Yeah, that won't cause much problem for bayesian. Essentially your filter will learn that news goes from good to neutral, and that javascript mouseovers go from bad to terrible.
However, this isn't what Joshua and the rest of MS are working on. His stuff is much more in the area of modifying SMTP so that untrusted clients have to perform some calculations before their email is accepted, or pay a few cents. My guess is it will fail since it doesn't account for zombie PCs but I'm sure he has something planned for them.
Firstly, they're the largest player in the industry. How on earth have they got themselves into the situation where they are 'incapable' of releasing specs? And isn't it funny how it only applies to gfx cards, not to TV tuner cards, sound cards, MPEG cards, SCSI cards, RAID cards,...
Secondly, notice they _never_ say what NDA prevents them releasing information. My bet is a PR person experimented and found that term caused the least friction with open source zealots, so they've stuck with it. They might even have convinced a.tw mfg to set up a NDA with them, making it true.
Thirdly, binary firmware is largely tolerated in linux. Any NDAed code could be trivially inserted in firmware instead of in the binary driver. Then the driver could be opened.
But the reality is that currently we have no choice. Buy NVidia and get good closed source drivers and awful open-source drivers, or buy ATI and get tolerable open-source drivers. If your principal concern is quality of driver, nvidia is the choice. If your principal concern is quality of open source driver then ATI is the choice. The 'open video card' project is apparently making good progress, but I've seen too much vapourware in the industry to put stock in anything that I can't buy right now.
One: Time has moved on since then -- one GHz laptops are no longer top of the line, but eclipse is not any more complex. That makes it run much more smoothly.
Two: Java interpreters have improved amazingly since then. Java 5 feels twice as fast to me. Even on your friend's 1GHz machine, I expect eclipse would feel comfortable now.
Three: I code for a living. How many other industries would balk at spending five percent of your gross income on your main tool? A $5000 computer is able to run eclipse without blinking as its dual 2.4GHz 64 bit processors compile classes in parallel, caching intermediate results and all the source files in 8 gigabytes of ram. My point? If eclipse makes you ten minutes more productive every day, that will pay for even an insanely powerful machine within a year.
Now, to be honest, I don't use eclipse because I'm so familiar with the editors I have been using for the past ten years that their key-bindings are second nature. However, I know people who do use it and are much more productive as a result. My point is that insane system requirements just don't matter when you're talking about a tool that determines your productivity.
If you are sued for infringement, then you have lost. End of story. The cost of invalidating a patent is so high that it cannot be afforded by anybody. So, if you are sued, you are screwed.
I would like to repeat this ten times to drill it into your head, but I will save space and just repeat it once. In the world we live in, being sued for patent infringement will destroy your business and your life. Ordinary small business do not have the resources necessary to invalidate any patents. The ONLY way to save them is to not grant them in the first place.
Now return to your scenario. Is it better to have a vague patent, or a non-vague one? Clearly the vague patent is better, because it is easier to start a lawsuit. Sure, you'll probably lose if they fight -- but who cares, only one in a hundred is stupid enough to fight.
I suppose you could argue that in a hypothetical world where the legal system was affordable, this would not be the case. Well, sorry. We don't live in a hypthetical world, and we write rules to fit with the world we live in.
Yeah, pretty much. Though in a few years time (five? less?) I think you'll see this flip. Analogue is more expensive to make, but you'll sell more because more people will be able to buy it. However soon a large enough percentage will have digital that you'll be able to sell an almost equal number and you'll see digital only on the cheapest options.
Computers aren't the only thing that works this way, try buying uniodised salt for instance, or blackcurrant concentrate without added sugar. It seems conventional manufacturing makes it cheaper to always do things the same way than it is to offer a less refined product for the few consumers that want it.
There was an announcement to/. about a USB based tuner for linux just the other month. Since mythtv is designed to have the tuner on a different machine it doesn't make a lot of sense to me. But I guess you might be in the market for it with your xbox.
I think it was USB2, though that seems unnecessary since e.g. my PVR500 encodes at 1MB/s (or actually only 900KB/s), under 12Mb/s anyway.
