This article is identical to what we covered... in 300 level Modern Physics in college in 1983.
I don't see how this is new or radical, except for the general population, who seem to think that for every "black hole" there is a corresponding "white hole", or that when you "fall into a black hole", you somehow end up somewhere else.
You should read Feynman's lecture series; he has one from the 50's that debunks the idea of a "graviton" or a particulate carrier for gravity because of the need for it to have mass.
Why are patch deployment costs on Windows so high? The only real rationale for this on the Wikipedia page you reference is "a patch issued by Microsoft would break existing functionality", and that's a matter of code, not physics constraints.
-- Terry
This feature is not called "shuffle"...
on
FCC Approves iPhone
·
· Score: 2, Funny
This feature is not called "shuffle"... it's called iDrunkDial.
I call "bullshit"... this is the union guys again.
I'm a former IBM Global Services employee; I left IBM on my own steam to start my own startup when they were closing down our division in 2001 (IBM Global Small Business - say that with a straight face). When that happened, they gave each of us six months to find a job elsewhere within IBM, including paying for flights for interviews, and sending HR people out on site to help write resumes, etc.. IBM bends over backwards to retain employees. IBM is not just a job, it's a job for life for most of its employees.
Here's the original "The Register" article which set off this current firestorm:
Notice who the press release is from: "Alliance@IBM, a group working to form a union at the company, reports that..."; this is The Communication Workers of America Union; here's their site:
They claim to be "The official national site for the IBM Employees' Union CWA Local 1701, AFL-CIO" - only there *is no* "IBM Employees' Union" (nowhere on this site will you find membership statistics of any kind).
These are the guys that have been trying to unionize IBM since smelling a payday, in order to get their claws into everyone's paycheck at $10/month and control of the IBM pension plans. When the IBM pension plan converted over to a cash-balance plan, about the only people who didn't have a choice were people who hadn't vested in the plan at all. Yet these same guys tried to stir up a firestorm about it, and attacked IBM over it in the press.
These guys are all about trying to FUD IBM employees into joining their union; while I was with IBM, no one was interested in their party line, even though they practically camped out in all the off-campus restaurants and tried to chat us up every chance they got - they were worse than religious missionaries.
It's too bad Cringely bought into this fake press release, I genuinely enjoyed his "Triumph of the Nerds" television programs.
(1) Insurance companies: - They get paid by you - They get paid by your employer - They get paid for malpractice insurance for the doctors - They get paid for liability insurance by the hospitals - They get paid for liability insurance by medical equipment manufacturers
Your insurance costs are so high because the insurance companies are charging the people that your insurance pays so much for their own insurance. And the costs for everyone else are so high because malpractice lawsuits pay out so much money; engage in tort reform for malpractice to cap damage claims to reasonable levels.
(2) Billing: - They don't present you with your actual bill (deductible plus copay) at time of service, when you'd be grateful to pay - Insurance companies pay "net 90", which means they don't pay until the account has practically gone to collections - Nobody pays the first bill because of this - Some people never pay their bill because of this - "Financial services" companies buy the paper at ~80 cents on the dollar, if the seller of the commercial paper gets a good deal (there's 20% cost right there, and it's a higher percentage than that, if the seller doesn't get a good deal)
This would all go away if there were a single entity that all billing had to go through such that insurance companies and hospitals and doctors were _forced_ to talk electronically, and didn't get paid if they didn't bill at time of service.
-
Want to fix the health care system? Fix the law and fix the money flow itself
The upshot of this is that you are potentially legally liable for consequences arising from someone taking and using the code. This includes, but is not limited to:
- implied or statutory warranties
- legal liability for direct, indirect, incidental, special, consequential, or other damages
Putting something in the public domain unfortunately does not protect you from this by making it truly public property. Instead, you open yourself up to a world of legal hurt - particularly in countries like the U.S., which has more lawyers than could be kept gainfully employed otherwise.
It's unfortunate, but current tort law has made it such that there is (effectively) no such thing as a Commons. It won't be until nearly 2030 until the first piece of software without this liability risk comes off Copyright; I doubt that when it does, you will be able to get useful source code for it anyway.
- US Citizenship - Ability to travel - BA or BS from a community college accredited by ABET _or_ 2nd year coursework in 5 of 7 areas: physics, chemistry, architecture, computer science, mathematics, hydrology, or geology - Registered as a professional engineer by a state, DC, Guam, or Puerto Rico - Pass a written test for "Engineer In Training" or professional registration test - 60 semester hours of courses in basic sciences/physics/math/engineering
So you've got to really want to be a patent examiner, be willing to live with a salary far below what you'd get in private industry with the same paper qualifications.
And then you get to do the scut work for a couple of years.
If you want to have your pay grade go up, you need time in grade and even more qualifications.
