Computer Science programs at accredited universities are accredited through the Accreditation Board for Engineering and Technology, Inc. (ABET), specifically though the Computing Accreditation Commission (CAC).
The ABET CAC web site is located at:
http://www.abet.org/cac1.html
Here is the accredidation criteria:
http://www.abet.org/criteria_cac.html
Here is the current list of accredited Universities:
Please note that the University of Waterloo is absent, but the University of California at Berkeley and Washington State University are both present.
If your College or University is not on this list, expect to have to take additional classes in order to be admitted to graduate degree programs at anywhere that *is* on the list.
"Thats not what they are doing at all.. read the article..."
Quoting from the article which I have read, and apprently you have not: "Completion of this course will be mandatory for students entering the E&CE program."
"Otherwise why would my college teach a class called c and a class called c++ and a class called visual basic."
Because your college is not accredited? It's a community collge limited to associates degrees? You are in a country other than the United States, which currently leads the world in software and other engineering, precisely because of the standards enforced by its institutes of higher education? Stop me if I'm getting close...
Tempting as it is to post "Let's build Gotham City, complete with the Rodan's ``The Thinker''/Heroic Soviet Industrial style public statues!"...
The information is archival in nature; it would not be used to rebuild the city, any more than the plans and photographs of the World Trade Center are now being used to rebuild the World Trade Center, instead of something new.
The value it has is as a tool today, as described in the article, makes it just that: a tool. And the security concerns aren't over its value in rebuilding, they're over it's value to someone who wants to raze the place, or model a "Cobra Event" style drop of a bioagent.
Massive datassemblies are probably some of the few valid objects for protection via security through obscurity.
Back in the early 1980's, there was a change in the accreditation rules which stated that a College or University could not teach a specific computer language and remain accredited.
The result was a lot of courses, like "Data structures using C++" and "Business programming with FORTRAN", rather than specific language courses, because computer languages were no longer allowed to be an ends in themselves: the accreditation committees had seen the writing on the wall that days of "COBOL is forever" would son be over.
It's interesting to see that the U of Waterloo is endorsing a pure language course as a program entry requirement; I would not like to be standing near the fan when the accreditation committee gets a hold of this (though I'm not adverse to turning the fan on by telling them).
If I were a student currently at, or considering entering, the U of Waterloo, I would be rather worried that I would not be able to go on to a graduate degree from Stanford, MIT, CIT, Berkeley, etc., because the University from which I received my undergraduate degree was not accredited.
I would be particularly concerned if this was my third or second year, or even my first, if I turned down other offers to attend there instead.
This sets a nice stage to provide a court test on the patentability of software at an appellate court level. If software is not patentable, then the agreement is not applicable to software.
He deserved to lose because he asked permission the way he did, and because he dissolved the relationship with his lawyer.
But that leaves several avenues for appeal, and one of them is to question the patentability of software in the first place.
Publically funded research should be in the public domain. Period.
Since this is not possible because of legal liability issues, the next best thing is a BSD/CMU/MIT license, which is effectively public domain with a hold harmless clause.
The legislation we *should* be calling for here is indemnification *in law* for authors who put their works into the public domain.
The problem with any license other than public domain is that it limits the people who can use it to a subset of the public.
Unfortunately, the GPL contains a "poison pill" in the "no additional restrictions clause"; this has been addressed by "the two clause BSD license", voluntarily, giving in to RMS' immutable world view.
But the GPL contains *another* "poison pill" against any license other than the GPL, by requiring derivative works to be licensed under the GPL. This is the license author's choice, but it ignores the possibility of combined derivation from both GPL and other sources. This time, it's up to RMS to give a little.
Rather than pushing for publically funded research to be GPL'ed, the GPL itself needs to change to permit the possibility of mixed derivation code; do that, and the entirety of the code that has gone into BSD and other large public code bases immediately becomes available to Linux -- and vice versa.
Limit miscibility to Open Source licenses, if you want (I even have a catch name: "the GNU Open Source Public License" -- GOSPL, pronounced "gospel").
The Internet is built on code that is as close to public domain as it is legally possible for the authors to make it. The model works. The model *wins* vs. closed source commercial attempts, like NetBEUI, IPX/SPX, XNS, DECNet, etc., all of which predated the public Internet, and had an opportunity to compete fairly... and lost to (effectively) public domain code. This code did not win because it was superior -- who designs a protocol that requires two responses to a single request by a client doing a connection shutdown to free up server resources, assuming that all clients will "play nice", or is succeptible to SYN-floodin *on purposed*?
