OK, so we have some sensors (from the article) picking up the movement of the vocal cords. Great. What you have there, my friends is fundamental frequency. Not speech. You also need the formants.
You could get (by picking up other movements in the head) a synthetic model of what the speaker is doing (raising the tongue in back, lowering it in front, opening the nasal passages) and use that to build a filter model to synthesize the speech, but such models sound like crap.
I'd love something like this to work. But it doesn't. And it probably never will. Your flesh is noisy. You move a lot of stuff, and that generates/requires voltage. To do a really good job of this, you need two things: 1. A lot of pickups, which would HAVE to be invasive, not filtered through the skin, which distorts signals, and 2. REALLY good synthesis models understanding the attached speaker's flesh and tubes.
The first would be unpleasant to have installed (Bone cement, anyone? Ever see a monkey or cat wired in this way?) and the second hasn't yet been written, after years of trying. (They've gotten much better, but they still don't sound like people.)
Certs may get you nothing in step two, three, etc of the hiring process, but (in many companies) they help you in step ONE.
Here at my company, I get resumes and check them out and say, this person, yes, that person, no, and the others in my group do the same, and whoever someone REALLY wants to meet, or who most in the group kinda want to bring in get brought in.
We're smart, know the field, know what certs show and don't, etc.
But we're not stage one. We're stage TWO. Where did those resumes come from in the first place? Who went out on Monster and other places and pulled resumes to show us? Who screened the resumes he/she got sent due to a posting?
Screening/First Selection is stage one. Certs are searchable as key terms. They get you placed above another person with equivilent qualifications in the mind of HR.
That's where you want them. If you have experience, and have a lot of buzzwords on your resume which can be searched for, you don't need certs for stage one. But they won't hurt.
And that gets you to stage two. Now, you might not make it past stage two. But your chance of making it past stage two are ZERO if you don't get grabbed in stage one.
Hence certs.
That said, I believe that certs can HURT you in stage two. Some of us think some certs are crap, and will actually diminsh you in our estimation. So for THAT reason, get good certs, if you go that route.
Read _Sweaty Palms_ by H. Anthony Medley. It's a great book on interviewing and the job application process.
Re:Convergence of Laptop parts and desktop
on
Pentium M Goes SFF
·
· Score: 1
I agree that they're less reliable (and we're getting far from the subject) but I suspect that there's good reason for their unreliability compared to desktops, and that if you have mobile CPUs in desktop machines, you won't have the same trouble (or will, but to a much lesser degree).
One -- the hardware in them changes constantly. Every year is a major redesign, as makers try to cram more into less space/weight.
Two -- They're more complicated, integrating non-standard mice and such as well as that screen (with not completely standard controller hardware, tweaked for the individual display).
Three -- they get moved around and dinged and dropped and such.
None of that has to be the case for desktops. Motherboard innovation could proceed at the usual desktop pace (slower than laptops) and case designs would be less of a factor, as we would settle on a brickish size, and not have to worry about weight.
You'd see some of the same problems, but certainly not all.
My bigger worry is that it'll cost me more to make a high-voltage server.
Convergence of Laptop parts and desktop
on
Pentium M Goes SFF
·
· Score: 1
If laptop parts come down some in price (more than this), and folks are willing to pay the difference to have quiet, smaller, cheaper to run machines, we could see an interesting trend over the next couple of years.
Laptop and desktop parts could converge, leaving what we currently think of as desktop parts (and their like) in the server realm. Meaning we might have a little burp in Moore's law (that the average macine bought in 2006 will be SLOWER than the average one bought in 2005) and the number of hefty CPUs sold would drop, so that the prices of them (therefore) would go up (or fail to come down as we have come to expect).
It'd be good for America's energy bill, but bad for gamers, and those of us who make servers at home with commodity parts.
I must say I have been horrified by my experience with a Primera Bravo box. Not because it's bad -- it's really great -- but because there are no linux drivers, and the Windows methods are absurdly awful.
To burn a disk, you go into a GUI and mockup (or just load) an image to print on the disk.