Also worth mentioning that most jurisdictions have sanity rules which say if an asset is too small to bother about then you don't have to amortise it. Servers at the end of their official life invariably fit into this category, so they aren't much of a headache from an accounting perspective.
Oh, groklaw has an agenda. However, they've very good about publishing enough details for you to a) make up your own mind and b) check the facts for yourself. You should visit it sometime.
Laura Didio is a paid shill, who writes whatever will give her the most hits, regardless of factual content. I've read tabloids with a better grasp of the truth than her.
According to the article, she's now claiming to have done yet another study which no doubt will get debunked within a couple hours but still cited by microsoft (sans the debunking) months later.
So, a simple question for everyone? Why bother debunking it? Anybody with more than half a brain already knows Didio is a paid liar, so she's not going to care if you drive a truck through her arguments. The other people on./ already know she's a shill, they won't care. And the PHBs reading the article via MS's links won't get to read your debunking. Why waste your time on her?
My feeling is that any criticism of dead people is very strongly frowned on. It doesn't matter if it is left wing, right wing, or if the person who died used to be an enemy (except for the straw-men enemies like saddam).
So, I'm not surprised that Bush came out saying nice things about him regardless of the papal eulogy was.
s/english/mandarin/ig and the argument still holds. Virtually any natual language is riddled with ambiguity, and virtually any natural language can have an unambiguious form.
Also remember that since, by definition, the computer can unambiguously parse the controlled english, it can trivially generate controlled mandarin/whatever.
Everybody can read and understand English, even PHBs. So, if we had a reliable C -> English translator, then even PHBs would be able to understand what a given function does (I doubt they'd be able to understand a complete system, since that involves holding together the interaction lots of functions.)
Rational (of ROSE fame) invented a language called 'Controlled English'. This is English with a formal lexicon and grammar, absolutely awful to write, but great for giving to PHBs to read.
While we cannot convert C to English as such, we can convert various languages (including English) to Controlled English. And we can convert Controlled English to C. Ergo, we can write English and produce C.
Now, I cannot see programmers being made redundant in any great hurry since the English to Controlled English process requires the standard set of programming skills, as does writing a full spec., regardless of it being in English. However, I can see it helping non-programmers to at least comprehend what programmers do.
You're right, but I think that it has pretty much changed already. I only really see pedants using come now, other people are too concerned with being understood so choose cum.
Besides, cum reduces ambiguity, so makes things easier.
My (limited) experience is that businesses with strong IT resources have all other IT management costs going down as well, to the point that money saved on license fees is just a small bonus. And that businesses with zero IT skills have those costs shoot up, swallowing all savings and then some.
Yep, that's why. Say you have two insurance companies, A Inc, and B Inc. A has the same rate for everybody, and B gives you a custom quote based on your chance of needing their services.
On average, B's quotes will be the same as A (or maybe a little higher since there is an extra cost in categorising people). However, for most people, B will be slightly cheaper (since 80% pay for the other 20%, and you can partially predict the 80%). If those people are smart, they'll swap from insurer A to insurer B.
Now A is left with a subset of the population which is higher risk on average. Unless they up their fees, they're going to go bankrupt. However, if they do up their fees, more average people are going to jump ship. Eventually, A will be left with crazily high fees but minimal checks. Ergo, those insurers that discriminated put those that didn't out of business a long time ago.
Of course, you could implement a government mandated scheme requiring people to stick with A, and therefore while the average person would have to pay a tiny bit extra, there would be no disaster-cases. Do you think that bill would wash at the moment?
You can't argue that it's more expensive to provide medical service for any given human when they're not working for a large firm.
No, not exactly, but you can argue something very similar. It is pretty normal to work for a large firm, so the health stats for such people will be average. But what about the health stats of people who aren't working for firms? Or for some reason aren't covered by the firm's insurance policy.
Of those people, every single one with medical problems will want to try and get insurance, but a lot of those without problems will want to risk it since they are less likely to have dependants, more likely to be on an immigrant visa (H1B is it?). Essentially, the average payout for
You see much the same thing in geriatric healthcare. Have a look at how much it costs for health insurance if you're over 70. As you'd expect, it is higher than four a 40 year old, but it is WAY higher, totally out of proportion to the expected illness rates at 70. The reason is simple: 70 year olds typically don't bother with health care at its high prices, unless they think they're very likely to get sick. So only the risky ones enrol, and so the insurers see they're making less than the standard profit/person and put fees up, etc.
