.
My coworkers were jealous that I got to leave 'early' and they had to stay until 9-10pm every night. I told them it was their own doing and they could leave when they wanted. When bonus time came around, my billing was on target and my bonus reflected that. I got more done during 8:30-6:30 when i felt good about my job, than i ever got done working 8:30AM-10PM and hating it.
This is one of my great revelations. It was beginning of restructuring my social life, getting physically fit, starting bike racing, and generally getting my life in order.
Now that's irony. Take it from me.
I think that you fail to understand the point of my friend, the grammar nazi.
Remember, the new corporate scrutiny will only increase revenue for consulting firms, since the consulting firms are hired to "straighten out the books".
As far as "service&consulting being the last refuge of the Damned" goes... I don't think it applies to financial consulting. Only to software/engineering/research consulting.
These figures are off the top of my head, please correct me if wrong.
Consulting made up ~25% of IBM's 18 Billion in revenue last year. It's been a significant part of IBM for a long time and many credit the consulting arm of IBM for pulling IBM through the '90s. IBM is *much* more than a harddrive manufacturer or computer hardware company.
If you have a hard time understanding this, put it this way... IBM was the Microsoft of the 60s, 70s, and 80s. Back then, Computers were mostly for *accounting* and various services that you would hire consultants for. IBM took advantage of this and became big. I question Microsoft, because after ~10 years as a substantial Monopoly, they haven't successfully branched into another area and their largest competition is given away for free.
Anyways, I figured a consulting explanation was in order for anybody who wants to post a *WHY WOULD IBM BUY A CONSULTING FIRM?* comment.
BTW, does anybody know what happened to the name "Monday"? I thought PWC consulting was changing it's name to "Monday". Am I wrong?
Yes, I know that setting the date ahead and then back isn't really copyright infringement, but you're still getting an improved version of a product that you aren't supposed to have, without paying for it.
Doing this with software such as Quicktime is also rebellious, as it sends the word to Apple that *hey your software is worthwhile to me*. It states that *I don't succumb to predatory bundle-it-with-the-OS techniques* and most importantly *I support capitalism and the free market*.
Do you get to do that by playing with the calendar?
It doesn't take the irony nazi to point out the sweet, sweet irony in using the random NYT account generator to read the story.
This reply is offtopic, but I just wanted to help this story become the #1 all time story on the Slashdot Hall of fame list.
The quartz rendering is nice, but I havn't noticed many other improvements.
To respond to the second part of your comment... That's not my problem.
I'm a little unsure about Jurassic Park being on the list. Sure it's another good movie, but these are supposed to be the best Sci-Fi movies. I would have put Aliens or T2-Judgement day on the list in order to have 2 from the same series.
Don't forget that T9 technology has nothing to do with the layout of the keys. You can use T9 in addition to having a nicely laid out keypad and that would speed you up VERY much. The layout of a number pad is bad for entering letters. Enough said.
;-)
In NO way does Despair's trademark spoof the USPTO. It is a valid trademark for a valid company.
The fact that they applied for it and were granted it was SOP (Standard Operating Procedure). They then went on to joke about *patenting* the emoticon and charging people for every time it is used. This might have been the *spoof* part, but not the trademark itself. I wish that people would get this straight.
1. Stealing MP3s is illegal. Somebody will post an intelligently worded comment that will be modded to +5 based upon this premise. Most of the other comments that point this out will be modded -1 Troll.
2. RIAA Screws over artists and consumers. Like we didn't know this already. </sarcasm>
3. Kazaa is a file sharing network, even if it is a *spyware* file sharing network. If the RIAA feels it can make more money without the Kazaa Spynet, then it won't negotiate with Kazaa. Plain and simple. If the RIAA geniuses have an idea that they think will make bigger $$$ then they will do it. They will then sue Kazaa who needs illegal MP3s to make their network popular, yet will be sued if they use them. (napster all over again).
