Personnaly, I just can't wait for the concept of releases to go away. I download updates almost every day, so that the software installed on my computer is always changing anyway. New releases can be installed as updates, so why bother at all?
Why not update the.iso file every day for newcomers, and simply update continually from there on. Ok, I'm aware this would make it harder to write specs (e.g. "compatible with release X or Y"), but nothing prevents generating freezes if needed.
Didn't he look around and see that no professional communication was using "Voiceless velar fricatives"? Did he decide that all other writers in the entire English language were wrong, and he was being more communicative than the professionals? Or was he intentionally making communication difficult?
Not at all. It's just that "TeX" is not a T, and E and a X. It's tau, epsilon, and chi, and it is pronounced as in Modern Greek because it *is* greek.
As noted in Wikipedia, "TeX is the abbreviation of tau-epsilon-chi-nu-eta (techne), Greek for both "art" and "craft", which is also the source word of technical."
In some fields, it seems that scientists grossly under-use computers. I've seen megabytes of experimental data painfully being loaded into Excel to produce mediocre-looking graphs and generate trivial post-analysis.
With basic knowledge in some multi-purpose languages (be it Matlab, Python, Octave, or even basic scripting in Bash), it's simply amazing to see what can be done. I've seen 10-year backlogs of never-opened data files processed in an afternoon, with the results plotted on a few well-thougth graphs.
Ok I admit this is in fact a programming job, but you'll probably find it much easier, interesting and fun compared to your standard business/web programming gig. There's a dire need for people with basic computing skills in probably every science lab, though you probably won't make a fortune because this need is not widely recognized.
Right now it is almost impossible to purchase a house using internet based tools. Every housing market is controlled by the local realtors and they are VERY territorial. In Canada, ByTheOwner.Com is gaining lots of traction. They're not online realtors; it's a (paying) website letting individual sellers and buyers find each other, completely bypassing realtors. I'd say around 10 to 15% of all houses are sold through them in my region, which isn't that much, but they seem to have reached critical mass.
Without the 6% friction rate, and with a *much* better web site than traditional realtors', they're gaining market share quickly. They really are 10 years ahead of the realtors and I predict that the concept will simply wipe out realtors.
I'm not affiliated with them but I'm a satisfied customer. Their fees are rather steep but it includes the services of a professional photograph, which is why the pictures are so good in the first place.
In Bell's case, switching on and off the "Ignore Unencrypted Peers" option in Transmission gets me an order of magnitude variation on the overall download speed. Non-encrypted traffic seems to be capped at around 25 KB/s while I can reach saturation (around 300 KB/s) for swarms that are well-populated with encryption-using peers.
It's acutally pretty much what I did, and still do: I keep a "to read" file, and fill it with SF books I might like, mostly following comments I read on Slashdot. From this, I order maybe two or three deliveries per year from Amazon and the like, and I now have a pile of book to read, and I'm pretty confident that they'll be worth it.
That's the funny thing with light. In light's reference frame, time is stopped; for a photon, the moment it is emitted is the same as the moment when it's absorbed. So does a photon really travel between the source and your eye? Or is it more a kind of instantaneous (always in the photon reference frame) transfer of energy between the source and some molecule in your eye? Time only makes sense for us sub-lightspeed entities;)
... But we still have to wait for Inkscape 0.50 (Milestone 16) for our arrowheads to be the same color as the arrow lines. In the mean time, keep your arrows black, please.
The war destabilized oil prices, and until we get out of there, they won't remain stable for any long periods.
The rise in oil prices has multiple causes, but the main one is the fact that oil is priced in US dollars and this currency has devalued very rapidly while the intrinsic value of oil stayed more or less the same. Admittedly, the devaluation of the US dollar is linked to the Iraq war though, so there is definitely a link between oil prices and the war. It's just not mainly a geopolitical stability issue.
I think ant hive behavior is kind of anthropic - if hives were not optimized that way, ant hives wouldn't exist because all the ants would be dead. This is clearly the antropic principle at work.
The main difference between a joystick/pad and a mouse is not a question of ergonomics or resistance. The quantity being controlled by the mouse in a FPS is the angular position of the line of sight. Using a mouse, there is only a gain between the mouse position and the angular position of the LOS. Thus, the control is direct.