Then, you print it to file -- something.prn
Then you go into another GUI and set up the task, picking an ISO image, and the image file you just made, click here, click there, then burn.
That works just fine for 40 of the same disk, but if you want a different image on each (different date or different text, or the ISO filesize) you need to make each change manually (or with tags) and then print to file and then set up each task.
Man.
In unix/linux, or with command-line tools for windows, even, that would be:
Done. You'd be able to do everything this guy wants and more with 10 seconds of typing. You'd be able to automate processes. And it's not hard. Primera's been selling this stuff for years, and yet, no Linux support, and no command-line support.
If this had Linux support, or even DOS command-lines, I'd recommend it to everyone I meet. As is, it's an anchor.
It's not something over nothing. If someone has power over you that he should not have, but decides not to use it for a while, does that mean he should be able to keep it?
There's always the threat of not abuse, but use, which will be hanging over our heads. We have less rights as Americans while this law is in place.
I remember sweaty, frightened, forty-year old guys in suits (really bad polyester suits) who were trying to get into microcomputer jobs when I was just starting out as a professional.
The were mainframe people, and mainframes were drying up, at the time, and they knew nothing about microcomputers. They had been doing the same thing for years, and they didn't know what to do. They looked like a deer in headlights.
Interviewing them, they kept trying to use mainframe concepts to answer questions about microcomputers. They were... not a good fit. I don't know what happened to those people -- we stopped seeing them after a few years.
The VB folks seem like the same sort of problem. It's an object lesson on not getting tightly bounc to just one thing.
I'm a pointy haired boss. Or rather, I'm in the hiring process for all our tech hires.
Open Source projects are an INSTANT door opener. You have work on one in on your resume, you get an interview. Period.
That's not because this is some corporate give-away, but because we're techies, here. We want to talk to you, to find out how that went, what you did, etc. We're curious. And that means you get the interview. Most of the resumes we see don't get an interview, so you're already in the running for the job, unlike the guy with similar resume who didn't get called.
Plus the drive to do this sort of thing means you're not a slug. We get a lot of slugs with VERY good resumes. They waste our time. I've never had a slug come in who had OSS authorship.
There are no lines. The layers of X above the physical layer broke it into pixels. You need to tell your vector display to draw a line. What's the line? Who knows, as all I have is a bunch of pixels because the layer above broke it into a bunch of pixels.
For a vector display to work, you have to have lines, not pixels, and an approximation of a bunch of pixels on a vector display looks awful.
So, no, it could not "work quite nicely." That's the problem with X. It's all (as I understand it) bitmaps below application/library layer.
Applications (using libraries and toolkits) Xserver Drivers Physical hardware
If you didn't have a bitmap based Xserver system, things wouldn't get broken down into bits until after you hit the driver (going down). That means that communications over a networked representation of an X screen (via a socket to your X server from anthoer box running a client) or via VNC wouldn't have to be mostly bitmaps. They'd be drawn, primarily, and would be much tighter than all but the smallest bitmaps, and they'd scale.
Meaning that if you got a really good monitor, say one that does 10240x7680, you could use more dots to depict the things you see on the screen, thus improving the image, instead of just making them all smaller with the same number (and in our 100x bigger screen example, so small they'd be impossible to make out).
You raise in interesting point, though. If you get away from bitmapped systems at the X layer, if you DID have a vector display, your driver could feasibly NEVER break into a bitmap. As is, running X on a vector graphics system would require you to make jagged little lines, and spoil all the wonderfullness of your vector display.
So X ought to lose its bitmapped nature and have a scalable system at the X layer above the drivers.
Does anyone who's read more on this, and knows what each of these ideas mentioned in the articles corresponds to have an clue to whether this is being moved toward (as a replacement for X) or are we stuck with a friggen bitmap for another decade?
I' reading these posts, and I'm confused, probably because they're addressing different issues than I'm focused on when I think of X. And because I don't know enough about X.
I think the problem with X is not features, and libraries on top of it, but rather that the basic core concept in an X display is a bitmap.
The problems with this are: slow communications, and lack of scalability on different displays. The classic cure for this was display Postscript, which had problems of A. Copyright, B. Bloat, C. Large blocks files of code to do small things, D. Arcane syntax.