So it's pretty much the same deal that entry level teachers get, only you don't get the summer off.
It is not always possible, when there is other traffic around, to tell whether the person in front of you is slowing down or not (look up "retrograde motion" on Wikipedia). Brake lights warn the other driver.
If the jackass had obeyed the law and stuck his left arm out the window, I *also* would have known he was stopping, and come to a stop.
So in the case in question, the pickup truck I was following was in violation of two laws in Utah, and would have failed inspection and the state would not have renewed registration over the brake lights (same for Arizona, for that matter).
I'm actually constantly amazed that California has turned registration from an opportunity to improve safety (by conditionalizing registration renewal on safety inspections of all vehicle systems, as other states do), into simple revenue collection.
But to carry your original point to its logical conclusion:
What's a safe following distance for a car which you don't know whether or not they have antilock brakes? Clearly, it's safer to follow closer if they don't have them, since if they have them, then they will have a shorter stopping distance.
Or even better... what's a safe following distance for an enclosed truck with a tare weight of 2000 pounds, a load capacity of 6000 pounds, and therefore brakes capable of stopping 8000 pounds? I guarantee you that if the truck is unloaded, its stopping distance is going to be far shorter than if it were fully loaded, by a significant amount.
A reductio ad absurdum of this argument, you would apparently have all cars at least one block apart so that they weren't "tailgating": there's ALWAYS a reason to stay farther back than you are ("maybe the driver in front of me is drunk, and I just can't tell for sure from a safe distance").
Defensive driving is a good baseline, but at some point, you have to assume that the other drivers on the road will be cooperating with each other to try and drive safely; it would be impossible to drive next to someone on a multilane highway - the distance between cars is *far* shorter than the difference in stopping distance or the time absorbed by reaction time, and after all, they might swerve into you.
"I don't understand the problem you had. Your clutch got stuck. Could you pull the transmission out of gear without depressing the clutch?"
The clutch would not disengage (the actual failure was a shear in the clutch plate). As a result, the car could not be forced out of gear, so the engine could not be disengaged from the wheels, without turning off the ignition.
"Did your brakes not work? Most brakes are much stronger than the engine."
That depends on whether you are in a low gear or a high gear (overdrive). If you are in a low gear, the best you can do is burn your breaks and go forward anyway. My brakes worked, but it took time to evaluate why the clutch was not disengaging, try to pop it out of gear anyway (to hell with the gears grinding!), discover that that wasn't going to work, and decide whether or not to ride it out. I had attempted to start decellerating as soon as the light turned yellow, and had about 4 seconds to make all these evaluations and decisions (I did very well in this: I considered 4 options, and average human response time is 2 seconds). If the transmission hadn't been locked into gear, or I had been able to pop it out of gear, I could have stopped before it went red; when the second failed, there was no chance of it.
"Did you sue the mechanic/part manufacturer?"
They redid the repair at no cost, covering towing and rental. The best I could get them on is $390 for the ticket. It would cost far far more than that to take them and the manufacturer of the part to court (both would have to be a party to the case - I can explain the tort liability, if you want me to... though I'm not a member of the bar, I know the law), get expert testimony, etc. etc..
After that, I would (and do) still have the points on my license, and my insurance rate would still go up (as it did).
"Of course there is. Tickets are not issued by a camera, they are issued by the court system. Go to court and try to explain. If the court system in California doesn't accept mechanical failure as an excuse, the fact that you got caught by a camera is irrelevant. A judge would accept a cop's testimony the same way."
I *did* go to court, and argued it to the judge. A cop would have exercised human judgement and not issued the ticket; an officer testified to that (I entered the intersection 1 second after the light turned red). It didn't matter: the camera, lacking human judgement, caused the ticket to be issued, despite what a reasonable person or officer would consider.
So yeah, *if* the officer was an asshole instead of a reasonable person, the court would have accepted his testimony and enforced the rule, but *if* human judgement had been exercised instead of an inhuman machine process, the question of enforcement of the rule would never have arisen.
In California, as in most states, you are not permitted to have a jury trial on traffic violations, or I would have insisted (the only way I could have done this is to follow it through appeal - which as I said before, I simply don't have the time to do right now.
Like I siad: until someone challenges the legality of these things through an appeals process, there's no chance. And I suspect that if it ever came to it, they'd drop the charges on the person challenging, to avoid setting a precedent that would bite into their automated revenue stream (think about how RIAA handles cases, and why the suit for civil damages and legal costs going forward, even though they "decided not to sue", is such a strongly watched case).
"Driver in front never at fault" laws are plan stupid.
Utah has such a law, and I was ticketed there (maybe 16 years ago?) for rear-ending someone; so to answer your question:
"So who the fuck else could it be that was at fault? Santa Claus? The tooth fairy? The devil made me do it? I have a really hard time believing that you seriously mean that it's the driver in the front cars fault that somebody decides to run into him."