Using a joystick, pad, or keyboard, the player controls the rate of change (i.e. speed) of the LOS movement (and with non-analog controls, even that is fixed, the player only controls when to start and stop moving). Thus, there is an integrator in the control loop. An integrator is a low-pass filter, and it is impossible for a human to react as quickly with such an indirect control.
It is indeed possible to play using both schemes, but the gameplay just cannot be the same. It's just a question of choice. For a direct control scheme, if you don't want to wave your arm all around you, you need a way to interrupt the input while you reset the controller. This is what we do when we lift the mouse at the end of the mouse pad to put it back at the other end to continue the movement. This is much harder to implement in a console controller.
Ok, for the interested readers, here is the explanation of the AC's computation.
The expression:
(9 * (10^9) * 8 * (log(2) / log(10)))
is the number of decimal digits in 9 GB. 9*(10^9)*8 is the number of bits in 9 GB, and log(2)/log(10) is the factor to convert into the number of decimal digits (look up how to perform base changes when using logarithms). This is about Nd = 21.7 billion decimal digits.
Now, there is about the same number of sequences of 9 digits in the stream of decimal digits as the number of digits itself... in fact, there is only 8 less sequences than there are digits, so Nd-8 sequences, which is about Nd.
When picking a random sequence of 9 digits, the probability of it being a particular 9-digits ICQ number is 1 in 10^9. Correspondingly, the probability of it *not* being the particular ICQ number is ( 10^9 - 1 ) in 10^9. Thus, repeating the experiment Nd times, the probability of not hitting the ICQ number becomes ((10^9-1)/10^9)^Nd, so about 3.86e-10. Thus, the probability of finding the ICQ number is 1 minus 3.86e-10, very very close to 1.
I think the AC made an error in evaluating the probability of not hitting the ICQ number as (9/10)^9. This would be the probability of guessing an incorrect digit nine times in a row in an independant fashion, instead of guessing a 9-digits number wrongly a single time. But feel free to correct me if anyone thinks I'm wrong.
Either way, this shows that it's almost impossible for a given 9-digit ICQ number *not* to show up in the decimal stream representation of a 9 GB movie.
(...) one of the three magnets within its enclosing cryostat broke at a pressure of 20 atmospheres, in response to asymmetric forces applied during the test. Such forces are expected on occasion during normal operation of the LHC.
It seems to me that asymmetric forces was one of the thing being tested, or at least, that asymmetric forces were expected to occur, so at least, some people tought of that at CERN. The question is, were the asymmetric load in the design specs? If yes, then the designers overlooked or underestimated this point; if no, then there was a specs-writing problem, evidently. Then again, I know nothing in the field of large hadron colliders design...
Game are full of such feature-bugs. The strafejumping bug was finally added as a feature in subsequent versions of Quake, but other "bugs" (or limitations) in the physics model of Quake (and presumably other games) are used by skilled players.
For example, in Quake 3, the sampling of the position of the player in 3 dimensions is tied to the frame rate. When the player jumps, in the physics model, the player appears in successive, discrete position along a pre-computed parabolic trajectory. Only at key frame rates, a position sample appears at the peak of the parabola. Players running at these key frame rates are thus able to jump higher. Some custom maps feature areas that are accessible only to players knowing this trick.
There is also a Quake 3 mod (Defrag) which was created to facilitate finding map bugs (e.g. rebounds, which happen when you can fall with absolutely no x or y speed component on some surfaces, happens when the z velocity gets reversed). This leads to impressive "trick jumps" which do not violate the rules of the game (since they are done without modifying the game engine). A quick look in Google and you can find many trick-jumping videos based both on synchronisation and on physics model bugs.
A better combination, IMHO, is to use alt-sysrq-S/U/B in that order. S syncs drives; U unmounts them; and B reboots.
Seriously, I've use that many, many times, even with the most user-friendly Linux distros... If the UI is unresponsive, ctrl-alt-backspace (restarting X) does not work, and you cannot change virtual consoles, and you cannot connect to your machine via ssh, this is the only thing left to do.
Personally I found that Debian's problem is that by the time they've gotten a new release out the door, it is already hideously out of date.
I never understood this argument. I don't consider myself an expert in the various Linux distributions, especially not Debian. However, I settled on Debian for the exact opposite reason. I wanted, for my home computer, a bleeding-edge system that I wouldn't have to re-install ever because a "newer version" was released.