There has to be a better way. But what I'm seeing here is all applications and libraries for use by applications on top of the bitmap based rendering. There are some things mentioned which I recall being replacements for this engine, but certainly Enlightentment DR17 is all on top of the X bitmap system, right?
Any movement on chucking that in favor of a bitmap independant system?
Read everything Tom DeMarco has ever written. Read it now. Run don't walk to Borders (or wherever) to see what they have, then Amazon (or whatever) for the rest.
The Deadline is silly, but it's a good read and has excellent information and might be the first one you read. Peopleware is the most important book. Read Slack last, as it's least connected to specific software projects, and more on management, in general.
If this doen't sound like the Bible Code, nothing does. Seek in a large blob of random crap and you shall find -- every friggin' time.
The key is not to look for things you don't want to find, like "nothing is happening today" -- cause you'd find that, too, and then you'd know it was a basketful of hooey.
I don't mean to slam Google, but just to go ahead and state the obvious:
What a great way to influence a project: pay for it.
Google will really be able to get any pet idea that they have at least brought up as a part of the project.
This is a very cheap way of touching millions of people. A smart, patient and friendly company should be able to find ways to get their agenda helped, even when their employee is generally remaining "independant".
And free advertising: BGoodger@google.com at the bottom of every communication? Though I suppose it'll be something more like BGoodger@gmail.com
I interview a lot of job candidates. I think Joel's advice is good -- very good -- but it isn't all there. That is: Blow off the CS, or rather, don't major in CS.
If you want to get a technical degree, don't go to a real college. Go to a trade school masquerading as one. In Chicago, we have DeVry, IIT, and to some extent (depending on your program) DePaul.
But I don't advise that really. Go to a real college with a real collegiate program, major in physical sciences, social sciences, the humanities, or math COMBINED with a bunch of electives in the other disciplines and you'll learn something CS doesn't typically teach you: To have a broad base of knowledge about the world that your computing can connect to.
He's right: Take C, if you want to do hard-core programming. I'd go further -- take (in order) Java, then C, then assembler. Don't let them make you take COBOL!!! (DePaul was doing that for a while, as was DeVry.)
And take a data structures class, if one of those above wasn't coupled with one.
If taking classes in compilers, operating systems, AI, etc. is important to you, consider doing it as a master's degree. Go to those tech schools, if you need to. Also, you can generally sit-in on all those classes in college, while doing your real work in sciences and liberal arts. Do any of the work you feel like and get the basic ideas, so that you can read all about it, or take a master's program class in it if you love it.
But use college (which is often much more expensive than those tech schools) for what it's been designed to do for over the past 700 or so years: let it turn you into a thinking, logical, well-rounded person.
That's not to say you shouldn't be doing computer stuff in college. But if you work on free software projects in school, you'll learn what the other guys learned in class, and you'll have killer resume line-items after graduation. As in: "Worked on open source project froboz 3.1 to 3.4 as part of a team -- coded parts of bad block collection, memory coalescing, garbage collection, blah, blah, blah."
GPA means a lot. Experience means much more. But projects!?!? Man, can you get your foot in the door with projects! If it's something I as an employer actually use? And lo, I find your name in the docs? You're hired, big fella.
I can see the NAFTA trade angle. But a highway isn't the solution, not without a lot of cost on individual self-propelled transport containers (ie. TRUCKS) . The solution is clearly rail. Intermodal transportation (http://www.robl.w1.com/Transport/intermod.htm) is vastly cheaper than the archaic system these guys are suggesting. Containers come on trucks, then go on via train. When they get near an urban center for delivery, they get put on a truck body.
Much less labor, much less fuel consumption. Much less cost for individual carrier equipment. (Can someone else comment on the cost of rail vs. highway maintenance?)
If this is a way to make NAFTA better for everyone, they need to scrap the highway (or at least scale it back to very little) and run rails. If it's a way to generate tariffs on transport, well, rails do that, too.