It was the fault of the asshole in front of me with the broken brake lights who didn't maintain his vehicle.
I.e.: the guy in front.
I still got the ticket, because that's the law (the officer had no choice), but I was able to fight (and win) in court as a result.
But it cost me the use of the bailment I had to pay until the court heard the case, the use of my vehicle and the cost of a rental car while the case was pending (the insurance would not pay for repairs or a rental if I was at fault), and a day in court -- all over a ticket which should never have been issued to me, but for the utterly stupid law.
I had had my vehicle repaired, and the repair was done with a faulty part.
When my clutch stuck the car in gear so that my choices were to run the red light or run off the bridge on the curve immediately on the other side of the intersection because my steering wheel lock engaged from turning off my ignition, I chose running the light and not dying over running the light anyway and dying.
California doesn't allow that a mechanical breakdown could be a mitigating factor in any moving violation, and so it was a $390 fine and points on my license. There is no arguing this.
Even if I agree with this on the theory that vehicle maintenance is the operators responsibility, I _had_ the vehicle maintained by a certified auto mechanic: it was a defective part that caused the malfunction.
I'm convinced that a reasonable police officer in the same situation would have called a tow truck when I finally got the vehicle to a safe point where I could turn off the ignition with the wheels pointing more or less straight, but there is no appeal against an unreasonable camera.
I wish I had had the ability (time) to follow this through an appeals process to case law (just on principle, but I didn't.
I paid the fine: there's pretty much zero appeal for these revenue collection devices, no matter what your story, at least in California.
"Do unto others as they would have you do unto them".
Reference: Andrew G. Haley "Space Law and Metalaw - A Synoptic View." Rome: Associazione Italiana Razzi Proceedings of the Seventh International Astronautical Congress, 1956.
It only tests memory under conditions that will only allow it to find real problems with bits in the memory. These include cross-talk related problems from adjacent memory cells, actual stuck bits, paired bits, and so on.
It does NOT test the memory for undervoltaging (which you have to test by actually dropping the voltage to the RAM), nor does it test for delayed DRAM referesh (which you could theoretically do on a PC, if you had a PCI card, and it forced an excessive bus-on time).
In short, software testing is not very good for transient failure testing, and it's not a good thing to include in an OS, rather than as a standalone diagnostic program, due to it being unable to test things that are wired down by the OS (unless your OS can relocate physical pages out of the way to allow you to test them, as well, then running a memory tester under the OS itself means that you don't test any of the memory that's wired down to the OS or to the test tool, if that memory's also wired).
Unfortunately, the most common memory failures in Macs occur during voltage stepping, speed stepping, or when other devices cause the machine to exceed its available power budget.
In other words, problems happen when RAM is run right to the edge of (but not over) its specifications.
You can't do most of that on purpose with normal machines, and not any of it, if it's not standallone.
Software memory checking is OK for gross problems - those problems that usually are not transient - but since most memory you buy has passed that sort of testing before it was sold to you, you are unlikely to ever receive RAM with permanent problems, unless you fail to observe proper precautions when handling or installing it.
So go ahead and use memtestosx; you may in fact find a problem... you're just exceedingly unlikely to find any of the problems that typically differentiate Apple-certified vs. non Apple-certified RAM.
Why you want Apple RAM... or at least certified RAM:
If you insist on third-party RAM, then at least get it in writing that if you have issues that the third party will replace it until you don't have problems, or take it back if they can't get you RAM that performs in spec. (someone else in this article already posted about how they ended up with some nice RAM for one of their PC boxes when it turned out it was out of spec. enough that it wouldn't work in their Mac).
Most third party RAM that fails qualification does so because it's technically out of manufacturer spec.. For many platforms that doesn't tend to matter much, but for Macintosh machines, this typically means problems like drawing current beyond the power budget on wake-from-sleep (and causing the system to be unstable) or being unable to step-up up from a stepped-down clock rate in the necessary time window when the rest of the system is stepping-up, etc. - things that cause the contents of the memory to end up unreliable, or cause the memory bus to be clocked down, and the system run slower.
Unsurprisingly, IBM Power architecture systems tend to have the same issues, given the way multiple simultaneous in-flight transactions are allowed on the memory bus, after the lines and latches have been timed out by the system. When you tune a system that closely, then the electrical characteristics of the components start to matter more and more.
I can't tell you how many customer reported "kernel bugs" I've personally tracked down to actually being faulty third party RAM.
Short of building in a hardware memory tester, which would be prohibitively expensive (including over/under clocking and under/over voltage/current testing, slow refresh cycles, etc.), there's no way to tell if a memory problem exists, short of running into a problem, and throwing the sticks into a real hardware memory tester (it's an NP incomplete problem if you are running test software in the RAM being tested, since it can malfunction and give you a false "no problem" result).