I tend to view health care much like a lottery, with the insurer playing the role of running the lottery. They don't rip off anybody who chooses to buy a ticket, they just charge the tickets at such a rate that they make about the same profit on each person. If some class of people are more likely to get sick then the expected payout for that class is higher, so their ticket price will be higher.
That's not to say there might be classes they discriminate against because they can. I vaguely know a statistican who was called as an expert witness in such a case to show the insurer was charging a fee out of proportion to the increased risk. That's just ordinary corporate greed, and is still grounds for a civil suit.
But before you think "the insurers are charging me stupid amounts for insurance", think about what bucket the insurers will be lumping your claim in with. Will it be in with other rich people who will pay whatever they charge? With lots of ex-middle-class people who can't quite give up their safety net after losing their jobs? Usually the insurers have a statistical reason for the fees they derive, and most of it depends on what bucket you're classified into.
Yeah, they exist, but they're f'ing expensive. Unless you're using windows or you're so rich that having proper solid state is worth paying an 50% markup for, you are much better putting more ram into your computer. Windows is to dumb to properly cache the data so you have to go with a drive, but for anything else...
Just get an A64 using a decent tyan MB and you should be able to get 64GB no trouble. The comp I was using yesterday (IBM PPC) had 16GB using just stock components (2GB * 8), which would've made it a fair bit cheaper.
It's as if individuals are subsidizing businesses. It really stinks.
Nah, it's just stats. What kind of company gets health insurance for their employees? Pretty much all of them right? So any company thinking of it is low risk. Odds are, they're going to be just like the others. However, what kind of person gets personal health insurance instead of going through their employer. Immediately you have to start justifying why, and that means they're high risk.
As a related anecdote, when I left home I wanted to insure my stuff when moving since that wasn't covered by my parent's policy. It cost the same for that one day's coverage as it would for me to pay for a year's coverage including moving. Risk again...
Not that I approve of the system, but you can't blame the insurance companies for charging more to individuals. They're just looking at the stats coming in and charging accordingly.
I think the Dutch don't like the idea that an American company would come along, start selling things locally, and not even bother to monitor local news enough to know they'd been charged.
Essentially, the MS lawyer managed to change it from "We didn't tell them" to "We did what we were required to do, but they weren't paying enough attention to this country to be here".
Remember that here, "windows" is a generic term and so rather hard to hold a trademark on, but "windows" is quite different to "fenster", and so it is much harder to see the difference to "lindows".
Still, it was clear MS were going to lose their trademark on windows if the case continued in the states, hence the settlement.
Lindows was found to be NOT infringing trademark. So, Microsoft sued them in Sweeden, on the grounds that Sweedish users could go to the site too. Microsoft lost, again. Then Microsoft sued them in Germany, over the same issue, and lost, again. Next microsoft sued them in the Netherlands and didn't even bother telling them about it, so they didn't turn up to the court and Microsoft won by default (apparently posting a notice in the local paper is sufficient in the Netherlands). Almost certainly it would be overturned on appeal.
But, how many countries does Microsoft have offices in? How much would it cost to win the same lawsuit again, and again, and again? So, they settled with Microsoft paying them around $20M to change their name.
I can see your point; I'm going to ignore it and give you a different viewpoint.
I've run a few classes, though sadly I'm not currently employed full-time doing it. When running those courses I got to choose a text, and a wide range of samples were 'generously' donated to me by the publishers. Each of those texts had their merits, but none of them fitted exactly how I wanted to run the course.
So I picked what I thought was best and supplimented it with my own notes when my course diverged too much from the text. I think this is the key thing undergrads tend to miss -- a textbook is designed to cover a topic and is very unlikely to have the same coverage that the course is supposed to. Perhaps bits are aimed too high, too low, not considered important, or just covered by other courses being run in the department.