But they wouldn't need 175 billion dollars for it. If they want to spend that kind of money, they should think about running rail lines through Texas (using some of the rails already there), building over and underpasses for existing rails in and around cities all over the country, running lines around cities to avoid marshalling yards (with their speed restrictions) and building efficient Intermodal systems in smaller towns (there are already such systems in the big ones).
But that would just mean investing in a rail company instead of press-releasing and creating a whole new way of thinking about roads, etc.
I bought a pair of the Plane Quiet headphones and am VERY VERY happy with them. But only on planes.
The sound they're good at cancelling is engine noies and other fairly low frequency stuff. They damp voices, but not by that much. If you wear them in a conversation, you hear the other folk -- they just sound low-fi.
My hugely favorite thing for damping noise and actually getting work done is called a DOOR. If you don't have one, ask your boss for one. If he's smart and able, he'll get one for you. Your productivity will go up.
As for using music to dampen, see Tom DeMarco's book, _Peopleware_. In it he runs tests with silence, office noise and music. The music ain't much better than the office noise for concentration. (page 78)
"New management at Canopy . . . may push [SCO] to try and settle."
"settle"? Not a chance that can happen. Drop the case is really the only thing that could happen. Why would anyone pay them a dollar? (Unless the settlement, is "we'll drop the case if you agree not to sue US.")
He's only sued people with enough money to defend themselves. I suspect he'll keep lawsuits open until SCO goes into bankruptcy, for fear that SCO will be countersued for legal fees.
Slashdot Mirror
I just now read that article she wrote. I think it's the first thing by her I've ever read. Someone copied it into a comment on slashdot.
She really reminds me of the gossip columnists in a Howard Hawks comedy of the 30's or 40's. It's weird how she takes you back.
Those were fun films, but I had assumed the somewhat stock characters were a parodies. I guess there ARE people who talk/write like that.
But I'd expect her to write for a paper called _The Tattler_.
Because I get paged at 10pm when the damn Windows servers crap out.
Because I need our PC guy to do stuff and instead he spends all day fighting spyware.
Because I have to walk users through web pages and they can't find the window they opened this or that up in (ie. they don't have tabs).
Because when our LAN guy goes out of town, they turn to ME to manage their exchange server. And guess what? It's unreliable.
So, as a resident techie, there is no "they use this, you use that, what's the big deal?" There is only "WE USE this." We need to stop.
OK, so we have some sensors (from the article) picking up the movement
of the vocal cords. Great. What you have there, my friends is
fundamental frequency. Not speech. You also need the formants.
You could get (by picking up other movements in the head) a synthetic
model of what the speaker is doing (raising the tongue in back,
lowering it in front, opening the nasal passages) and use that to
build a filter model to synthesize the speech, but such models sound
like crap.
I'd love something like this to work. But it doesn't. And it
probably never will. Your flesh is noisy. You move a lot of stuff,
and that generates/requires voltage. To do a really good job of this,
you need two things: 1. A lot of pickups, which would HAVE to be
invasive, not filtered through the skin, which distorts signals, and
2. REALLY good synthesis models understanding the attached speaker's
flesh and tubes.
The first would be unpleasant to have installed (Bone cement, anyone?
Ever see a monkey or cat wired in this way?) and the second hasn't yet
been written, after years of trying. (They've gotten much better, but
they still don't sound like people.)
Nothing could be more useless than an MSCE.
Certs may get you nothing in step two, three, etc of the hiring process, but (in many companies) they help you in step ONE.
Here at my company, I get resumes and check them out and say, this person, yes, that person, no, and the others in my group do the same, and whoever someone REALLY wants to meet, or who most in the group kinda want to bring in get brought in.
We're smart, know the field, know what certs show and don't, etc.
But we're not stage one. We're stage TWO. Where did those resumes come from in the first place? Who went out on Monster and other places and pulled resumes to show us? Who screened the resumes he/she got sent due to a posting?
Screening/First Selection is stage one. Certs are searchable as key terms. They get you placed above another person with equivilent qualifications in the mind of HR.
That's where you want them. If you have experience, and have a lot of buzzwords on your resume which can be searched for, you don't need certs for stage one. But they won't hurt.
