The short of it is that layout.css.visited_links_enabled is still true so _you_ get to see the visited links styled differently, but the website is blocked from being able to tell apart visited and unvisited links. And yes, every single Firefox 4 beta has had that fix, including this one.
> And yet if I open Firefox it takes nearly 30 - 45 seconds, while Chrome opens almost > instantly.
I assume that's with a clean profile (or at least one without extensions) and that you're using the Firefox 4 beta, right?
On Windows Vista/7, there are some existing dwrite issues being worked through where trying to initialize the dwrite library will go and read all the fonts on your operating system in their entirety; issues that are partially fixed if you install the updates IE9 requires and partially being worked around on the Firefox end for Firefox 4; those patches aren't done yet.
Also on Windows it's common for OEMs (and things like Skype!) to install system-wide Firefox extensions that severely impact Firefox performance. That's being worked on too. I sort of assume you checked for this possibility already.
If you're not on Vista or Windows 7, I'd really appreciate you reporting a bug about this to Mozilla and ccing me (put the string ":bz" in the cc field). It should be quite possible to measure what's going on during those 30-45 seconds and figure out why you're seeing slow startup when others aren't. It's pretty late in the game, but we might be able to fix whatever is causing you problems before final!
If you look at http://www.finaid.org/calculators/finaidestimate.phtml and select the "Institutional Methodology" (which is what the elite schools use), 4 person family, 1 person in college, MD as the state, 55 for older parent's age, $120k AGI, $18k federal tax paid (it's about the right ballpark for that AGI given no weird deductions), $120k for father's earned income, $20k in a rainy-day savings account, zeroes for everything else for the parents except home equity, but assume the house is paid off so there's $300k of home equity. Assume the student also lives in MD, is 18, and has no money at all.
Estimated family contribution: $40103 per year.
While this is only 2/3 of "sticker price" if you include room and board and other things they estimate at Harvard, say, it's more than the yearly Harvard tuition.
Note that $300k is about 50% below median for the close-to-DC counties of Maryland, by the way. And that the home equity number is influenced by how much your home is assessed at, not anything tangible. So it can just drop when the housing market drops.
Had this family, or just the student actually had any savings, they would be paying a good bit more (12% of parent assets; 20% of student assets; both per year).
Now things are definitely changing; ten years ago I could have made the income way lower and gotten similar numbers, but now there are some income cutoffs below which the institutional methodology no longer considers assets. Which means that it may well be worth it to quit your job or get a much lower-paying one as your kids go to college, depending on the state of your home loan and your bank account.
Of course if I modify the example to have zero home equity the EFC drops to $25230....
All of which says that the only people who pay sticker for the elites are those who are stupid or uncaring about money enough to have actually saved some money and to still have jobs while their kids are in college.
Whether this is a good incentive system is an open question.
Whatever their reasons (and I know what they are, yes, and agree with them), the upshot was that they said they would not implement it. So it was DOA for standardization.
Opera and Gecko refused to implement one of the possible contenders (H.264) for patent reasons. Furthermore, H.264 doesn't comply with the spirit of the W3C patent policy, though it does comply with the letter (because while a W3C spec can't require implementation of a W3C-designed techonlogy that has W3C members holding patents on it and not licensing them, it _can_ require implementation of a patented technology developed by someone else, via citing it by reference).
Apple refused to implement anything other than H.264.
Microsoft refused to comment, basically.
Google implemented H.264 and the other containers+codecs Gecko and Opera implement (WebM/VP8 and Ogg/Theora).
So anything that was going to be specified was going to be a fiction in practice....
No, but I am a mathematician and physicist by training with some so-so knowledge of statistics (not enough to do original research, but enough to do error analysis on my experiments, say).
> All those buzzwords you used in your failed attempt to look smart refer to measurable > numeric characteristics
Yes, and "computed experience" is not all that hard to define (in various ways, agreed) and then measure. And given pretty much any reasonable definition and measurement technique my statement would be true. Now you can accuse me of being insufficiently pedantic in that I didn't define such a measurement technique, but that's where the fact that in this case it really doesn't matter much comes in.
> But saying Mr X is the mode of Lalaland is just retarded.
When measuring somewhat imprecise quantities (like "computer experience"; this is less applicable to salaries) it's common to deal with the imprecision by binning (e.g. instead of asking people for the exact to-the-second amount of time they've programmed in C, whicih they couldn't tell you if they tried, you ask them for the number of years, possibly with some predefined non-single-year ranges). At which point it does in fact make sense to speak of the mode of the resulting distribution.