Regardless, the textbook was better written than what I could achieve, but it not sticking to the course would have caused students some trouble. Had I kept running that course, I expect I would have gradually expanded my notes. And if I'd been particularly dedicated, I may well have combined them into a 'textbook' and sold it. This text would then fit the course perfectly, and it would change year-to-year as the course adapted to fit into my changing teaching style, changes in student demand, changes in the other courses, etc. Oh yeah, and maybe a little because the subject has moved on.
So, the point I'd like you to get is that most profs I know writing their own texts are doing it because they're experienced enough teachers that they don't want a textbook telling them how to structure their course. And they'd rather have a textbook fitting their course perfectly than have a slightly better textbook.
Finally, the only person I know who has got rich off writing textbooks was not doing it for their own class. The market is just too small.
Since we don't have trusted computing implemented yet, it is hard to say definitively. Given current chips, it is possible for the answer to be yes, and also possible for the answer to be no. Therefore in answering your question I am second-guessing how microsoft will set the defaults rather than what the chip can do.
My guess as to the most likely scenario is that you will be able to boot knoppix. I just cannot imagine the amount of pressure the linux community brings to bear if this happens. The TC chip will however PROBABLY not permit knoppix to unlock the hard drive. So at this point you're kinda stuck. It is likely you'll get low-level access to the disk, but dumping encrypted data isn't of much use.
It is possible the TC chip will trust knoppix once the user enters their password, in which case your recovery would go just fine. It is also possible there is a back door and if you send the encrypted partition to the NSA they will unlock it for you, though I doubt they'll offer the service even if it is possible.
Slashdot Mirror
Yeah, that won't cause much problem for bayesian. Essentially your filter will learn that news goes from good to neutral, and that javascript mouseovers go from bad to terrible.
However, this isn't what Joshua and the rest of MS are working on. His stuff is much more in the area of modifying SMTP so that untrusted clients have to perform some calculations before their email is accepted, or pay a few cents. My guess is it will fail since it doesn't account for zombie PCs but I'm sure he has something planned for them.
Sadly, you've bought into their FUD.
...
.tw mfg to set up a NDA with them, making it true.
Firstly, they're the largest player in the industry. How on earth have they got themselves into the situation where they are 'incapable' of releasing specs? And isn't it funny how it only applies to gfx cards, not to TV tuner cards, sound cards, MPEG cards, SCSI cards, RAID cards,
Secondly, notice they _never_ say what NDA prevents them releasing information. My bet is a PR person experimented and found that term caused the least friction with open source zealots, so they've stuck with it. They might even have convinced a
Thirdly, binary firmware is largely tolerated in linux. Any NDAed code could be trivially inserted in firmware instead of in the binary driver. Then the driver could be opened.
But the reality is that currently we have no choice. Buy NVidia and get good closed source drivers and awful open-source drivers, or buy ATI and get tolerable open-source drivers. If your principal concern is quality of driver, nvidia is the choice. If your principal concern is quality of open source driver then ATI is the choice. The 'open video card' project is apparently making good progress, but I've seen too much vapourware in the industry to put stock in anything that I can't buy right now.
Hiya, three points.
One: Time has moved on since then -- one GHz laptops are no longer top of the line, but eclipse is not any more complex. That makes it run much more smoothly.
Two: Java interpreters have improved amazingly since then. Java 5 feels twice as fast to me. Even on your friend's 1GHz machine, I expect eclipse would feel comfortable now.
Three: I code for a living. How many other industries would balk at spending five percent of your gross income on your main tool? A $5000 computer is able to run eclipse without blinking as its dual 2.4GHz 64 bit processors compile classes in parallel, caching intermediate results and all the source files in 8 gigabytes of ram. My point? If eclipse makes you ten minutes more productive every day, that will pay for even an insanely powerful machine within a year.
Now, to be honest, I don't use eclipse because I'm so familiar with the editors I have been using for the past ten years that their key-bindings are second nature. However, I know people who do use it and are much more productive as a result. My point is that insane system requirements just don't matter when you're talking about a tool that determines your productivity.