It won *because* it was capable of being included in commercially produced routers and operating systems at a tiny marginal cost, in addition to other protocols, without enrighing one of the players on the field to the disadvantage of the others.
It's time to acknowledge the fact that there *is* an "intellectual commons" -- AND IT WORKS.
Yeah, I would be just as happy to have Open Source required for all government purchases -- particularly after the recent Oracle vs. California debacle on continuing licensing fees -- if only so that once tax money is spent on something, it doesn't have to be respent each year to *keep* that something. It doesn't so much matter to me that the winner be Linux or something else, so long as no one gets a lamprey-like grip on the public teat.
But that's a *totally* seperate issue from publically funded research, and just because the two are mixed together in the same position paper by an organization that it's otherwise easy to dislike, doesn't mean that all their positions are defacto unreasonable, or can be tarred with the same brush as the worst of their positions.
Publically funded research's results should be *public*, or as public as possible, until we can get the liability laws changed to the point that the results can be *truly public*.
So how long before there are stegonographic comments in Linux source coming out of China to get around the gvernment censorship of the media, but not of source code?
"Take the first letter of each fortune in the fortune file, and then..."
Please research the state of the art before you post. Just because your favorite software company has not seen fit to license Kurtzweil's code, or the (arguably better, but based on his ideas) Russian handwriting recognition code doesn't mean that it doesn't exist or isn't effective.
I use Kurtzweil's code for handwriting recognition on a daily basis on an ~8MHz embedded 68010 class system, and have no complaints.
Today we have very fast processors, and there's no excuse for not doing handwriting recognition, if you have the hardware to do the input.
The human brain operates at about 120Hz; it takes 1/120th of a second for a neuron to repolarize so that it can fire again. So I'm not underestimating anything.
As to your straw man examples: your average human will at *least* be more likely to be able to understand human written text, when presented one sample of human written text, and another sample of human written Grafiti.
The point that we build systems for humans an not humans for systems remains valid.
And my expectation of being able to model the recognition behaviour of humans on computers is based on the work by Ray Kurtzweil using trained back propagation neural networks, which can be coded in software very easily, and which can take an absolutely huge amount of information chains into account when run on serial hardware at 2GHz.
Case insensitivity required globbing in the kernel. UNIX has historically done all globbing in user space, and passed only literal values to the kernel.
Case folding on storage is realtively easy (Just Do It(tm)), but case folding on lookup for system calls that take explicit names (open, stat, creat, rename, unlink, etc.) requires that the literal value passed match multiple potential values as soted in a directory.
Doing globbing in the kernel is actually more efficient in a lot of ways (consider that when doing an "ls -c", rather than pushing the whole directory contents over the user/kernel boundary via the getdirentries/getdents system calls, you would push only names which matched the expression, drastically reducing the protection domain crossing overhead).
However, POSIX practically *demands* that you not do your globbing in the kernel.
Apple gets around this by force-folding. Basically, it is case sensitive on storage, and case insensitive on lookup (Windows outside the legacy 8.3 namespace is the same). The problem with this, as you would know, if you had ever written internationalized FS code on Windows to be installe dunder the IFSMgr, is that there is a requirement for "active character set case folding tables", and these are locale specific (e.g. what do you do with the Schlossen if you are in Germany vs. France, and how do you handle Hangul, Kanji, etc.).
Basically, without destorying certain standards mandated behaviour, it's not possible. Most likely, you will not find a programmer willing to give up POSIX compliance or internationalization in order to change this behaviour.
"Your capital T curves up, mine looks like a T Square, his looks like italic script, hers looks like a small T anyway. How is a computer supposed to know what you're talking about?"
The same way other humans know.
Only faster.
-- Terry
If we are going to go ballistic...
on
Gone Fission
·
· Score: 5, Funny
If we are going to go ballistic over non-native species, then we should start by wiping out all the pigeons in North and South America, please.
The author discussed, in true binary fashion, "available" vs. "unavailable" for a second roll.
People have also discussed "over the top vs. under the bottom".