But as an aside, you could in fact talk about the modal salary. It's just not a very useful measure when binned at the 1 cent level, so no one does.
Again, I could have been more pedantic and said "The amount of computer experience you have is likely several standard deviations away from any of the mean, median, or mode amounts of computer experience in the population of grandmothers in the United States in the year 2010." But _that_ would have sounded pretentious.;)
To be fair, you are not the mean, median, or mode grandmother. Nor anywhere within several standard deviations of one...
But yes, the article should probably just have said "just about anyone" instead of "grandmother". I would bet that the average kid using the Web would have a harder time with do not track mechanisms than the average grandmother, if nothing else. For one thing, the kid doesn't even understand what the problem is...
Actually, the driveway analogy is a bad one; it's quite different from the Borings' situation. I have several neighbors; I'm ok with them turning around in my driveway because I know them and trust them to, say, watch out for my kids in the process, because they know said kids exist. I don't thus trust the world.
As another example, people are commonly ok with their family members (including somewhat extended family, easily dozends of people) or close friends doing things that are completely not ok for random strangers to do. Like, say, using their hidden emergency key to get in the house as needed.
> What did Mr Greg Bear contribute to the literary world that he may reap these royalty > fees?
Who says he's reaping them? The author's widow is, no?
Greg Bear just happens to be:
1) The son-in-law of the author in question, hence have a personal interest in the matter. 2) A published author in his own right; quite famous in science fiction circles (if you
were not aware of this, a short search would turn it up). This mostly means that he's
maybe thought about copyright more than average.
And as a matter of fact, the works in question are not public domain yet; that's the point of the Gutenberg response that was cited.
> You don't "control the web" by adopting open standards.
No, but you do control it by making possibly-unilateral changes to the way the web works and getting them adopted as "open standards" (which is a pretty low bar, actually; witness a number of the W3C standards).
Well, the apparent fragility of the code is not speculation. Whatever is going on here is certainly very fragile.
Whether it's innocent or purposeful is certainly speculation. Rob isn't doing such speculation in public (though I bet he has a private opinion), nor am I willing to, precisely to uphold the character of Mozilla.
Well, sure. If you read carefully, the vast majority of Sunspider is actually dead code; one of the many reasons it's a bad benchmark. More modern benchmarks are more careful to not do that....
1) If you actually read the article, you may have noticed that the engineer is named. It's
right there there at the beginning of paragraph 2: "While Mozilla engineer Rob Sayre" 2) The "cheating" stuff is all from the Hacker News thread and the fucking articl. I
suggest you further read item 1 under "Further Readings" on the fucking article, which
is what Rob actually wrote. The link is: http://blog.mozilla.com/rob-sayre/2010/11/16/reporting-a-bug-on-a-fragile-analysis/
Just to save you the trouble of reading it, if don't want to, it's pretty clear that IE9 is eliminating the heart of the math-cordic loop as dead code. It _is_ dead code, so the optimization is correct. What's weird is that very similar code (in fact, code that compiles to identical bytecode in some other JS engines) that's just as dead is not dead-code eliminated. This suggests that the dead-code-elimination algorithm is somewhat fragile. In particular, testing has yet to turn up a single other piece of dead code it eliminates other than this one function in Sunspider. So Rob filed a bug about this apparent fragility with Microsoft and blogged about it. The rest is all speculation by third parties.
The interesting part is that setting up that sort of process would make it very difficult to claim you're just interested in power generation... Reactor-grade uranium is a few percent (3-4 according to http://en.wikipedia.org/wiki/Enriched_uranium) U-235; weapons-grade is closer to 80+% U-235 (85% for Little Boy according to the first link above; about 90% for modern weapons-grade according to the second link). Getting the latter by methods designed to look like you're just aiming for the former is a bit of a pain, I imagine.
All this assumes we can trust wikipedia on the subject.;)
> No holiday time, no sick leave, no maternity leave, no restrictions on hours worked, no > mandated breaks, few health and safety regulations, can be fired without notice or > reason, can legally discriminate, etc.