So I switched to Debian unstable, got packages almost as fast as in Gentoo and other bleeding-edge distros, and since then, never bothered re-installing anything from scratch. I just let my system evolve. Sometime this leads to broken bits, but I don't mind much, fixes are generally released fast enough for my tastes.
So... if you want a rock-solid server OS, get Debian stable. If you want a bleeding-edge, configurable OS, get Debian unstable, and you have testing inbetween for a more mainstream-type distro. I'm not saying that everybody should use Debian only (I myself use other distros quite often) -- just that the "out of date" argument is really getting old.
The situation is in fact ridiculous. You now have two choices to submit your tax report: by snail mail, which is free. Or electronically, which requires you to buy an accredited program in order to generate the e-forms, thus excluding any open source software.
Given the resources at the disposal of the Government, why couldn't they just build a website similar to the ones online tax programs use for data entry and obvious computations (adding columns, tax brackets, etc.). I don't care if the results are not automatically optimized, but at least, you could *fill* the tax forms online and submit the e-forms without the fees required by proprietary programs. IMHO, that should be a basic IRS service. Of course it would mean a well-rounded one-time investment, but the following years, only minor adjustment would be needed.
An open, electronic journal could work with a moderation system not unlike Slashdot, or even better, Everything2. User could get registered, and acquire reputation by writing articles having good reviews. Moderation systems work well for comments and E2 nodes; why not for scientific articles?
Citation and references would also offer a dizzying world of possibilities for moderating, ranking and following up on research. It would be trivial to automatically add to each article a list of articles that link back to it. Just think what Google or any search engine would do with such an open database (I know Google Scholar, but right now, it mostly indexes abstracts and cannot follow much in terms of references and citations).
I for one know I would read and write much more articles in such a system, and be much more active in general in my scientific community. Performing bibliography work when you can access on the spot every reference is so fast and effective, it's just a pity that right now, it litterally bleeds off by University's library.
I think the parent is actually correct. "Research" jobs are mostly development, and the horizon is most often quite short. I know companies where "research" refers to upgrades of existing products, deliverable in two months. Very few corporations nowadays do medium-to-long-term research. And if they do, I guess not everybody can be accepted in the select research team.
Looking toward academic research is probably a good thing if you are interested in fundamental or exploratory research. However, if you want a "job" there, and you're not a Professor, you'll have to be lucky, as I don't know of many academic labs with the budget for non-grad-student staff. (Then again I'm doing my PhD in Canada).
Thus I guess the best way to do research would be to contact a Professor whose research is of interest to you (most list their research topics on their web pages), and begin graduate studies under his direction. This may be a very drastic change in your life, as financing this project depends on your capacity to get a grant, unless you're very lucky/rich.
Once a grad student, you'll do research, but most of your work will be to learn (through courses, reading papers, etc.). I agree that your coding skills will probably be a boon on both you and your team. On average, I would say that research teams are doing a very poor job with data acquisition, processing, exploration, etc. Priority goes to expensive instruments, not to IT staff unless it's absolutely necessary, so that being able to code and perform IT tasks will boost your capacity to analyze experimental and simulation data.
I think most research teams underestimate the need for IT resources, and it is sad to see that funds are generally kept for a new shiny instrument that generates tons of data that get unused because of IT limitations, or to new reasearchers instead of IT staff to support the existing research staff... Hell there's probably a need for 2-3 people like you right now on my team of ten or so. However for now, I guess the easiest way to enter a research team is by the traditional way -- through grad studies.
That is actually quite true, in my experience too. Calculators are mandatory in school and college because they are the only computing appliances typically allowed during exams. However, in science/engineering, computers (and math software) are so ubiquitous that a calculator, graphing or not, is simply useless now. A good advice then: buy a good model, but a used one (e.g. on eBay), and keep your money for yourself.
There's an excellent open-source MATLAB clone called Octave. I've used it for a lot of real-world physics work in my lab. Worth checking out before you shell out for MATLAB.
While Octave is indeed excellent and a worthy replacement for Matlab in many contexts, it is not, in itself, a *graphing* solution. Octave's 'plot' command uses gnuplot. Thus, in order to use Octave as a graphing calculator, you would have to use, say, a standard Linux distro. It presumably wouldn't run on a 150$ handheld.