There is a middle ground: two rollers, one "over the top" and the other "under the bottom". This would also self-regulate the usage of the rolls, ensuring one ran out before the other.
Which one runs out first depends on the physical characteristics of the holder.
For a fixed holder, the answer is probably "over the top".
But for a hanging holder, where the weight of the paper itself levers the paper into the wall, increasing the overall friction, "under the bottom" allows you to mitigate friction effects, while "over the top" increases them.
Since a fixed holder results in the choice being user preference, that doesn't solve anything; clearly, the fix is in two parts: (1) use hanging holders, and (2) make the primary roll "under the bottom" and the secondary roll "over the top".
Another solution (which is only statistical) is to locate rolls on either side of the toilet. Left handed people are outnumbered by right handed people 20 to 1, on average. But this fix only works "on average", as a result.
Your theory of public transportation is a game of statistics. I am supposed to be convinced that, at a particular population density, people will start using it.
The problem with this theory is that when I want to go home at 9:37 PM, I want to *go home* at 9:37 PM, not walk 4 blocks, wait 15 minutes for a bus, ride for a while, get off, wait another 15 minutes to transfer to another bus, ride for another while, get off, and walk another 4 blocks.
Public transportation is always a great idea *for other people, not for you*.
I've discussed this in letters with Governor Gray Davis of California. The only real way to get a lot of people onto public transportation is to make it fully subsidized.
The only way to get people to take public transportation is to make the marginal cost of taking it less than the marginal cost of not taking it. Making it free would drastically increase ridership. But there will still be people who will not take it, even if you were to pay them.
It's more about "how many drunken idiots can we get to strap drowning machines to themselves and jump into the San Francisco Bay?"
The Gossamer Condor, which won the Kremer Prize in 1977 is, so far, the smallest human powered aircraft. It has a wingspan of 70 feet. The Gossamer Albatross, which crossed the English Channel, and won the second Kremer Prize, has a wingspan of 90 feet.
A wingspan of 30 feet is pretty mugh guaranteed to not result in flight, if the wings alone are intended to provide the lift, with just human power.
The weight limit lets out most practicaly designs, such as a neutral buoyancy ornithopter with a helium lifting baloon with a 30 foot maximum width (hint: search for "one person helium balloon").
Prove to me they were personal, not corporate loans.
In other words, that the money was going into the director's pockets, instead of the pockets of the organization whose interests the directors were there to defend in the first place. This *really* is not different than a stock grant with private placement. It's better, because it doesn't dilute the shares that are the founders and the employee ISOs, like it would if it were a true private placement.
Realize... there *ARE* no other investors, than the people who put up the money. You can talk about "sweat equity", but it's not worth very much in the long run, unless there's a big upside because there are different classes of stock. Investors have preferred liquidation, and early investors have ratchet antidilution. If they didn't get these things, they would not invest, and there's be no company and no ISOs for the employees in the first place.
If it *was* into the directors pockets... it's no more a handout than a salary is a handout, if it was intended as tax deferred compensation represented as a loan (which is what personal loans mostly are).
Yeah, it's possible to rape a company this way, but it generally doesn't happen, because if it's for stock, the stock ends up being collateralized. The company only suffers if their stock loses value. There are laws against raping a company this way; it's called "embezzleing", so they tend to keep it as legal and above board as possible (particularly directors, who are legally culpable, and subject to shareholder lawsuits).
It's my understanding of the article that it was a loan for stock in another company. If so, it's clearly a case of bailing money out as a form of risk mitigation.
Note that the capital gains taxes in the U.S. have made it common practice for companies to sell unvested incentive stock options to the employee using a company guaranteed loan, on which interest is not paid until the stock has vested (simple answer: exercise stock as it becomes vested). The only downside is that if the company folds, the loans become debt instruments, like any other; but you can quit-claim the options, and still get out.
All of these are really tax dodges to get around the draconian tax laws that were only supposed to apply to the top 1% of income earners, but end up screwing the little guys.
I know tons of engineers who lost cars, houses, and even one who committed suicide over their exercise in Feb/March of 2000 of stock options that were so far under water by the time they left the lock-out period, that they owed literally millions of dollars as of January 1st, and things didn't get any better by April 15th of 2001.
It seems to me that it's a double standard, if it's "unethical and scummy" on one hand, and "the only way to avoid losing everything" on the other.