If you are sued for infringement, then you have lost. End of story. The cost of invalidating a patent is so high that it cannot be afforded by anybody. So, if you are sued, you are screwed.
I would like to repeat this ten times to drill it into your head, but I will save space and just repeat it once. In the world we live in, being sued for patent infringement will destroy your business and your life. Ordinary small business do not have the resources necessary to invalidate any patents. The ONLY way to save them is to not grant them in the first place.
Now return to your scenario. Is it better to have a vague patent, or a non-vague one? Clearly the vague patent is better, because it is easier to start a lawsuit. Sure, you'll probably lose if they fight -- but who cares, only one in a hundred is stupid enough to fight.
I suppose you could argue that in a hypothetical world where the legal system was affordable, this would not be the case. Well, sorry. We don't live in a hypthetical world, and we write rules to fit with the world we live in.
Yeah, pretty much. Though in a few years time (five? less?) I think you'll see this flip. Analogue is more expensive to make, but you'll sell more because more people will be able to buy it. However soon a large enough percentage will have digital that you'll be able to sell an almost equal number and you'll see digital only on the cheapest options.
Computers aren't the only thing that works this way, try buying uniodised salt for instance, or blackcurrant concentrate without added sugar. It seems conventional manufacturing makes it cheaper to always do things the same way than it is to offer a less refined product for the few consumers that want it.
There was an announcement to /. about a USB based tuner for linux just the other month. Since mythtv is designed to have the tuner on a different machine it doesn't make a lot of sense to me. But I guess you might be in the market for it with your xbox.
I think it was USB2, though that seems unnecessary since e.g. my PVR500 encodes at 1MB/s (or actually only 900KB/s), under 12Mb/s anyway.
Also worth mentioning that most jurisdictions have sanity rules which say if an asset is too small to bother about then you don't have to amortise it. Servers at the end of their official life invariably fit into this category, so they aren't much of a headache from an accounting perspective.
Oh, groklaw has an agenda. However, they've very good about publishing enough details for you to a) make up your own mind and b) check the facts for yourself. You should visit it sometime.
If you haven't heard of her, you're doing well.
a DiDio
i o&keyType=phrase&datestart=&dateend=&topic=0&type= all&author=0&mode=search
1 15008560&query=Laura+Didio 0 40251888&query=Laura+Didio
Perhaps the best summary:
http://twiki.iwethey.org/twiki/bin/view/Main/Laur
Or for a more broad-brush approach, here's the output of a search for her name on groklaw
http://www.groklaw.net/search.php?query=Laura+Did
And some articles specifically discussing her:
http://www.groklaw.net/article.php?story=20040422
or http://www.groklaw.net/article.php?story=20040103
Laura Didio is a paid shill, who writes whatever will give her the most hits, regardless of factual content. I've read tabloids with a better grasp of the truth than her.
./ already know she's a shill, they won't care. And the PHBs reading the article via MS's links won't get to read your debunking. Why waste your time on her?
According to the article, she's now claiming to have done yet another study which no doubt will get debunked within a couple hours but still cited by microsoft (sans the debunking) months later.
So, a simple question for everyone? Why bother debunking it? Anybody with more than half a brain already knows Didio is a paid liar, so she's not going to care if you drive a truck through her arguments. The other people on
An 'actual newspaper' press doesn't have an axe to grind about blogging? Yeah, right.
My feeling is that any criticism of dead people is very strongly frowned on. It doesn't matter if it is left wing, right wing, or if the person who died used to be an enemy (except for the straw-men enemies like saddam).
So, I'm not surprised that Bush came out saying nice things about him regardless of the papal eulogy was.
Just look at how well paid those few cobol programmers are...
Personally I'd hate your job, but you do have a good point.
s/english/mandarin/ig and the argument still holds. Virtually any natual language is riddled with ambiguity, and virtually any natural language can have an unambiguious form.
Also remember that since, by definition, the computer can unambiguously parse the controlled english, it can trivially generate controlled mandarin/whatever.
Nice post. I'll just present one counter point.
Everybody can read and understand English, even PHBs. So, if we had a reliable C -> English translator, then even PHBs would be able to understand what a given function does (I doubt they'd be able to understand a complete system, since that involves holding together the interaction lots of functions.)