While there are no _federal_ laws covering some of those items, there are plenty of _state_ laws (e.g. a number of states have laws requiring paid maternity leave; several have required paid paternity leave as well). Further, there are federal laws requiring certain treatment of maternity leave (e.g. treating it at least as well as you treat any medical short-term disability). There are certainly federal laws against discriminating (though perhaps your definition of "discriminate" is different from the one in those laws?). Health and safety regulations clearly vary by field; I'd like to see your data on "few".
In quite a number of jobs there are legal requirements that anything past 40 hours a week be paid at overtime rates; while this may not be "restrictions on hours worked", it's not quite the same thing as "slavery" either.
At-will employment, I'll grant, for most employees. Bad for job security; possibly better for finding jobs in the first place. Hard to say.
Not sure what the latency is, but if it's too high for uses like this, please let the people involved know? They want this to actually be useful for exactly the sort of things you're talking about.
> It's generally taught as an advanced class senior year.
At least 50% of the seniors in my high school took calculus senior year. That was 13 years ago. The percentage is higher now.
And more importantly, "getting to calculus" is seen as the goal.
> Which we're not doing.
But not through lack of _trying_. The entire high school math curriculum is centered on getting the students to calculus by senior year; the ones who don't get there take the same exact classes, but just one year later (so they're taking a course typically officially named "precalculus" senior year, with an eye to taking calculus freshman year in college).
> You've argued that we should be teaching "practical" math to students
I'm arguing that we should be making sure that baseline math education includes the things that are most needed to function well in society. That's not quite the same as "practical" math, nor is it what we're doing right now.
I agree that stopping teaching math after elementary school is not so useful; there's way too much people really need to know to make informed decisions. They're just not taught it now, either.
> You also need to give them the chance to opt out of literature, music, dance and history > and politics classes
How many high schools in the US require 4 years of music to graduate? 4 years of dance? 4 years of literature? (Quite a number require 4 years of English, but that includes composition, reading comprehension, etc; you don't really get to "literature" until maybe senior year, if at all.)
In many places there aren't "politics classes" at all; in some there are semester-long "civics" classes that are required.
So we're actually most of the way to what you propose; not only can most people opt out of that stuff, but in many cases they can't even opt into it.
History is the one thing on your list that's commonly required to be taught for multiple years on the high-school level.
> How can statistics possibly be useful in today's world?
Statistics is not part of the required math curriculum most places in the US. It would be far better if it were (replacing, say, calculus).
> Or an understanding of continuously changing variables, like mortgages?
A good financial math class would be far more useful to people than delta-epsilon proofs.
I'd very much support an effort to reach basic financial math (which used to be taught in Home Economics classes before those got dumbed down, by the way) and basic statistics (to the level of being able to read an article and understand when you're being bullshitted, not to the level of being able to do your own statistical analysis on all of your own raw data) to everyone in high school. We'd be way better off than we are now trying to teach everyone calculus.
If the goal is not practical skills but rather satisfying intellectual curiosity, there are lots of parts of math that make a better subject for it than your typical second semester of calculus (which largely focuses on recipes for integrating different kinds of functions... sometimes with a bit of jazzing-up by adding Taylor series, but the main focus is definitely on the recipes).
> when did "ramblings of a random guy" become "news"?
This guy isn't quite random, being someone who actually knows a little something about math.
Oddly enough, many of other people I know who know a good bit of math agree with him on the basic points: the problem is that we're teaching the _wrong_ sorts of math to people in high school and early college, and killing off their enjoyment of the subject while at the same time not actually providing them with the skills they actually need.
Slashdot Mirror
> Can this thing prevent covert, un-removable install of add-ons .NET Framework Assistant)?
> (e.g.
Not yet. Being worked on.
> Does it set layout.css.visited_links_enabled to false?
You did read http://dbaron.org/mozilla/visited-privacy right? It was linked to from the article mentioned in response to the comment you cite.
The short of it is that layout.css.visited_links_enabled is still true so _you_ get to see the visited links styled differently, but the website is blocked from being able to tell apart visited and unvisited links. And yes, every single Firefox 4 beta has had that fix, including this one.
> And yet if I open Firefox it takes nearly 30 - 45 seconds, while Chrome opens almost
> instantly.
I assume that's with a clean profile (or at least one without extensions) and that you're using the Firefox 4 beta, right?
On Windows Vista/7, there are some existing dwrite issues being worked through where trying to initialize the dwrite library will go and read all the fonts on your operating system in their entirety; issues that are partially fixed if you install the updates IE9 requires and partially being worked around on the Firefox end for Firefox 4; those patches aren't done yet.