And that gets you to stage two. Now, you might not make it past stage two. But your chance of making it past stage two are ZERO if you don't get grabbed in stage one.
Hence certs.
That said, I believe that certs can HURT you in stage two. Some of us think some certs are crap, and will actually diminsh you in our estimation. So for THAT reason, get good certs, if you go that route.
Read _Sweaty Palms_ by H. Anthony Medley. It's a great book on interviewing and the job application process.
I agree that they're less reliable (and we're getting far from the subject) but I suspect that there's good reason for their unreliability compared to desktops, and that if you have mobile CPUs in desktop machines, you won't have the same trouble (or will, but to a much lesser degree).
One -- the hardware in them changes constantly. Every year is a major redesign, as makers try to cram more into less space/weight.
Two -- They're more complicated, integrating non-standard mice and such as well as that screen (with not completely standard controller hardware, tweaked for the individual display).
Three -- they get moved around and dinged and dropped and such.
None of that has to be the case for desktops. Motherboard innovation could proceed at the usual desktop pace (slower than laptops) and case designs would be less of a factor, as we would settle on a brickish size, and not have to worry about weight.
You'd see some of the same problems, but certainly not all.
My bigger worry is that it'll cost me more to make a high-voltage server.
If laptop parts come down some in price (more than this), and folks are willing to pay the difference to have quiet, smaller, cheaper to run machines, we could see an interesting trend over the next couple of years.
Laptop and desktop parts could converge, leaving what we currently think of as desktop parts (and their like) in the server realm. Meaning we might have a little burp in Moore's law (that the average macine bought in 2006 will be SLOWER than the average one bought in 2005) and the number of hefty CPUs sold would drop, so that the prices of them (therefore) would go up (or fail to come down as we have come to expect).
It'd be good for America's energy bill, but bad for gamers, and those of us who make servers at home with commodity parts.
Agreed. I feel it, too.
There will be apps, but not "All Apps", and certainly not word processors/spreadsheets.
But those apps may surprise and horrify us. Let's hope so, anyhow.
I must say I have been horrified by my experience with a Primera Bravo box. Not because it's bad -- it's really great -- but because there are no linux drivers, and the Windows methods are absurdly awful.
To burn a disk, you go into a GUI and mockup (or just load) an image to print on the disk.
Then, you print it to file -- something.prn
Then you go into another GUI and set up the task, picking an ISO image, and the image file you just made, click here, click there, then burn.
That works just fine for 40 of the same disk, but if you want a different image on each (different date or different text, or the ISO filesize) you need to make each change manually (or with tags) and then print to file and then set up each task.
Man.
In unix/linux, or with command-line tools for windows, even, that would be:
create_postscript_with_substitutions [inputs] > printerfile.prn
burn_image_and_print isofile printerfile.prn
Done. You'd be able to do everything this guy wants and more with 10 seconds of typing. You'd be able to automate processes. And it's not hard. Primera's been selling this stuff for years, and yet, no Linux support, and no command-line support.
If this had Linux support, or even DOS command-lines, I'd recommend it to everyone I meet. As is, it's an anchor.
It's not something over nothing. If someone has power over you that he should not have, but decides not to use it for a while, does that mean he should be able to keep it?
There's always the threat of not abuse, but use, which will be hanging over our heads. We have less rights as Americans while this law is in place.
It's gotta go.
I remember sweaty, frightened, forty-year old guys in suits (really bad polyester suits) who were trying to get into microcomputer jobs when I was just starting out as a professional.
The were mainframe people, and mainframes were drying up, at the time, and they knew nothing about microcomputers. They had been doing the same thing for years, and they didn't know what to do. They looked like a deer in headlights.
Interviewing them, they kept trying to use mainframe concepts to answer questions about microcomputers. They were... not a good fit. I don't know what happened to those people -- we stopped seeing them after a few years.
The VB folks seem like the same sort of problem. It's an object lesson on not getting tightly bounc to just one thing.
I'm a pointy haired boss. Or rather, I'm in the hiring process for all our tech hires.