Certified memory is memory that has been thrown into a hardware memory tester, and certified to be in spec. before it's sold to the consumer.
Other peripherals and add-ons are much less sensitive to close-to-spec/out-of-spec problems, compared to RAM. Feel free to take chances with them long before you take chances with third party RAM.
There are people who cannot grasp the difference between "a thing" and "a respresentation of a thing". In fact, there was an entire century where almost the entire school of French philosophers were unable to grasp that there was a difference.
For pointers, it comes down to realizing that a pointer to memory is not the memory, it's a representation, and the real memory lives somewhere else: at the dereference of the pointer. And at that location it takes up real space. And so does the representation (but it's comparatively tiny).
The ability to make this disctinction comes naturally when you've written assembly code of any complexity, since you've had to manage the memory yourself. But if you've never done that, or, worse, you've only ever programmed in languages which eschew pointers and try to pretend that their own internal implementations don't use them, even if they won't let you use them, well, then, you've got a conceptual problem. An that conceptual problem is going to be very hard for you to overcome.
It also makes it very, very hard for someone to understand row-major vs. column major languages, or how you would link a program in one against a library in another, and be able to usefully communicate your data between the two. Or it makes it hard to understand the difference between "call by value" and "call by reference.
If I'm interviewing someone, and it becomes clear that they need a "code interview" because some things don't add up between their banter and their resume, the first thing I whip out is a pointer problem.
"If things worked out the way you are supposing, then BSD or equiv. would be the preferred license rather than the GPL."
I'm not the guy you were replying to, but I did get something back from the BSD license... I got an Internet.
The reason TCP/IP is so widespread isn't because AT&T and Novell didn't try to strike a deal to make sure the Internet backbone ran SPX and IPX, or because Microsoft didn't try to get everyone to use NetBEUI, it's because a DARPA grant covered some research, and because the code that resulted from that was under a BSD license, everyone could grab it and use it for free. And when everyone did exactly that, things interoperated.
Just to point out the obvious, but the vast majority of inventors are corporate employees who assign the rights for anything they invent over to the company that's paying their paycheck before they invent anything. If they are lucky, then they work for someone who gives them a plaque or a small bonus when they get a patent issued. So the benefit to the actual inventor is minimal. If they are really lucky, their company collects patents for defensive purposes, rather than beating people over the head with the patent bat to stifle competition. Such defensive portfolios tend to FUD the small guy out of business before they start.
Even with a defensive portdfolio, the companies actually producing product are effectively stripped of their defensive capability by patent holding companies, who don't counter-infringe, and so are not open to negotiation because they produce *absolutely nothing of value*.
So as for reform, you would need to start by disallowing third party assignment (which is how these patent holding companies get the vast majority of their patents).
Then rather than going off half-cocked with a "throw it all away" approach. see what happens after that and maybe one or two other small adjustments, before turning the patent system into your own personal experiment in Second System Syndrome.
So basically, if you use this, then there's only one entity to sue to shutdown a project now?
How handy for people who want to use litigation to shut down a competing Open Source project... just think how much more convenient it would have been to have one small group of people with no funding to sue, so they could shut down Linux without having to take on IBM.
(If you can't tall from the above, I'm currently having problems typing at the same time I'm doing the "bad idea dance"...)
``What's wrong with licensing privately-owned, competitive "ID Verification Entities"?''
Use your "Discover" card much?
When I worked for IBM, due to the fallout of cobranding and comarketing agreements that I could not possily map out and decode for you, if you had an employee ID number ending in an odd digit, the corprorate credit card you were issues was American Express, and if you had one ending in an even digit, you got a Discover card.
So if you needed to travel, if your employeed ID was an odd number, you made a phone call. If it was an even number, you had to go get a cash advance on the card, and then use it to pay for the travel in person with a cachiers check.
The moral to this story is that, no matter how you try to level the playing field, you will end up with natural monopolies forming, so long as there's an inequality of trust.
As an experiment to prove this to yourself, I suggest you get copies of the top 6 Internet broswers, get the list of certificate authorities each of them recognize as providing valid root certificates by default, and then throw out all the ones that are not supported by all 6 browsers. When you are done doing this, post:
(1) The list of certificate authorities whose root certificates are by default known by all 6 browsers
(2) The count of root CAs for the browser with the most root CAs
I guarantee you that if you do this, and you are ever needing to start an Internet commerce site that needed secure transactions for payment, you will be picking a particular root CA to buy that certificate from.
Drivers licenses are not for identification, they are used for control and information gathering.
That should be reason enough for you. If you don't believe me, have your driver's license stolen (mine was), and try to get the stolen license invalidated. It's practicaly impossible to do, even if you have a police report in hand.