Also on Windows it's common for OEMs (and things like Skype!) to install system-wide Firefox extensions that severely impact Firefox performance. That's being worked on too. I sort of assume you checked for this possibility already.
If you're not on Vista or Windows 7, I'd really appreciate you reporting a bug about this to Mozilla and ccing me (put the string ":bz" in the cc field). It should be quite possible to measure what's going on during those 30-45 seconds and figure out why you're seeing slow startup when others aren't. It's pretty late in the game, but we might be able to fix whatever is causing you problems before final!
Sort of.
If you look at http://www.finaid.org/calculators/finaidestimate.phtml and select the "Institutional Methodology" (which is what the elite schools use), 4 person family, 1 person in college, MD as the state, 55 for older parent's age, $120k AGI, $18k federal tax paid (it's about the right ballpark for that AGI given no weird deductions), $120k for father's earned income, $20k in a rainy-day savings account, zeroes for everything else for the parents except home equity, but assume the house is paid off so there's $300k of home equity. Assume the student also lives in MD, is 18, and has no money at all.
Estimated family contribution: $40103 per year.
While this is only 2/3 of "sticker price" if you include room and board and other things they estimate at Harvard, say, it's more than the yearly Harvard tuition.
Note that $300k is about 50% below median for the close-to-DC counties of Maryland, by the way. And that the home equity number is influenced by how much your home is assessed at, not anything tangible. So it can just drop when the housing market drops.
Had this family, or just the student actually had any savings, they would be paying a good bit more (12% of parent assets; 20% of student assets; both per year).
Now things are definitely changing; ten years ago I could have made the income way lower and gotten similar numbers, but now there are some income cutoffs below which the institutional methodology no longer considers assets. Which means that it may well be worth it to quit your job or get a much lower-paying one as your kids go to college, depending on the state of your home loan and your bank account.
Of course if I modify the example to have zero home equity the EFC drops to $25230....
All of which says that the only people who pay sticker for the elites are those who are stupid or uncaring about money enough to have actually saved some money and to still have jobs while their kids are in college.
Whether this is a good incentive system is an open question.
Whatever their reasons (and I know what they are, yes, and agree with them), the upshot was that they said they would not implement it. So it was DOA for standardization.
They tried to set a standard video codec.
Opera and Gecko refused to implement one of the possible contenders (H.264) for patent reasons. Furthermore, H.264 doesn't comply with the spirit of the W3C patent policy, though it does comply with the letter (because while a W3C spec can't require implementation of a W3C-designed techonlogy that has W3C members holding patents on it and not licensing them, it _can_ require implementation of a patented technology developed by someone else, via citing it by reference).
Apple refused to implement anything other than H.264.
Microsoft refused to comment, basically.
Google implemented H.264 and the other containers+codecs Gecko and Opera implement (WebM/VP8 and Ogg/Theora).
So anything that was going to be specified was going to be a fiction in practice....
> To be fair, you are not a statistician.
No, but I am a mathematician and physicist by training with some so-so knowledge of statistics (not enough to do original research, but enough to do error analysis on my experiments, say).
> All those buzzwords you used in your failed attempt to look smart refer to measurable
> numeric characteristics
Yes, and "computed experience" is not all that hard to define (in various ways, agreed) and then measure. And given pretty much any reasonable definition and measurement technique my statement would be true. Now you can accuse me of being insufficiently pedantic in that I didn't define such a measurement technique, but that's where the fact that in this case it really doesn't matter much comes in.
> But saying Mr X is the mode of Lalaland is just retarded.
When measuring somewhat imprecise quantities (like "computer experience"; this is less applicable to salaries) it's common to deal with the imprecision by binning (e.g. instead of asking people for the exact to-the-second amount of time they've programmed in C, whicih they couldn't tell you if they tried, you ask them for the number of years, possibly with some predefined non-single-year ranges). At which point it does in fact make sense to speak of the mode of the resulting distribution.
But as an aside, you could in fact talk about the modal salary. It's just not a very useful measure when binned at the 1 cent level, so no one does.
Again, I could have been more pedantic and said "The amount of computer experience you have is likely several standard deviations away from any of the mean, median, or mode amounts of computer experience in the population of grandmothers in the United States in the year 2010." But _that_ would have sounded pretentious. ;)
To be fair, you are not the mean, median, or mode grandmother. Nor anywhere within several standard deviations of one...