Thus, Matlab, Octave + Gnuplot, or Python + SciPy + MatPlotLib would all be great on a laptop, but alas cannot (yet) replace a graphing calculator in the submitter's context.
Marathon set the standard for me, too, and that is one of the reasons for which I'm not playing much, these days. Yes, I played a few interesing story-driven games after that, Half-Life probably being the best. But none would have me struggling in order to reach a terminal an actually read the rest of the story, which was in my opinion of SF litterature-grade.
(OK I was a teenager when I was playing Marathon, maybe I just didn't know much about SF litterature at that age; but then again, the Marathon and Halo stories are heavily influenced by SF bestsellers from Iain M. Banks, William Gibson, etc. The Marathon ship itself is very close to a Culture ship, while the struggle between the Leela and Durandal AI are not that far from the one between Wintermute and Necromancer.)
At the same time, the game introduced the grenade jump ("Frog Blast the Vent Core!"), two-triggered weapons, vertical aiming and much more, so it wasn't only story-driven: the action was incredible too.
So... why similar games aren't on the shelf these days? They're not economically feasible anymore? I can enjoy deathmatches alright, it's an entirely different game type, but I'm really longing for a compelling story.
So you first assume I meant speed of ligth in vacuum, and then infer from that that I know shit about physics and bitch about how I'm moderated. The problem is, as this is Slashdot and not an engineering department, I tought is wasn't necessary to write a white paper on transmission lines and waveguides as an introduction.
For the record, the whole point of my post was that the speed of light (be it in the vacuum or in a dielectric medium, or the propagation speed inside a waveguide) never was the limiting factor for communication bandwidth. It may cause delay problems on large distances (which isn't related), but I doubt that chipmakers are bitching about the slowness of c (then again I'll admit I know nothing about chip design). Switching (modulation) speed is what we're talking about and that's where photonics shine (bad pun accidental).
And as for the speed of light inside glass fiber, it may be much slower than 0.8 c as refraction indices can be quite high, and even slower again if you're talking about a specific waveguide mode and not a straight path in a glass slab.
Electrical wave guides, like for example coax cable, have wave velocities that are faster than light in glass, so they would logically be even better.
Err, actual wave speed never was a problem. Electromagnetic force already propagates at the speed of light in an electric conductor.
It's the modulation speed (e.g. how fast you can vary the signal inside the channel) that is much higher in photonic devices. In conductors, losses are very high for rapidly varying signals, and as you said it, microwave guides are much too large for chips. Evanescent fields are also a problem since they can spread very far from small guides.
This is even true when you light: for example, a standard coupler ("Y") for visible or IR wavelengths must normally be several centimeters long. However, so-called photonic-bandgap devices are solving this problem.
Slashdot Mirror
Personnaly, I just can't wait for the concept of releases to go away. I download updates almost every day, so that the software installed on my computer is always changing anyway. New releases can be installed as updates, so why bother at all?
Why not update the .iso file every day for newcomers, and simply update continually from there on. Ok, I'm aware this would make it harder to write specs (e.g. "compatible with release X or Y"), but nothing prevents generating freezes if needed.
Didn't he look around and see that no professional communication was using "Voiceless velar fricatives"? Did he decide that all other writers in the entire English language were wrong, and he was being more communicative than the professionals? Or was he intentionally making communication difficult?
Not at all. It's just that "TeX" is not a T, and E and a X. It's tau, epsilon, and chi, and it is pronounced as in Modern Greek because it *is* greek.
As noted in Wikipedia, "TeX is the abbreviation of tau-epsilon-chi-nu-eta (techne), Greek for both "art" and "craft", which is also the source word of technical."
In some fields, it seems that scientists grossly under-use computers. I've seen megabytes of experimental data painfully being loaded into Excel to produce mediocre-looking graphs and generate trivial post-analysis.
With basic knowledge in some multi-purpose languages (be it Matlab, Python, Octave, or even basic scripting in Bash), it's simply amazing to see what can be done. I've seen 10-year backlogs of never-opened data files processed in an afternoon, with the results plotted on a few well-thougth graphs.
Ok I admit this is in fact a programming job, but you'll probably find it much easier, interesting and fun compared to your standard business/web programming gig. There's a dire need for people with basic computing skills in probably every science lab, though you probably won't make a fortune because this need is not widely recognized.