Rational (of ROSE fame) invented a language called 'Controlled English'. This is English with a formal lexicon and grammar, absolutely awful to write, but great for giving to PHBs to read.
While we cannot convert C to English as such, we can convert various languages (including English) to Controlled English. And we can convert Controlled English to C. Ergo, we can write English and produce C.
Now, I cannot see programmers being made redundant in any great hurry since the English to Controlled English process requires the standard set of programming skills, as does writing a full spec., regardless of it being in English. However, I can see it helping non-programmers to at least comprehend what programmers do.
You're right, but I think that it has pretty much changed already. I only really see pedants using come now, other people are too concerned with being understood so choose cum.
Besides, cum reduces ambiguity, so makes things easier.
My (limited) experience is that businesses with strong IT resources have all other IT management costs going down as well, to the point that money saved on license fees is just a small bonus. And that businesses with zero IT skills have those costs shoot up, swallowing all savings and then some.
Yep, that's why. Say you have two insurance companies, A Inc, and B Inc. A has the same rate for everybody, and B gives you a custom quote based on your chance of needing their services.
On average, B's quotes will be the same as A (or maybe a little higher since there is an extra cost in categorising people). However, for most people, B will be slightly cheaper (since 80% pay for the other 20%, and you can partially predict the 80%). If those people are smart, they'll swap from insurer A to insurer B.
Now A is left with a subset of the population which is higher risk on average. Unless they up their fees, they're going to go bankrupt. However, if they do up their fees, more average people are going to jump ship. Eventually, A will be left with crazily high fees but minimal checks. Ergo, those insurers that discriminated put those that didn't out of business a long time ago.
Of course, you could implement a government mandated scheme requiring people to stick with A, and therefore while the average person would have to pay a tiny bit extra, there would be no disaster-cases. Do you think that bill would wash at the moment?
You can't argue that it's more expensive to provide medical service for any given human when they're not working for a large firm.
No, not exactly, but you can argue something very similar. It is pretty normal to work for a large firm, so the health stats for such people will be average. But what about the health stats of people who aren't working for firms? Or for some reason aren't covered by the firm's insurance policy.
Of those people, every single one with medical problems will want to try and get insurance, but a lot of those without problems will want to risk it since they are less likely to have dependants, more likely to be on an immigrant visa (H1B is it?). Essentially, the average payout for
You see much the same thing in geriatric healthcare. Have a look at how much it costs for health insurance if you're over 70. As you'd expect, it is higher than four a 40 year old, but it is WAY higher, totally out of proportion to the expected illness rates at 70. The reason is simple: 70 year olds typically don't bother with health care at its high prices, unless they think they're very likely to get sick. So only the risky ones enrol, and so the insurers see they're making less than the standard profit/person and put fees up, etc.
I tend to view health care much like a lottery, with the insurer playing the role of running the lottery. They don't rip off anybody who chooses to buy a ticket, they just charge the tickets at such a rate that they make about the same profit on each person. If some class of people are more likely to get sick then the expected payout for that class is higher, so their ticket price will be higher.
That's not to say there might be classes they discriminate against because they can. I vaguely know a statistican who was called as an expert witness in such a case to show the insurer was charging a fee out of proportion to the increased risk. That's just ordinary corporate greed, and is still grounds for a civil suit.
But before you think "the insurers are charging me stupid amounts for insurance", think about what bucket the insurers will be lumping your claim in with. Will it be in with other rich people who will pay whatever they charge? With lots of ex-middle-class people who can't quite give up their safety net after losing their jobs? Usually the insurers have a statistical reason for the fees they derive, and most of it depends on what bucket you're classified into.
Yeah, they exist, but they're f'ing expensive. Unless you're using windows or you're so rich that having proper solid state is worth paying an 50% markup for, you are much better putting more ram into your computer. Windows is to dumb to properly cache the data so you have to go with a drive, but for anything else...
Just get an A64 using a decent tyan MB and you should be able to get 64GB no trouble. The comp I was using yesterday (IBM PPC) had 16GB using just stock components (2GB * 8), which would've made it a fair bit cheaper.