The real problem is that such contortions are necessary to protect little guys from tax laws which are blatantly anti-little-guy, for all of their intent to hurt only people for which no one has any sympathy.
The libertarians who found an underwater atol close to the surface in the tropis, and then built an island on top of it to get out from under the yoke.
Oh yeah... they failed, when the U.S. paid the Tongan Navy $1M to go plant a Tongan flag on it to keep it from happening.
Apparently, you missed the part where I noted that it was unclear as to whether or not these were personal loans, or corporate loans for the purchase of alternate stock.
Personal loans are a tax dodge to get around compensation issues by turning a tax liability into a deferred liability with a short term asset.
This type of thing wouldn't be necessary, if there were no capital gains tax. As it is, between state and fed, 55% of all capital gains ends up in the government's pocket, otherwise, instead of funding new business, paying people's salaries, creating jobs, etc..
More likely, it was corporate, with the investors bailing out of a bad investments.
Yeah, it undermines the business in question, but it protects the money from some idiot who thinks there's such a thing as "page views" and "selling eyeballs".
It's the fiduciary responsibility to directors to protect the money of the investors they represent, and sometimes, this means taking it out of a bad investment, and putting it some place else.
Yeah, businesses fail over this. Euthanasia is better than the alternative. If you are looking for compassionate spending of millions of dollars to prop up a failed business model, you are dreaming.
Slashdot Mirror
Computer Science programs at accredited universities are accredited through the Accreditation Board for Engineering and Technology, Inc. (ABET), specifically though the Computing Accreditation Commission (CAC).
i te .html
~ csab
The ABET CAC web site is located at:
http://www.abet.org/cac1.html
Here is the accredidation criteria:
http://www.abet.org/criteria_cac.html
Here is the current list of accredited Universities:
http://www.abet.org/accredited_programs/CACWebs
Please note that the University of Waterloo is absent, but the University of California at Berkeley and Washington State University are both present.
If your College or University is not on this list, expect to have to take additional classes in order to be admitted to graduate degree programs at anywhere that *is* on the list.
See also:
http://www.acm.org/education/
http://csab.org/
Have a nice day.
-- Terry
"Thats not what they are doing at all.. read the article..."
Quoting from the article which I have read, and apprently you have not: "Completion of this course will be mandatory for students entering the E&CE program."
"Otherwise why would my college teach a class called c and a class called c++ and a class called visual basic."
Because your college is not accredited? It's a community collge limited to associates degrees? You are in a country other than the United States, which currently leads the world in software and other engineering, precisely because of the standards enforced by its institutes of higher education? Stop me if I'm getting close...
-- Terry
"There are stronger cases to use to fight software patents."
Which will never get their day in an appellate level court, where they have the danger of becoming case law against software patents.
Any time this issue ever gets raised, it ends up getting settled out of court to prevent it becoming a legal precedent.
-- Terry
Tempting as it is to post "Let's build Gotham City, complete with the Rodan's ``The Thinker''/Heroic Soviet Industrial style public statues!"...
The information is archival in nature; it would not be used to rebuild the city, any more than the plans and photographs of the World Trade Center are now being used to rebuild the World Trade Center, instead of something new.
The value it has is as a tool today, as described in the article, makes it just that: a tool. And the security concerns aren't over its value in rebuilding, they're over it's value to someone who wants to raze the place, or model a "Cobra Event" style drop of a bioagent.
Massive datassemblies are probably some of the few valid objects for protection via security through obscurity.
-- Terry
Back in the early 1980's, there was a change in the accreditation rules which stated that a College or University could not teach a specific computer language and remain accredited.
The result was a lot of courses, like "Data structures using C++" and "Business programming with FORTRAN", rather than specific language courses, because computer languages were no longer allowed to be an ends in themselves: the accreditation committees had seen the writing on the wall that days of "COBOL is forever" would son be over.
It's interesting to see that the U of Waterloo is endorsing a pure language course as a program entry requirement; I would not like to be standing near the fan when the accreditation committee gets a hold of this (though I'm not adverse to turning the fan on by telling them).
If I were a student currently at, or considering entering, the U of Waterloo, I would be rather worried that I would not be able to go on to a graduate degree from Stanford, MIT, CIT, Berkeley, etc., because the University from which I received my undergraduate degree was not accredited.