4. RIAA will come out with their own draconian file format and continue to blame illegal file sharing whenever the consumer chooses getting more *value* for their money (via trading MP3s) as opposed to getting less *value* for more money (restricted and encrypted non-MP3 formats). Personally, I blame capitalism for this consumer decision, not file sharing. I feel that I have a right, as a consumer, to choose the file format, distribution technique, and amount that I pay. Although this argument shouldn't be used to justify illegal copying of music, in a sense, one can assign a $$ value for breaking the law. I would surely pay a small amount of money for music just to know that I wasn't doing anything illegal. I would place even more value (i.e. pay more) if I knew that the artist was getting a significant portion of what I was paying.
A long time ago (~1996), I said that the recording industry will never be able to do away with an open audio format (at least as open as MP3 is). The reason why is because they will have to offer something with much more *percieved* value than what I can download for free and play on any music device. No more $14 CDs (a new CD was $14 at the time).
IBM
Red Hat (although it may be accounting tricks, they did claim profit last quater)
Tivo (I'm not sure about them, but insert a hardware vender here, Axis cameras if not Tivo)
O'reilly Publishers (some of their books fall under GPLish licenses)
I tried to cover all the bases too. Services, Services/Hardware, Services/OS, Consumer Hardware, and a book company.
Nobody is claiming that games/propriatary applications have to be GPL'd. I don't think they should be. The basics of the OS, and the tools of programming, OTOH, should be opened and free so that a maximum number of people can make them better and understand them.
On the topic of dividends:
MS doesn't need to pay dividends. They will continue to NOT pay dividends on the stock and the stockholders will be happy. Here's why:
1. MS earns revenues, pays taxes on revenues (although MS pays little or no taxes, but that's a another tale).
2. after-tax income would go to investors in the form of dividends
3.investors pay taxes on dividends.
What I desribe here is that the investors end up with income that has had taxes paid twice on it, which is rather watered down if you ask me. What I would rather have happen, and THIS is what MS is known for is the following:
1. MS earns revenues and pays taxes on it.
2. MS buys back its own stock
3. Stock price increases substantially
4. Investors have a capital gain, which they don't have to pay any taxes on until they sell their stock, and if they are smart, they will hardly pay any capital gains tax.
On the topic of a MS having STAGGERING Cash on hand:
This is acceptable and even expected. Look at balance sheets. MS has more cash-on-hand than Walmart, Exxon, and Ford combined. Is this a surprise? No. Walmart has thousands of chain stores accross the country. Exxon has billions of dollars worth of fuel pumping, refining, and distribution equipment. Ford has billions of dollars of factories, raw materials, and vehicles. If any of these companies had excess cash-on-hand, that means they wouldn't be making money off of it by building stores, or pipelines, or Vehicles. When you look at it this way, what does MS really have? They don't have much. They have one campus in Redmond Washington and a bunch of Intellectual property, which doesn't translate well to a financial statement.
The only thing that gives MS *any* investment value is this $40B. The same holds for any software company and this is the reason that so many of the "mind-share" dot-com companies went bankrupt. If a software company doesn't have cash, then it doesn't have much of anything, and nobody wants to invest in it. You can't add *nix guru's and Aeron Chairs to your end-of-year report. *nix gurus can always get jobs elsewhere and Aeron chairs are, well, a waste of money.
About catastrophe Hedging software:
A lot of people on Slashdot have been writing that this software is to protect MS against a DOJ "disaster", a marketing flop (X-box?) or other lawsuits. This couldn't be further from the truth. Catastrophe Hedging is a very active part of finance and I can assure you that they hedging against large interest rate changes, stock-market disasters, foreign currency melt-downs, and Enrons... just like any other Risk Analysis software does. This is basic quantitative finance.
To end this rant session, I would like to state that Quantitative finance is a very interesting and geeky field. It's the 100% math and computer field that is often overlooked by the geek community. Did you know that a large number of hedge funds use Linux primarily for their analysis computers? Did you know that PDE's and Stochastic Calculus runs rampant all throughout finance?
One nitpick, however...
Do you feel a 'since' or a 'sense'?It's been a while since I felt a sense of 'since', but to be honest, I haven't had a sense for 'since' since.
or
I'll give my 2 cents, since you seem to need sense.
I just wanted to give you a sense for the grammar nazi since he is not around.
Sorry. I've lost my sense since I haven't had much sleep lately. Please forgive me after you read this.