But yes, the article should probably just have said "just about anyone" instead of "grandmother". I would bet that the average kid using the Web would have a harder time with do not track mechanisms than the average grandmother, if nothing else. For one thing, the kid doesn't even understand what the problem is...
Actually, the driveway analogy is a bad one; it's quite different from the Borings' situation. I have several neighbors; I'm ok with them turning around in my driveway because I know them and trust them to, say, watch out for my kids in the process, because they know said kids exist. I don't thus trust the world.
As another example, people are commonly ok with their family members (including somewhat extended family, easily dozends of people) or close friends doing things that are completely not ok for random strangers to do. Like, say, using their hidden emergency key to get in the house as needed.
> What did Mr Greg Bear contribute to the literary world that he may reap these royalty
> fees?
Who says he's reaping them? The author's widow is, no?
Greg Bear just happens to be:
1) The son-in-law of the author in question, hence have a personal interest in the matter.
2) A published author in his own right; quite famous in science fiction circles (if you
were not aware of this, a short search would turn it up). This mostly means that he's
maybe thought about copyright more than average.
And as a matter of fact, the works in question are not public domain yet; that's the point of the Gutenberg response that was cited.
> You don't "control the web" by adopting open standards.
No, but you do control it by making possibly-unilateral changes to the way the web works and getting them adopted as "open standards" (which is a pretty low bar, actually; witness a number of the W3C standards).
> And I'm not in the boonies, I live near Boston
Depending on _where_ near Boston, this could be http://www.dslreports.com/shownews/Boston-Wonders-Where-Its-FiOS-Is-105269 or failures to negotiate TV franchise rights with your actual municipality (as was the case for Boston proper for years) or any of a number of other things....
Well, the apparent fragility of the code is not speculation. Whatever is going on here is certainly very fragile.
Whether it's innocent or purposeful is certainly speculation. Rob isn't doing such speculation in public (though I bet he has a private opinion), nor am I willing to, precisely to uphold the character of Mozilla.
Well, sure. If you read carefully, the vast majority of Sunspider is actually dead code; one of the many reasons it's a bad benchmark. More modern benchmarks are more careful to not do that....
Ah, yes. The Hacker News thread has the var foo example. It's just Mozilla's internal testing that didn't find other examples that are still DCEd.
So yeah, looks like just a very fragile dead code eliminator.
The cordic test didn't come from any web libraries, as far as I can tell.
1) If you actually read the article, you may have noticed that the engineer is named. It's
right there there at the beginning of paragraph 2: "While Mozilla engineer Rob Sayre"
2) The "cheating" stuff is all from the Hacker News thread and the fucking articl. I
suggest you further read item 1 under "Further Readings" on the fucking article, which
is what Rob actually wrote. The link is: http://blog.mozilla.com/rob-sayre/2010/11/16/reporting-a-bug-on-a-fragile-analysis/
Just to save you the trouble of reading it, if don't want to, it's pretty clear that IE9 is eliminating the heart of the math-cordic loop as dead code. It _is_ dead code, so the optimization is correct. What's weird is that very similar code (in fact, code that compiles to identical bytecode in some other JS engines) that's just as dead is not dead-code eliminated. This suggests that the dead-code-elimination algorithm is somewhat fragile. In particular, testing has yet to turn up a single other piece of dead code it eliminates other than this one function in Sunspider. So Rob filed a bug about this apparent fragility with Microsoft and blogged about it. The rest is all speculation by third parties.
In 1944 they used a three-stage process involving several very large facilities; the tail end of http://en.wikipedia.org/wiki/S-50_(Manhattan_Project) has details and links.
The interesting part is that setting up that sort of process would make it very difficult to claim you're just interested in power generation... Reactor-grade uranium is a few percent (3-4 according to http://en.wikipedia.org/wiki/Enriched_uranium) U-235; weapons-grade is closer to 80+% U-235 (85% for Little Boy according to the first link above; about 90% for modern weapons-grade according to the second link). Getting the latter by methods designed to look like you're just aiming for the former is a bit of a pain, I imagine.
All this assumes we can trust wikipedia on the subject. ;)
People _say_ that, but when they vote with their feet (or mice, as the case may be) it turns out that they do care about it. Often subconsciously.