I would be particularly concerned if this was my third or second year, or even my first, if I turned down other offers to attend there instead.
-- Terry
This sets a nice stage to provide a court test on the patentability of software at an appellate court level. If software is not patentable, then the agreement is not applicable to software.
He deserved to lose because he asked permission the way he did, and because he dissolved the relationship with his lawyer.
But that leaves several avenues for appeal, and one of them is to question the patentability of software in the first place.
-- Terry
Publically funded research should be in the public domain. Period.
Since this is not possible because of legal liability issues, the next best thing is a BSD/CMU/MIT license, which is effectively public domain with a hold harmless clause.
The legislation we *should* be calling for here is indemnification *in law* for authors who put their works into the public domain.
The problem with any license other than public domain is that it limits the people who can use it to a subset of the public.
Unfortunately, the GPL contains a "poison pill" in the "no additional restrictions clause"; this has been addressed by "the two clause BSD license", voluntarily, giving in to RMS' immutable world view.
But the GPL contains *another* "poison pill" against any license other than the GPL, by requiring derivative works to be licensed under the GPL. This is the license author's choice, but it ignores the possibility of combined derivation from both GPL and other sources. This time, it's up to RMS to give a little.
Rather than pushing for publically funded research to be GPL'ed, the GPL itself needs to change to permit the possibility of mixed derivation code; do that, and the entirety of the code that has gone into BSD and other large public code bases immediately becomes available to Linux -- and vice versa.
Limit miscibility to Open Source licenses, if you want (I even have a catch name: "the GNU Open Source Public License" -- GOSPL, pronounced "gospel").
The Internet is built on code that is as close to public domain as it is legally possible for the authors to make it. The model works. The model *wins* vs. closed source commercial attempts, like NetBEUI, IPX/SPX, XNS, DECNet, etc., all of which predated the public Internet, and had an opportunity to compete fairly... and lost to (effectively) public domain code. This code did not win because it was superior -- who designs a protocol that requires two responses to a single request by a client doing a connection shutdown to free up server resources, assuming that all clients will "play nice", or is succeptible to SYN-floodin *on purposed*?
It won *because* it was capable of being included in commercially produced routers and operating systems at a tiny marginal cost, in addition to other protocols, without enrighing one of the players on the field to the disadvantage of the others.
It's time to acknowledge the fact that there *is* an "intellectual commons" -- AND IT WORKS.
Yeah, I would be just as happy to have Open Source required for all government purchases -- particularly after the recent Oracle vs. California debacle on continuing licensing fees -- if only so that once tax money is spent on something, it doesn't have to be respent each year to *keep* that something. It doesn't so much matter to me that the winner be Linux or something else, so long as no one gets a lamprey-like grip on the public teat.
But that's a *totally* seperate issue from publically funded research, and just because the two are mixed together in the same position paper by an organization that it's otherwise easy to dislike, doesn't mean that all their positions are defacto unreasonable, or can be tarred with the same brush as the worst of their positions.
Publically funded research's results should be *public*, or as public as possible, until we can get the liability laws changed to the point that the results can be *truly public*.
And that's my 2 cents.
-- Terry
So how long before there are stegonographic comments in Linux source coming out of China to get around the gvernment censorship of the media, but not of source code?
"Take the first letter of each fortune in the fortune file, and then..."
-- Terry
Please research the state of the art before you post. Just because your favorite software company has not seen fit to license Kurtzweil's code, or the (arguably better, but based on his ideas) Russian handwriting recognition code doesn't mean that it doesn't exist or isn't effective.
I use Kurtzweil's code for handwriting recognition on a daily basis on an ~8MHz embedded 68010 class system, and have no complaints.
Today we have very fast processors, and there's no excuse for not doing handwriting recognition, if you have the hardware to do the input.
-- Terry
The human brain operates at about 120Hz; it takes 1/120th of a second for a neuron to repolarize so that it can fire again. So I'm not underestimating anything.
As to your straw man examples: your average human will at *least* be more likely to be able to understand human written text, when presented one sample of human written text, and another sample of human written Grafiti.
The point that we build systems for humans an not humans for systems remains valid.
And my expectation of being able to model the recognition behaviour of humans on computers is based on the work by Ray Kurtzweil using trained back propagation neural networks, which can be coded in software very easily, and which can take an absolutely huge amount of information chains into account when run on serial hardware at 2GHz.