The problem is that everywhere a driver's license is used does not phone home to verify that the driver's license is in fact not stolen; so anyone who looks roughly like the picture on the front of the license (a biometric identification device intended to prevent fradulent use, BTW) can use the license to identify themselves as you, and there's no cross-check that they are in fact NOT you. This is roughly the same as if you did not do a reverse DNS check followed by a forward lookup on a contacting IP address to verify that the machine contacting you is in the domain which the claim to be from. Your SMTP server might as well be an open relay.
My personal experience ended up with them doing effectively nothing but charging me $25. I suggested that they place a sign on their desk that said "This Side Towards Enemy", since their processes were clearly not directed at the criminal(s) who stole my license.
-
As to privacy, when they swipe your license in a reader to allow you to purchase cold medicine in most large drugstore chains these days, they are in technical violation of the Patriot Act Section 711, 21 USC 830(e)(1)(A)(iv)(I)(bb), in not taking a written signature for the log book, and they are in violation of HIPPA.
The HIPPA violation is collection of more information than they are authorized by law to collect (name, address, and amount purchased); instead, they collect the entire three stripes of the license, which includes all the information in the AMMVA mandated standard ANSI D320-2003, which also includes type of license, whether or not you are a senior citizen, your age, sex, birthdate, identification number, expiration date of the license, endorsements, hair color, eye color, height, weight, restictions (handicaps relevent to driving), and the issue date.
The HIPPA violation, depending on whether the information is controlled according to HIPPA standards, could also constitute a second violation of the Patriot Act, Section 711, 21 USC 830(e)(1)(C)(ii): "prohibit accessing, using, or sharing information in the logbooks for any purpose other than to ensure compliance with this title or to facilitate a product recall to protect public health and safety" -- in other words, they better be damn careful about their information systems attached to their cach registers.
Think about that the next time someone asks you about a national ID card, or you have a cold and consider buying Sudafed.
They will simply stop letting users in those countries purchase from iTMS in order to comply with the legislation.
That would probably do very little to hurt iPod sales anyway.
If they wanted to pick a fight, they'd have iTunes block access to iTMS and put up a "sorry, your government has legislatively disabled your access to iTMS; to email your government representative and complain, click here: [insert appropriate contact information derived from credit card billing information here]".
You say ``Regarding "trademark": "You keep using that word. I do not think it means what you think it means."''
According to the USPTO: "A trademark is a word, phrase, symbol or design, or a combination of words, phrases, symbols or designs, that identifies and distinguishes the source of the goods of one party from those of others."
So there are many aspects of the iPhone that could be Trademarked, as well as the artwork itself being Copyright, vs. Design Patent, and you could make a case for Process Patent on menu/item navigation, if you wanted to get sticky about it.
But this probably needs to be looked at in the right context.
The point of a theory is to allow you to predict future results, based first on the current state of the universe, and future actions.
Falsifiability is what makes a sound theory; what this means is that you can predict something using it, and then measure the results of an experiment based on that prediction, and decide categorically, based on the outcome of the experiment, whether the theory is true or false. If it's false, then it's no longer a theory, and we throw it away -- or, if it still gives useful approximations, like Newtonian mechanisc, then we keep it around, but constrain the circumstances in which it should be used as a tool.
Any theory that's not falsifiable is not a theory - it's a hypothesis at best, and at worst, it's a conjecture.
So, for example, creationism isn't a sound theory, and it's not even a reasonable hypothesis, since it's not falsifiable. To falsify it, you would have to be able to come up with a repeatable laboratory experiment that could prove, one way or the other, whether or not there is a creator. Since the conjecture that there's a creator is a tautology, it's impossible to do this. So the next best thing is Occam's Razor, which, to paraphrase into plain English, states that "All things being equal, the simplest solution tends to be the best one".
-
Global warming, at this point, is a theory (based on observation, without contradiction), but it's not a very good one. It's falsifiable, but not in our lifetimes, and not under laboratory conditions.
Human activity being the root cause of the currently observed global warming is, at this point, a hypothesis.
And the movies idea of what will happen if human activity continues in the current direction is merely conjecture.
-
So to get back to your question: the more ways, and the easier, and the more controlled the conditions under which you can falsify a theory, the higher the quality of the theory.
As to soundness of a particular thory, the more ways that can (and have been attempted to) falsify it, and failed to do so, the more sound the theory.
You are conflating definitions to make your point. Or you are not a native English speaker.
The word "sell" has different meanings, based on its usage and context. In this particular context, it's being used as a transitive verb, and there are eight proper definitions, of which two are applicable from context.
I am using the word "sell" as in "sell reading to children", not in the sense of "sell children to the rendering plant".
You have to cause the employer to develop a belief in the truth, value, or desirability of employing you, as opposed to someone else.