Without the 6% friction rate, and with a *much* better web site than traditional realtors', they're gaining market share quickly. They really are 10 years ahead of the realtors and I predict that the concept will simply wipe out realtors.
I'm not affiliated with them but I'm a satisfied customer. Their fees are rather steep but it includes the services of a professional photograph, which is why the pictures are so good in the first place.
In Bell's case, switching on and off the "Ignore Unencrypted Peers" option in Transmission gets me an order of magnitude variation on the overall download speed. Non-encrypted traffic seems to be capped at around 25 KB/s while I can reach saturation (around 300 KB/s) for swarms that are well-populated with encryption-using peers.
It's acutally pretty much what I did, and still do: I keep a "to read" file, and fill it with SF books I might like, mostly following comments I read on Slashdot. From this, I order maybe two or three deliveries per year from Amazon and the like, and I now have a pile of book to read, and I'm pretty confident that they'll be worth it.
That's the funny thing with light. In light's reference frame, time is stopped; for a photon, the moment it is emitted is the same as the moment when it's absorbed. So does a photon really travel between the source and your eye? Or is it more a kind of instantaneous (always in the photon reference frame) transfer of energy between the source and some molecule in your eye? Time only makes sense for us sub-lightspeed entities ;)
Hey thanks, didn't even know this feature...
... But we still have to wait for Inkscape 0.50 (Milestone 16) for our arrowheads to be the same color as the arrow lines. In the mean time, keep your arrows black, please.
The rise in oil prices has multiple causes, but the main one is the fact that oil is priced in US dollars and this currency has devalued very rapidly while the intrinsic value of oil stayed more or less the same. Admittedly, the devaluation of the US dollar is linked to the Iraq war though, so there is definitely a link between oil prices and the war. It's just not mainly a geopolitical stability issue.
The main difference between a joystick/pad and a mouse is not a question of ergonomics or resistance. The quantity being controlled by the mouse in a FPS is the angular position of the line of sight. Using a mouse, there is only a gain between the mouse position and the angular position of the LOS. Thus, the control is direct.
Using a joystick, pad, or keyboard, the player controls the rate of change (i.e. speed) of the LOS movement (and with non-analog controls, even that is fixed, the player only controls when to start and stop moving). Thus, there is an integrator in the control loop. An integrator is a low-pass filter, and it is impossible for a human to react as quickly with such an indirect control.
It is indeed possible to play using both schemes, but the gameplay just cannot be the same. It's just a question of choice. For a direct control scheme, if you don't want to wave your arm all around you, you need a way to interrupt the input while you reset the controller. This is what we do when we lift the mouse at the end of the mouse pad to put it back at the other end to continue the movement. This is much harder to implement in a console controller.
Ok, for the interested readers, here is the explanation of the AC's computation.
The expression:
(9 * (10^9) * 8 * (log(2) / log(10)))
is the number of decimal digits in 9 GB. 9*(10^9)*8 is the number of bits in 9 GB, and log(2)/log(10) is the factor to convert into the number of decimal digits (look up how to perform base changes when using logarithms). This is about Nd = 21.7 billion decimal digits.
Now, there is about the same number of sequences of 9 digits in the stream of decimal digits as the number of digits itself... in fact, there is only 8 less sequences than there are digits, so Nd-8 sequences, which is about Nd.
When picking a random sequence of 9 digits, the probability of it being a particular 9-digits ICQ number is 1 in 10^9. Correspondingly, the probability of it *not* being the particular ICQ number is ( 10^9 - 1 ) in 10^9. Thus, repeating the experiment Nd times, the probability of not hitting the ICQ number becomes ((10^9-1)/10^9)^Nd, so about 3.86e-10. Thus, the probability of finding the ICQ number is 1 minus 3.86e-10, very very close to 1.
I think the AC made an error in evaluating the probability of not hitting the ICQ number as (9/10)^9. This would be the probability of guessing an incorrect digit nine times in a row in an independant fashion, instead of guessing a 9-digits number wrongly a single time. But feel free to correct me if anyone thinks I'm wrong.
Either way, this shows that it's almost impossible for a given 9-digit ICQ number *not* to show up in the decimal stream representation of a 9 GB movie.
(...) one of the three magnets within its enclosing cryostat broke at a pressure of 20 atmospheres, in response to asymmetric forces applied during the test. Such forces are expected on occasion during normal operation of the LHC.