-- Terry
Case insensitivity required globbing in the kernel. UNIX has historically done all globbing in user space, and passed only literal values to the kernel.
Case folding on storage is realtively easy (Just Do It(tm)), but case folding on lookup for system calls that take explicit names (open, stat, creat, rename, unlink, etc.) requires that the literal value passed match multiple potential values as soted in a directory.
Doing globbing in the kernel is actually more efficient in a lot of ways (consider that when doing an "ls -c", rather than pushing the whole directory contents over the user/kernel boundary via the getdirentries/getdents system calls, you would push only names which matched the expression, drastically reducing the protection domain crossing overhead).
However, POSIX practically *demands* that you not do your globbing in the kernel.
Apple gets around this by force-folding. Basically, it is case sensitive on storage, and case insensitive on lookup (Windows outside the legacy 8.3 namespace is the same). The problem with this, as you would know, if you had ever written internationalized FS code on Windows to be installe dunder the IFSMgr, is that there is a requirement for "active character set case folding tables", and these are locale specific (e.g. what do you do with the Schlossen if you are in Germany vs. France, and how do you handle Hangul, Kanji, etc.).
Basically, without destorying certain standards mandated behaviour, it's not possible. Most likely, you will not find a programmer willing to give up POSIX compliance or internationalization in order to change this behaviour.
-- Terry
"Your capital T curves up, mine looks like a T Square, his looks like italic script, hers looks like a small T anyway. How is a computer supposed to know what you're talking about?"
The same way other humans know.
Only faster.
-- Terry
If we are going to go ballistic over non-native species, then we should start by wiping out all the pigeons in North and South America, please.
-- Terry
It occurs to me that there *is* a paredo optimal solution.
There has to be.
The problem is "P-P complete".
-- Terry
The author discussed, in true binary fashion, "available" vs. "unavailable" for a second roll.
People have also discussed "over the top vs. under the bottom".
There is a middle ground: two rollers, one "over the top" and the other "under the bottom". This would also self-regulate the usage of the rolls, ensuring one ran out before the other.
Which one runs out first depends on the physical characteristics of the holder.
For a fixed holder, the answer is probably "over the top".
But for a hanging holder, where the weight of the paper itself levers the paper into the wall, increasing the overall friction, "under the bottom" allows you to mitigate friction effects, while "over the top" increases them.
Since a fixed holder results in the choice being user preference, that doesn't solve anything; clearly, the fix is in two parts: (1) use hanging holders, and (2) make the primary roll "under the bottom" and the secondary roll "over the top".
Another solution (which is only statistical) is to locate rolls on either side of the toilet. Left handed people are outnumbered by right handed people 20 to 1, on average. But this fix only works "on average", as a result.
All in all, a "P-P complete" problem.
-- Terry
Your theory of public transportation is a game of statistics. I am supposed to be convinced that, at a particular population density, people will start using it.
The problem with this theory is that when I want to go home at 9:37 PM, I want to *go home* at 9:37 PM, not walk 4 blocks, wait 15 minutes for a bus, ride for a while, get off, wait another 15 minutes to transfer to another bus, ride for another while, get off, and walk another 4 blocks.
Public transportation is always a great idea *for other people, not for you*.
I've discussed this in letters with Governor Gray Davis of California. The only real way to get a lot of people onto public transportation is to make it fully subsidized.
The only way to get people to take public transportation is to make the marginal cost of taking it less than the marginal cost of not taking it. Making it free would drastically increase ridership. But there will still be people who will not take it, even if you were to pay them.
-- Terry
Why do conspiracy theroies abound?
It's pretty simple: it's very hard for an unintelligent person to credit stupidity for something that could have been the result of malice.
-- Terry
If you can build one, more power to you: do it. You will win in at leastone category, and may win overall for effectiely making fun of the making fun.
Given the cost, and the lack of a long enough launching/landing strip, I think it's a non-issue.
-- Terry
I have a theory, and the theory is mine.
Public transportation is a wonderful thing.
Everyone will use it.
The number of cars on roads will decrease.
The amount of air pollution will decrease.
Then I can drive my car in peace.
-- Terry
"Homeless people are not in our target market".
-- Terry
It's a joke.
You are not expected to fly.