Or, more readably: You have to sell the employer on employing you.
Slashdot Mirror
This article is identical to what we covered... in 300 level Modern Physics in college in 1983.
I don't see how this is new or radical, except for the general population, who seem to think that for every "black hole" there is a corresponding "white hole", or that when you "fall into a black hole", you somehow end up somewhere else.
You should read Feynman's lecture series; he has one from the 50's that debunks the idea of a "graviton" or a particulate carrier for gravity because of the need for it to have mass.
-- Terry
You cite "Patch Deployment Costs" as a reason...
That just begs the question:
Why are patch deployment costs on Windows so high? The only real rationale for this on the Wikipedia page you reference is "a patch issued by Microsoft would break existing functionality", and that's a matter of code, not physics constraints.
-- Terry
This feature is not called "shuffle"... it's called iDrunkDial.
-- Terry
I call "bullshit"... this is the union guys again.
a yoffs_2007/
I'm a former IBM Global Services employee; I left IBM on my own steam to start my own startup when they were closing down our division in 2001 (IBM Global Small Business - say that with a straight face). When that happened, they gave each of us six months to find a job elsewhere within IBM, including paying for flights for interviews, and sending HR people out on site to help write resumes, etc.. IBM bends over backwards to retain employees. IBM is not just a job, it's a job for life for most of its employees.
Here's the original "The Register" article which set off this current firestorm:
http://www.theregister.co.uk/2007/05/02/ibm_may_l
Notice who the press release is from: "Alliance@IBM, a group working to form a union at the company, reports that..."; this is The Communication Workers of America Union; here's their site:
http://www.allianceibm.org/
They claim to be "The official national site for the IBM Employees' Union CWA Local 1701, AFL-CIO" - only there *is no* "IBM Employees' Union" (nowhere on this site will you find membership statistics of any kind).
These are the guys that have been trying to unionize IBM since smelling a payday, in order to get their claws into everyone's paycheck at $10/month and control of the IBM pension plans. When the IBM pension plan converted over to a cash-balance plan, about the only people who didn't have a choice were people who hadn't vested in the plan at all. Yet these same guys tried to stir up a firestorm about it, and attacked IBM over it in the press.
These guys are all about trying to FUD IBM employees into joining their union; while I was with IBM, no one was interested in their party line, even though they practically camped out in all the off-campus restaurants and tried to chat us up every chance they got - they were worse than religious missionaries.
It's too bad Cringely bought into this fake press release, I genuinely enjoyed his "Triumph of the Nerds" television programs.
-- Terry
It goes to two places:
(1) Insurance companies:
- They get paid by you
- They get paid by your employer
- They get paid for malpractice insurance for the doctors
- They get paid for liability insurance by the hospitals
- They get paid for liability insurance by medical equipment manufacturers
Your insurance costs are so high because the insurance companies are charging the people that your insurance pays so much for their own insurance. And the costs for everyone else are so high because malpractice lawsuits pay out so much money; engage in tort reform for malpractice to cap damage claims to reasonable levels.
(2) Billing:
- They don't present you with your actual bill (deductible plus copay) at time of service, when you'd be grateful to pay
- Insurance companies pay "net 90", which means they don't pay until the account has practically gone to collections
- Nobody pays the first bill because of this
- Some people never pay their bill because of this
- "Financial services" companies buy the paper at ~80 cents on the dollar, if the seller of the commercial paper gets a good deal (there's 20% cost right there, and it's a higher percentage than that, if the seller doesn't get a good deal)
This would all go away if there were a single entity that all billing had to go through such that insurance companies and hospitals and doctors were _forced_ to talk electronically, and didn't get paid if they didn't bill at time of service.
-
Want to fix the health care system? Fix the law and fix the money flow itself
-- Terry
Wikipedia reference
;^).
Someone has already updated the Wikipedia entry, citing this Slashdot article as a source.
Maybe it's about time to start trusting it a bit more...
-- Terry
Why more people don't use public domain...
There's no "hold harmless" clause.
The upshot of this is that you are potentially legally liable for consequences arising from someone taking and using the code. This includes, but is not limited to:
- implied or statutory warranties
- legal liability for direct, indirect, incidental, special, consequential, or other damages
Putting something in the public domain unfortunately does not protect you from this by making it truly public property. Instead, you open yourself up to a world of legal hurt - particularly in countries like the U.S., which has more lawyers than could be kept gainfully employed otherwise.
It's unfortunate, but current tort law has made it such that there is (effectively) no such thing as a Commons. It won't be until nearly 2030 until the first piece of software without this liability risk comes off Copyright; I doubt that when it does, you will be able to get useful source code for it anyway.