Open Source projects are an INSTANT door opener. You have work on one in on your resume, you get an interview. Period.
That's not because this is some corporate give-away, but because we're techies, here. We want to talk to you, to find out how that went, what you did, etc. We're curious. And that means you get the interview. Most of the resumes we see don't get an interview, so you're already in the running for the job, unlike the guy with similar resume who didn't get called.
Plus the drive to do this sort of thing means you're not a slug. We get a lot of slugs with VERY good resumes. They waste our time. I've never had a slug come in who had OSS authorship.
No.
There are no lines. The layers of X above the physical layer broke it into pixels. You need to tell your vector display to draw a line. What's the line? Who knows, as all I have is a bunch of pixels because the layer above broke it into a bunch of pixels.
For a vector display to work, you have to have lines, not pixels, and an approximation of a bunch of pixels on a vector display looks awful.
So, no, it could not "work quite nicely." That's the problem with X. It's all (as I understand it) bitmaps below application/library layer.
Sigh.
Wrong layer.
As an example of the layers of X:
Applications (using libraries and toolkits)
Xserver
Drivers
Physical hardware
If you didn't have a bitmap based Xserver system, things wouldn't get broken down into bits until after you hit the driver (going down). That means that communications over a networked representation of an X screen (via a socket to your X server from anthoer box running a client) or via VNC wouldn't have to be mostly bitmaps. They'd be drawn, primarily, and would be much tighter than all but the smallest bitmaps, and they'd scale.
Meaning that if you got a really good monitor, say one that does 10240x7680, you could use more dots to depict the things you see on the screen, thus improving the image, instead of just making them all smaller with the same number (and in our 100x bigger screen example, so small they'd be impossible to make out).
You raise in interesting point, though. If you get away from bitmapped systems at the X layer, if you DID have a vector display, your driver could feasibly NEVER break into a bitmap. As is, running X on a vector graphics system would require you to make jagged little lines, and spoil all the wonderfullness of your vector display.
So X ought to lose its bitmapped nature and have a scalable system at the X layer above the drivers.
Does anyone who's read more on this, and knows what each of these ideas mentioned in the articles corresponds to have an clue to whether this is being moved toward (as a replacement for X) or are we stuck with a friggen bitmap for another decade?
I' reading these posts, and I'm confused, probably because they're addressing different issues than I'm focused on when I think of X. And because I don't know enough about X.
I think the problem with X is not features, and libraries on top of it, but rather that the basic core concept in an X display is a bitmap.
The problems with this are: slow communications, and lack of scalability on different displays. The classic cure for this was display Postscript, which had problems of A. Copyright, B. Bloat, C. Large blocks files of code to do small things, D. Arcane syntax.
There has to be a better way. But what I'm seeing here is all applications and libraries for use by applications on top of the bitmap based rendering. There are some things mentioned which I recall being replacements for this engine, but certainly Enlightentment DR17 is all on top of the X bitmap system, right?
Any movement on chucking that in favor of a bitmap independant system?
Does he already know that GPS doesn't work indoors, and can be disabled with tinfoil?
"I didn't stalk her again. I was at home all day. It must have been some one else!"
can translate to
"I put some tinfoil on my GPS receiver and hunted her down just like you told me not to! HA HA!"
The Deadline is silly, but it's a good read and has excellent information and might be the first one you read. Peopleware is the most important book. Read Slack last, as it's least connected to specific software projects, and more on management, in general.
If this doen't sound like the Bible Code, nothing does. Seek in a large blob of random crap and you shall find -- every friggin' time.
The key is not to look for things you don't want to find, like "nothing is happening today" -- cause you'd find that, too, and then you'd know it was a basketful of hooey.
Sure it is. What's this country coming to? Free speech keeps us free. Are we afraid of what they're saying?
Big strong us.
There will always be places for coded messages, if that's the fear. If it's about propaganda, we need dissent. Without dissent we grow very weak.
I don't mean to slam Google, but just to go ahead and state the obvious:
What a great way to influence a project: pay for it.
Google will really be able to get any pet idea that they have at least brought up as a part of the project.