It's more about "how many drunken idiots can we get to strap drowning machines to themselves and jump into the San Francisco Bay?"
The Gossamer Condor, which won the Kremer Prize in 1977 is, so far, the smallest human powered aircraft. It has a wingspan of 70 feet. The Gossamer Albatross, which crossed the English Channel, and won the second Kremer Prize, has a wingspan of 90 feet.
A wingspan of 30 feet is pretty mugh guaranteed to not result in flight, if the wings alone are intended to provide the lift, with just human power.
The weight limit lets out most practicaly designs, such as a neutral buoyancy ornithopter with a helium lifting baloon with a 30 foot maximum width (hint: search for "one person helium balloon").
-- Terry
I have discovered a truly marvelous method, which the margin of this Slashdot posting is too narrow to contain.
-- Terry
Prove to me they were personal, not corporate loans.
In other words, that the money was going into the director's pockets, instead of the pockets of the organization whose interests the directors were there to defend in the first place. This *really* is not different than a stock grant with private placement. It's better, because it doesn't dilute the shares that are the founders and the employee ISOs, like it would if it were a true private placement.
Realize... there *ARE* no other investors, than the people who put up the money. You can talk about "sweat equity", but it's not worth very much in the long run, unless there's a big upside because there are different classes of stock. Investors have preferred liquidation, and early investors have ratchet antidilution. If they didn't get these things, they would not invest, and there's be no company and no ISOs for the employees in the first place.
If it *was* into the directors pockets... it's no more a handout than a salary is a handout, if it was intended as tax deferred compensation represented as a loan (which is what personal loans mostly are).
Yeah, it's possible to rape a company this way, but it generally doesn't happen, because if it's for stock, the stock ends up being collateralized. The company only suffers if their stock loses value. There are laws against raping a company this way; it's called "embezzleing", so they tend to keep it as legal and above board as possible (particularly directors, who are legally culpable, and subject to shareholder lawsuits).
It's my understanding of the article that it was a loan for stock in another company. If so, it's clearly a case of bailing money out as a form of risk mitigation.
Note that the capital gains taxes in the U.S. have made it common practice for companies to sell unvested incentive stock options to the employee using a company guaranteed loan, on which interest is not paid until the stock has vested (simple answer: exercise stock as it becomes vested). The only downside is that if the company folds, the loans become debt instruments, like any other; but you can quit-claim the options, and still get out.
All of these are really tax dodges to get around the draconian tax laws that were only supposed to apply to the top 1% of income earners, but end up screwing the little guys.
I know tons of engineers who lost cars, houses, and even one who committed suicide over their exercise in Feb/March of 2000 of stock options that were so far under water by the time they left the lock-out period, that they owed literally millions of dollars as of January 1st, and things didn't get any better by April 15th of 2001.
It seems to me that it's a double standard, if it's "unethical and scummy" on one hand, and "the only way to avoid losing everything" on the other.
The real problem is that such contortions are necessary to protect little guys from tax laws which are blatantly anti-little-guy, for all of their intent to hurt only people for which no one has any sympathy.
-- Terry
The libertarians who found an underwater atol close to the surface in the tropis, and then built an island on top of it to get out from under the yoke.
Oh yeah... they failed, when the U.S. paid the Tongan Navy $1M to go plant a Tongan flag on it to keep it from happening.
Richer people than you have tried...
-- Terry
Apparently, you missed the part where I noted that it was unclear as to whether or not these were personal loans, or corporate loans for the purchase of alternate stock.
Personal loans are a tax dodge to get around compensation issues by turning a tax liability into a deferred liability with a short term asset.
This type of thing wouldn't be necessary, if there were no capital gains tax. As it is, between state and fed, 55% of all capital gains ends up in the government's pocket, otherwise, instead of funding new business, paying people's salaries, creating jobs, etc..
More likely, it was corporate, with the investors bailing out of a bad investments.
Yeah, it undermines the business in question, but it protects the money from some idiot who thinks there's such a thing as "page views" and "selling eyeballs".
It's the fiduciary responsibility to directors to protect the money of the investors they represent, and sometimes, this means taking it out of a bad investment, and putting it some place else.
Yeah, businesses fail over this. Euthanasia is better than the alternative. If you are looking for compassionate spending of millions of dollars to prop up a failed business model, you are dreaming.
-- Terry