-- Terry
A patent examiner is GS-5 $38K job to start
http://usptocareers.gov/jobsearch.asp
Key requirements (redux):
- US Citizenship
- Ability to travel
- BA or BS from a community college accredited by ABET _or_ 2nd year coursework in 5 of 7 areas: physics, chemistry, architecture, computer science, mathematics, hydrology, or geology
- Registered as a professional engineer by a state, DC, Guam, or Puerto Rico
- Pass a written test for "Engineer In Training" or professional registration test
- 60 semester hours of courses in basic sciences/physics/math/engineering
So you've got to really want to be a patent examiner, be willing to live with a salary far below what you'd get in private industry with the same paper qualifications.
And then you get to do the scut work for a couple of years.
If you want to have your pay grade go up, you need time in grade and even more qualifications.
So it's pretty much the same deal that entry level teachers get, only you don't get the summer off.
-- Terry
Working brake lights are the law *for a reason*.
It is not always possible, when there is other traffic around, to tell whether the person in front of you is slowing down or not (look up "retrograde motion" on Wikipedia). Brake lights warn the other driver.
If the jackass had obeyed the law and stuck his left arm out the window, I *also* would have known he was stopping, and come to a stop.
So in the case in question, the pickup truck I was following was in violation of two laws in Utah, and would have failed inspection and the state would not have renewed registration over the brake lights (same for Arizona, for that matter).
I'm actually constantly amazed that California has turned registration from an opportunity to improve safety (by conditionalizing registration renewal on safety inspections of all vehicle systems, as other states do), into simple revenue collection.
But to carry your original point to its logical conclusion:
What's a safe following distance for a car which you don't know whether or not they have antilock brakes? Clearly, it's safer to follow closer if they don't have them, since if they have them, then they will have a shorter stopping distance.
Or even better... what's a safe following distance for an enclosed truck with a tare weight of 2000 pounds, a load capacity of 6000 pounds, and therefore brakes capable of stopping 8000 pounds? I guarantee you that if the truck is unloaded, its stopping distance is going to be far shorter than if it were fully loaded, by a significant amount.
A reductio ad absurdum of this argument, you would apparently have all cars at least one block apart so that they weren't "tailgating": there's ALWAYS a reason to stay farther back than you are ("maybe the driver in front of me is drunk, and I just can't tell for sure from a safe distance").
Defensive driving is a good baseline, but at some point, you have to assume that the other drivers on the road will be cooperating with each other to try and drive safely; it would be impossible to drive next to someone on a multilane highway - the distance between cars is *far* shorter than the difference in stopping distance or the time absorbed by reaction time, and after all, they might swerve into you.
-- Terry
Explanation of clutch failure
"I don't understand the problem you had. Your clutch got stuck. Could you pull the transmission out of gear without depressing the clutch?"
The clutch would not disengage (the actual failure was a shear in the clutch plate). As a result, the car could not be forced out of gear, so the engine could not be disengaged from the wheels, without turning off the ignition.
"Did your brakes not work? Most brakes are much stronger than the engine."
That depends on whether you are in a low gear or a high gear (overdrive). If you are in a low gear, the best you can do is burn your breaks and go forward anyway. My brakes worked, but it took time to evaluate why the clutch was not disengaging, try to pop it out of gear anyway (to hell with the gears grinding!), discover that that wasn't going to work, and decide whether or not to ride it out. I had attempted to start decellerating as soon as the light turned yellow, and had about 4 seconds to make all these evaluations and decisions (I did very well in this: I considered 4 options, and average human response time is 2 seconds). If the transmission hadn't been locked into gear, or I had been able to pop it out of gear, I could have stopped before it went red; when the second failed, there was no chance of it.
"Did you sue the mechanic/part manufacturer?"
They redid the repair at no cost, covering towing and rental. The best I could get them on is $390 for the ticket. It would cost far far more than that to take them and the manufacturer of the part to court (both would have to be a party to the case - I can explain the tort liability, if you want me to... though I'm not a member of the bar, I know the law), get expert testimony, etc. etc..
After that, I would (and do) still have the points on my license, and my insurance rate would still go up (as it did).
"Of course there is. Tickets are not issued by a camera, they are issued by the court system. Go to court and try to explain. If the court system in California doesn't accept mechanical failure as an excuse, the fact that you got caught by a camera is irrelevant. A judge would accept a cop's testimony the same way."
I *did* go to court, and argued it to the judge. A cop would have exercised human judgement and not issued the ticket; an officer testified to that (I entered the intersection 1 second after the light turned red). It didn't matter: the camera, lacking human judgement, caused the ticket to be issued, despite what a reasonable person or officer would consider.
So yeah, *if* the officer was an asshole instead of a reasonable person, the court would have accepted his testimony and enforced the rule, but *if* human judgement had been exercised instead of an inhuman machine process, the question of enforcement of the rule would never have arisen.