It seems to me that asymmetric forces was one of the thing being tested, or at least, that asymmetric forces were expected to occur, so at least, some people tought of that at CERN. The question is, were the asymmetric load in the design specs? If yes, then the designers overlooked or underestimated this point; if no, then there was a specs-writing problem, evidently. Then again, I know nothing in the field of large hadron colliders design
Game are full of such feature-bugs. The strafejumping bug was finally added as a feature in subsequent versions of Quake, but other "bugs" (or limitations) in the physics model of Quake (and presumably other games) are used by skilled players.
For example, in Quake 3, the sampling of the position of the player in 3 dimensions is tied to the frame rate. When the player jumps, in the physics model, the player appears in successive, discrete position along a pre-computed parabolic trajectory. Only at key frame rates, a position sample appears at the peak of the parabola. Players running at these key frame rates are thus able to jump higher. Some custom maps feature areas that are accessible only to players knowing this trick.
There is also a Quake 3 mod (Defrag) which was created to facilitate finding map bugs (e.g. rebounds, which happen when you can fall with absolutely no x or y speed component on some surfaces, happens when the z velocity gets reversed). This leads to impressive "trick jumps" which do not violate the rules of the game (since they are done without modifying the game engine). A quick look in Google and you can find many trick-jumping videos based both on synchronisation and on physics model bugs.
A better combination, IMHO, is to use alt-sysrq-S/U/B in that order.
S syncs drives; U unmounts them; and B reboots.
Seriously, I've use that many, many times, even with the most user-friendly Linux distros... If the UI is unresponsive, ctrl-alt-backspace (restarting X) does not work, and you cannot change virtual consoles, and you cannot connect to your machine via ssh, this is the only thing left to do.
Personally I found that Debian's problem is that by the time they've gotten a new release out the door, it is already hideously out of date.
I never understood this argument. I don't consider myself an expert in the various Linux distributions, especially not Debian. However, I settled on Debian for the exact opposite reason. I wanted, for my home computer, a bleeding-edge system that I wouldn't have to re-install ever because a "newer version" was released.
So I switched to Debian unstable, got packages almost as fast as in Gentoo and other bleeding-edge distros, and since then, never bothered re-installing anything from scratch. I just let my system evolve. Sometime this leads to broken bits, but I don't mind much, fixes are generally released fast enough for my tastes.
So... if you want a rock-solid server OS, get Debian stable. If you want a bleeding-edge, configurable OS, get Debian unstable, and you have testing inbetween for a more mainstream-type distro. I'm not saying that everybody should use Debian only (I myself use other distros quite often) -- just that the "out of date" argument is really getting old.
The situation is in fact ridiculous. You now have two choices to submit your tax report: by snail mail, which is free. Or electronically, which requires you to buy an accredited program in order to generate the e-forms, thus excluding any open source software.
Given the resources at the disposal of the Government, why couldn't they just build a website similar to the ones online tax programs use for data entry and obvious computations (adding columns, tax brackets, etc.). I don't care if the results are not automatically optimized, but at least, you could *fill* the tax forms online and submit the e-forms without the fees required by proprietary programs. IMHO, that should be a basic IRS service. Of course it would mean a well-rounded one-time investment, but the following years, only minor adjustment would be needed.
Citation and references would also offer a dizzying world of possibilities for moderating, ranking and following up on research. It would be trivial to automatically add to each article a list of articles that link back to it. Just think what Google or any search engine would do with such an open database (I know Google Scholar, but right now, it mostly indexes abstracts and cannot follow much in terms of references and citations).
I for one know I would read and write much more articles in such a system, and be much more active in general in my scientific community. Performing bibliography work when you can access on the spot every reference is so fast and effective, it's just a pity that right now, it litterally bleeds off by University's library.
I think the parent is actually correct. "Research" jobs are mostly development, and the horizon is most often quite short. I know companies where "research" refers to upgrades of existing products, deliverable in two months. Very few corporations nowadays do medium-to-long-term research. And if they do, I guess not everybody can be accepted in the select research team.
Looking toward academic research is probably a good thing if you are interested in fundamental or exploratory research. However, if you want a "job" there, and you're not a Professor, you'll have to be lucky, as I don't know of many academic labs with the budget for non-grad-student staff. (Then again I'm doing my PhD in Canada).