It's as if individuals are subsidizing businesses. It really stinks.
Nah, it's just stats. What kind of company gets health insurance for their employees? Pretty much all of them right? So any company thinking of it is low risk. Odds are, they're going to be just like the others. However, what kind of person gets personal health insurance instead of going through their employer. Immediately you have to start justifying why, and that means they're high risk.
As a related anecdote, when I left home I wanted to insure my stuff when moving since that wasn't covered by my parent's policy. It cost the same for that one day's coverage as it would for me to pay for a year's coverage including moving. Risk again...
Not that I approve of the system, but you can't blame the insurance companies for charging more to individuals. They're just looking at the stats coming in and charging accordingly.
I think the Dutch don't like the idea that an American company would come along, start selling things locally, and not even bother to monitor local news enough to know they'd been charged.
Essentially, the MS lawyer managed to change it from "We didn't tell them" to "We did what we were required to do, but they weren't paying enough attention to this country to be here".
Remember that here, "windows" is a generic term and so rather hard to hold a trademark on, but "windows" is quite different to "fenster", and so it is much harder to see the difference to "lindows".
Still, it was clear MS were going to lose their trademark on windows if the case continued in the states, hence the settlement.
Lindows was found to be NOT infringing trademark. So, Microsoft sued them in Sweeden, on the grounds that Sweedish users could go to the site too. Microsoft lost, again. Then Microsoft sued them in Germany, over the same issue, and lost, again. Next microsoft sued them in the Netherlands and didn't even bother telling them about it, so they didn't turn up to the court and Microsoft won by default (apparently posting a notice in the local paper is sufficient in the Netherlands). Almost certainly it would be overturned on appeal.
But, how many countries does Microsoft have offices in? How much would it cost to win the same lawsuit again, and again, and again? So, they settled with Microsoft paying them around $20M to change their name.
Trademarks on generic words are extremely weak.
I can see your point; I'm going to ignore it and give you a different viewpoint.
I've run a few classes, though sadly I'm not currently employed full-time doing it. When running those courses I got to choose a text, and a wide range of samples were 'generously' donated to me by the publishers. Each of those texts had their merits, but none of them fitted exactly how I wanted to run the course.
So I picked what I thought was best and supplimented it with my own notes when my course diverged too much from the text. I think this is the key thing undergrads tend to miss -- a textbook is designed to cover a topic and is very unlikely to have the same coverage that the course is supposed to. Perhaps bits are aimed too high, too low, not considered important, or just covered by other courses being run in the department.
Regardless, the textbook was better written than what I could achieve, but it not sticking to the course would have caused students some trouble. Had I kept running that course, I expect I would have gradually expanded my notes. And if I'd been particularly dedicated, I may well have combined them into a 'textbook' and sold it. This text would then fit the course perfectly, and it would change year-to-year as the course adapted to fit into my changing teaching style, changes in student demand, changes in the other courses, etc. Oh yeah, and maybe a little because the subject has moved on.
So, the point I'd like you to get is that most profs I know writing their own texts are doing it because they're experienced enough teachers that they don't want a textbook telling them how to structure their course. And they'd rather have a textbook fitting their course perfectly than have a slightly better textbook.
Finally, the only person I know who has got rich off writing textbooks was not doing it for their own class. The market is just too small.
Since we don't have trusted computing implemented yet, it is hard to say definitively. Given current chips, it is possible for the answer to be yes, and also possible for the answer to be no. Therefore in answering your question I am second-guessing how microsoft will set the defaults rather than what the chip can do.
My guess as to the most likely scenario is that you will be able to boot knoppix. I just cannot imagine the amount of pressure the linux community brings to bear if this happens. The TC chip will however PROBABLY not permit knoppix to unlock the hard drive. So at this point you're kinda stuck. It is likely you'll get low-level access to the disk, but dumping encrypted data isn't of much use.
It is possible the TC chip will trust knoppix once the user enters their password, in which case your recovery would go just fine. It is also possible there is a back door and if you send the encrypted partition to the NSA they will unlock it for you, though I doubt they'll offer the service even if it is possible.