This is a very cheap way of touching millions of people. A smart, patient and friendly company should be able to find ways to get their agenda helped, even when their employee is generally remaining "independant".
And free advertising: BGoodger@google.com at the bottom of every communication? Though I suppose it'll be something more like BGoodger@gmail.com
This should be happening much more than it does.
I know it's rendered, but is that supposed to be liquid water?
I'm behind the curve here, and this is off topic, but "whoa!"
Liquid water implies a very tight temperature range. Is Titan that habitable?
I interview a lot of job candidates. I think Joel's advice is good -- very good -- but it isn't all there. That is: Blow off the CS, or rather, don't major in CS.
If you want to get a technical degree, don't go to a real college. Go to a trade school masquerading as one. In Chicago, we have DeVry, IIT, and to some extent (depending on your program) DePaul.
But I don't advise that really. Go to a real college with a real collegiate program, major in physical sciences, social sciences, the humanities, or math COMBINED with a bunch of electives in the other disciplines and you'll learn something CS doesn't typically teach you: To have a broad base of knowledge about the world that your computing can connect to.
He's right: Take C, if you want to do hard-core programming. I'd go further -- take (in order) Java, then C, then assembler. Don't let them make you take COBOL!!! (DePaul was doing that for a while, as was DeVry.)
And take a data structures class, if one of those above wasn't coupled with one.
If taking classes in compilers, operating systems, AI, etc. is important to you, consider doing it as a master's degree. Go to those tech schools, if you need to. Also, you can generally sit-in on all those classes in college, while doing your real work in sciences and liberal arts. Do any of the work you feel like and get the basic ideas, so that you can read all about it, or take a master's program class in it if you love it.
But use college (which is often much more expensive than those tech schools) for what it's been designed to do for over the past 700 or so years: let it turn you into a thinking, logical, well-rounded person.
That's not to say you shouldn't be doing computer stuff in college. But if you work on free software projects in school, you'll learn what the other guys learned in class, and you'll have killer resume line-items after graduation. As in: "Worked on open source project froboz 3.1 to 3.4 as part of a team -- coded parts of bad block collection, memory coalescing, garbage collection, blah, blah, blah."
GPA means a lot. Experience means much more. But projects!?!? Man, can you get your foot in the door with projects! If it's something I as an employer actually use? And lo, I find your name in the docs? You're hired, big fella.
Much less labor, much less fuel consumption. Much less cost for individual carrier equipment. (Can someone else comment on the cost of rail vs. highway maintenance?)
If this is a way to make NAFTA better for everyone, they need to scrap the highway (or at least scale it back to very little) and run rails. If it's a way to generate tariffs on transport, well, rails do that, too.
But they wouldn't need 175 billion dollars for it. If they want to spend that kind of money, they should think about running rail lines through Texas (using some of the rails already there), building over and underpasses for existing rails in and around cities all over the country, running lines around cities to avoid marshalling yards (with their speed restrictions) and building efficient Intermodal systems in smaller towns (there are already such systems in the big ones).
But that would just mean investing in a rail company instead of press-releasing and creating a whole new way of thinking about roads, etc.
I bought a pair of the Plane Quiet headphones and am VERY VERY happy with them. But only on planes.
The sound they're good at cancelling is engine noies and other fairly low frequency stuff. They damp voices, but not by that much. If you wear them in a conversation, you hear the other folk -- they just sound low-fi.
My hugely favorite thing for damping noise and actually getting work done is called a DOOR. If you don't have one, ask your boss for one. If he's smart and able, he'll get one for you. Your productivity will go up.
As for using music to dampen, see Tom DeMarco's book, _Peopleware_. In it he runs tests with silence, office noise and music. The music ain't much better than the office noise for concentration. (page 78)
"settle"? Not a chance that can happen. Drop the case is really the only thing that could happen. Why would anyone pay them a dollar? (Unless the settlement, is "we'll drop the case if you agree not to sue US.")
He's only sued people with enough money to defend themselves. I suspect he'll keep lawsuits open until SCO goes into bankruptcy, for fear that SCO will be countersued for legal fees.