Thus I guess the best way to do research would be to contact a Professor whose research is of interest to you (most list their research topics on their web pages), and begin graduate studies under his direction. This may be a very drastic change in your life, as financing this project depends on your capacity to get a grant, unless you're very lucky/rich.
Once a grad student, you'll do research, but most of your work will be to learn (through courses, reading papers, etc.). I agree that your coding skills will probably be a boon on both you and your team. On average, I would say that research teams are doing a very poor job with data acquisition, processing, exploration, etc. Priority goes to expensive instruments, not to IT staff unless it's absolutely necessary, so that being able to code and perform IT tasks will boost your capacity to analyze experimental and simulation data.
I think most research teams underestimate the need for IT resources, and it is sad to see that funds are generally kept for a new shiny instrument that generates tons of data that get unused because of IT limitations, or to new reasearchers instead of IT staff to support the existing research staff... Hell there's probably a need for 2-3 people like you right now on my team of ten or so. However for now, I guess the easiest way to enter a research team is by the traditional way -- through grad studies.
That is actually quite true, in my experience too. Calculators are mandatory in school and college because they are the only computing appliances typically allowed during exams. However, in science/engineering, computers (and math software) are so ubiquitous that a calculator, graphing or not, is simply useless now. A good advice then: buy a good model, but a used one (e.g. on eBay), and keep your money for yourself.
While Octave is indeed excellent and a worthy replacement for Matlab in many contexts, it is not, in itself, a *graphing* solution. Octave's 'plot' command uses gnuplot. Thus, in order to use Octave as a graphing calculator, you would have to use, say, a standard Linux distro. It presumably wouldn't run on a 150$ handheld.
Thus, Matlab, Octave + Gnuplot, or Python + SciPy + MatPlotLib would all be great on a laptop, but alas cannot (yet) replace a graphing calculator in the submitter's context.
Amen to that.
Marathon set the standard for me, too, and that is one of the reasons for which I'm not playing much, these days. Yes, I played a few interesing story-driven games after that, Half-Life probably being the best. But none would have me struggling in order to reach a terminal an actually read the rest of the story, which was in my opinion of SF litterature-grade.
(OK I was a teenager when I was playing Marathon, maybe I just didn't know much about SF litterature at that age; but then again, the Marathon and Halo stories are heavily influenced by SF bestsellers from Iain M. Banks, William Gibson, etc. The Marathon ship itself is very close to a Culture ship, while the struggle between the Leela and Durandal AI are not that far from the one between Wintermute and Necromancer.)
At the same time, the game introduced the grenade jump ("Frog Blast the Vent Core!"), two-triggered weapons, vertical aiming and much more, so it wasn't only story-driven: the action was incredible too.
So... why similar games aren't on the shelf these days? They're not economically feasible anymore? I can enjoy deathmatches alright, it's an entirely different game type, but I'm really longing for a compelling story.
So you first assume I meant speed of ligth in vacuum, and then infer from that that I know shit about physics and bitch about how I'm moderated. The problem is, as this is Slashdot and not an engineering department, I tought is wasn't necessary to write a white paper on transmission lines and waveguides as an introduction.
For the record, the whole point of my post was that the speed of light (be it in the vacuum or in a dielectric medium, or the propagation speed inside a waveguide) never was the limiting factor for communication bandwidth. It may cause delay problems on large distances (which isn't related), but I doubt that chipmakers are bitching about the slowness of c (then again I'll admit I know nothing about chip design). Switching (modulation) speed is what we're talking about and that's where photonics shine (bad pun accidental).
And as for the speed of light inside glass fiber, it may be much slower than 0.8 c as refraction indices can be quite high, and even slower again if you're talking about a specific waveguide mode and not a straight path in a glass slab.
Err, actual wave speed never was a problem. Electromagnetic force already propagates at the speed of light in an electric conductor.
It's the modulation speed (e.g. how fast you can vary the signal inside the channel) that is much higher in photonic devices. In conductors, losses are very high for rapidly varying signals, and as you said it, microwave guides are much too large for chips. Evanescent fields are also a problem since they can spread very far from small guides.
This is even true when you light: for example, a standard coupler ("Y") for visible or IR wavelengths must normally be several centimeters long. However, so-called photonic-bandgap devices are solving this problem.