> No holiday time, no sick leave, no maternity leave, no restrictions on hours worked, no
> mandated breaks, few health and safety regulations, can be fired without notice or
> reason, can legally discriminate, etc.
While there are no _federal_ laws covering some of those items, there are plenty of _state_ laws (e.g. a number of states have laws requiring paid maternity leave; several have required paid paternity leave as well). Further, there are federal laws requiring certain treatment of maternity leave (e.g. treating it at least as well as you treat any medical short-term disability). There are certainly federal laws against discriminating (though perhaps your definition of "discriminate" is different from the one in those laws?). Health and safety regulations clearly vary by field; I'd like to see your data on "few".
In quite a number of jobs there are legal requirements that anything past 40 hours a week be paid at overtime rates; while this may not be "restrictions on hours worked", it's not quite the same thing as "slavery" either.
At-will employment, I'll grant, for most employees. Bad for job security; possibly better for finding jobs in the first place. Hard to say.
Indeed. https://wiki.mozilla.org/Audio_Data_API#Writing_Audio and https://github.com/corbanbrook/dsp.js are closer to what's needed for realtime synthesis. Hopefully it will make its way to other browsers too.
You might be interested in https://wiki.mozilla.org/Audio_Data_API#Writing_Audio and https://github.com/corbanbrook/dsp.js
Not sure what the latency is, but if it's too high for uses like this, please let the people involved know? They want this to actually be useful for exactly the sort of things you're talking about.
This particular standard is: 1) huge (about 800 printed pages and growing) and 2) still a work in progress.
There's no way to be 100% compatible with something that changes every day, unless implementing changes takes you zero time.
> It's generally taught as an advanced class senior year.
At least 50% of the seniors in my high school took calculus senior year. That was 13 years ago. The percentage is higher now.
And more importantly, "getting to calculus" is seen as the goal.
> Which we're not doing.
But not through lack of _trying_. The entire high school math curriculum is centered on getting the students to calculus by senior year; the ones who don't get there take the same exact classes, but just one year later (so they're taking a course typically officially named "precalculus" senior year, with an eye to taking calculus freshman year in college).
> You've argued that we should be teaching "practical" math to students
I'm arguing that we should be making sure that baseline math education includes the things that are most needed to function well in society. That's not quite the same as "practical" math, nor is it what we're doing right now.
I agree that stopping teaching math after elementary school is not so useful; there's way too much people really need to know to make informed decisions. They're just not taught it now, either.
> You also need to give them the chance to opt out of literature, music, dance and history
> and politics classes
How many high schools in the US require 4 years of music to graduate? 4 years of dance? 4 years of literature? (Quite a number require 4 years of English, but that includes composition, reading comprehension, etc; you don't really get to "literature" until maybe senior year, if at all.)
In many places there aren't "politics classes" at all; in some there are semester-long "civics" classes that are required.
So we're actually most of the way to what you propose; not only can most people opt out of that stuff, but in many cases they can't even opt into it.
History is the one thing on your list that's commonly required to be taught for multiple years on the high-school level.
> How can statistics possibly be useful in today's world?
Statistics is not part of the required math curriculum most places in the US. It would be far better if it were (replacing, say, calculus).
> Or an understanding of continuously changing variables, like mortgages?
A good financial math class would be far more useful to people than delta-epsilon proofs.
I'd very much support an effort to reach basic financial math (which used to be taught in Home Economics classes before those got dumbed down, by the way) and basic statistics (to the level of being able to read an article and understand when you're being bullshitted, not to the level of being able to do your own statistical analysis on all of your own raw data) to everyone in high school. We'd be way better off than we are now trying to teach everyone calculus.
If the goal is not practical skills but rather satisfying intellectual curiosity, there are lots of parts of math that make a better subject for it than your typical second semester of calculus (which largely focuses on recipes for integrating different kinds of functions... sometimes with a bit of jazzing-up by adding Taylor series, but the main focus is definitely on the recipes).
> when did "ramblings of a random guy" become "news"?
This guy isn't quite random, being someone who actually knows a little something about math.
Oddly enough, many of other people I know who know a good bit of math agree with him on the basic points: the problem is that we're teaching the _wrong_ sorts of math to people in high school and early college, and killing off their enjoyment of the subject while at the same time not actually providing them with the skills they actually need.
> free for end-users at the moment as long as they are using a licensed decoder
In other words, free as long as they have already paid for it, right? ;)