People who don't understand why you would/wouldn't log in as root are *extremely* unlikely to be playing around with 'rm', 'chmod', and 'mv'.
You would have a better argument saying something to the effect of "dragging an important system file into the trash" or moving/renaming an important file/folder.
I find it amazing how many people don't want to *login* to their computers.
They tell me, "I know that it's safer to log into my computer, but it's such a pain." --to which my usual reply is "You don't know that it's safer to log in."
First of all, have you ever heard of Occum's razor? I would have assumed that you knew of such things, beinging a professional astronomer and all. It states that all things being equal, the simplest explanation holds more weight than the complicated explanations. It is simpiler to go with the current theories rather than trying to figure out the unnecessary complications involved with a cyclic universe. The cyclic universe that you describe in the abstract is similar to something that I read in Brief History of Time in 11th grade. Again, I am suspect to calling you a professional astronomer if you have never read this book. Although much of it has been disproved (and, as your abstract states, re-hypothesized), I would still assume that every astronomer would have this book as second knowledge.
Anyways, to get to the point of my post, the website that you link to has PDF and PS and DVI included.
Working at a hobby shop, selling telescopes, does NOT qualify you as a professional astronomer.
Please make a note of that in your future /. comments.
However, like you, I can never actually take a a vacation. I *actually* put in ~60 hours per week and I'm too busy to get sick or take a vacation. The only thing that keeps me going is a shot at a hefty annual bonus. Some people in my department have earned bonuses of 100% of their salary.
However, 32MB of video ram isn't something that I can shake a stick at. It's the difference between a smooth Castle Wolfenstein experience and a not-quite-adequate Castle Wolfenstein experience.
Sigh.
We have some room for error for false-negatives, but I agree that those are annoying.
The number of input nodes depends upon the length of the feature vector, which varies from machine to machine (and with firmware upgrades). The output layer also varies from machine to machine, depending upon if we want a confidence measurement, and a few other factors. The hidden layer is dependent upon the training. for preprocessing, a PC uses a backprop/simulated annealing/GA combo. The genetic algorithm only changes the number of nodes in the hidden layer and picks from a small subset of the features, thus it isn't much of a GA. We figure, this way the GA can remove the temperature features without the marketing people bothering us about it. The simulated annealing is because the training is rerun on the PC whenever a new thumb is entered into the database, and a IMHO, every trainer should use simulated annealing to speed things up. Especially on low-end hardware.
Pretty much all biometric tools work this same way. We've reverse engineered some competitors, and found similar systems, so I'm not really giving away any trade secrets here.
I design software for biometric systems and although I don't know where they are installed at, the US Gov. is our largest client. *NO* current systems verify a third dimensional component. The neural network that IDs the print is fed many parameters. Amongst them is color (as you stated), thumbprint temperature,ambient and outdoor temperature (because the human extremity body-temperature is so dependent upon the environment), plus many more features from the actual 2-Dimensional image. There is no 3-D component.
You might argue that the angling of the scanning lasers adds a third dimensional component (a shadow) to the 2D image, but this is still something that could be duplicated given an image.
A very basic components analysis of the Neural Network will show that the thumb temperature is an ineffective means of classifieing the print, yet where I work, marketing insists that we continue to use this. That is why we have tried to increase the temperature importance by also including ambient temperatures, but mostly, the temperature is useless as a classification feature.
As far as taping a photocopy of somebody's fingerprint to the scanner this won;t work. Our scanners are color images, and the light from the photocopier has to come in at the same angle as the lasers. Using a pane of glass, a red light angled in the right direction, and a camera, we have been able to create photos that pass for fingerprints ~97% of the time. The percentage would be slightly increased if you kept the image in your pocket (body-heat) until placing it on the thumbprint scanner. This number approaches the number of false-negatives that you get with any thumbscanner.
Using biometric information creates a *real* problem for identity theft. Bruce Schneier points this out in his second book. If the advanced criminals can't reproduce your thumbprint, then they might as well intercept your biometric going from the scanner to the computer and reproduce that on all subsequent machines.
This is something that I will definitely opt out of in the future. Using a pseudo-random key generator on a cel-phone and having it transmit the key would be more accurate than a biometric.