In California, as in most states, you are not permitted to have a jury trial on traffic violations, or I would have insisted (the only way I could have done this is to follow it through appeal - which as I said before, I simply don't have the time to do right now.
Like I siad: until someone challenges the legality of these things through an appeals process, there's no chance. And I suspect that if it ever came to it, they'd drop the charges on the person challenging, to avoid setting a precedent that would bite into their automated revenue stream (think about how RIAA handles cases, and why the suit for civil damages and legal costs going forward, even though they "decided not to sue", is such a strongly watched case).
-- Terry
"Driver in front never at fault" laws are plan stupid.
Utah has such a law, and I was ticketed there (maybe 16 years ago?) for rear-ending someone; so to answer your question:
"So who the fuck else could it be that was at fault? Santa Claus? The tooth fairy? The devil made me do it? I have a really hard time believing that you seriously mean that it's the driver in the front cars fault that somebody decides to run into him."
It was the fault of the asshole in front of me with the broken brake lights who didn't maintain his vehicle.
I.e.: the guy in front.
I still got the ticket, because that's the law (the officer had no choice), but I was able to fight (and win) in court as a result.
But it cost me the use of the bailment I had to pay until the court heard the case, the use of my vehicle and the cost of a rental car while the case was pending (the insurance would not pay for repairs or a rental if I was at fault), and a day in court -- all over a ticket which should never have been issued to me, but for the utterly stupid law.
-- Terry
I had a ticket for a similar situation:
I had had my vehicle repaired, and the repair was done with a faulty part.
When my clutch stuck the car in gear so that my choices were to run the red light or run off the bridge on the curve immediately on the other side of the intersection because my steering wheel lock engaged from turning off my ignition, I chose running the light and not dying over running the light anyway and dying.
California doesn't allow that a mechanical breakdown could be a mitigating factor in any moving violation, and so it was a $390 fine and points on my license. There is no arguing this.
Even if I agree with this on the theory that vehicle maintenance is the operators responsibility, I _had_ the vehicle maintained by a certified auto mechanic: it was a defective part that caused the malfunction.
I'm convinced that a reasonable police officer in the same situation would have called a tow truck when I finally got the vehicle to a safe point where I could turn off the ignition with the wheels pointing more or less straight, but there is no appeal against an unreasonable camera.
I wish I had had the ability (time) to follow this through an appeals process to case law (just on principle, but I didn't.
I paid the fine: there's pretty much zero appeal for these revenue collection devices, no matter what your story, at least in California.
-- Terry
How about Metalaw instead?
"Do unto others as they would have you do unto them".
Reference:
Andrew G. Haley
"Space Law and Metalaw - A Synoptic View."
Rome: Associazione Italiana Razzi
Proceedings of the Seventh International Astronautical Congress, 1956.
-- Terry
Deficiencies of memtestosx...
It only tests memory under conditions that will only allow it to find real problems with bits in the memory. These include cross-talk related problems from adjacent memory cells, actual stuck bits, paired bits, and so on.
It does NOT test the memory for undervoltaging (which you have to test by actually dropping the voltage to the RAM), nor does it test for delayed DRAM referesh (which you could theoretically do on a PC, if you had a PCI card, and it forced an excessive bus-on time).
In short, software testing is not very good for transient failure testing, and it's not a good thing to include in an OS, rather than as a standalone diagnostic program, due to it being unable to test things that are wired down by the OS (unless your OS can relocate physical pages out of the way to allow you to test them, as well, then running a memory tester under the OS itself means that you don't test any of the memory that's wired down to the OS or to the test tool, if that memory's also wired).
Unfortunately, the most common memory failures in Macs occur during voltage stepping, speed stepping, or when other devices cause the machine to exceed its available power budget.
In other words, problems happen when RAM is run right to the edge of (but not over) its specifications.
You can't do most of that on purpose with normal machines, and not any of it, if it's not standallone.
Software memory checking is OK for gross problems - those problems that usually are not transient - but since most memory you buy has passed that sort of testing before it was sold to you, you are unlikely to ever receive RAM with permanent problems, unless you fail to observe proper precautions when handling or installing it.
So go ahead and use memtestosx; you may in fact find a problem... you're just exceedingly unlikely to find any of the problems that typically differentiate Apple-certified vs. non Apple-certified RAM.
-- Terry
Why you want Apple RAM... or at least certified RAM:
If you insist on third-party RAM, then at least get it in writing that if you have issues that the third party will replace it until you don't have problems, or take it back if they can't get you RAM that performs in spec. (someone else in this article already posted about how they ended up with some nice RAM for one of their PC boxes when it turned out it was out of spec. enough that it wouldn't work in their Mac).
Most third party RAM that fails qualification does so because it's technically out of manufacturer spec.. For many platforms that doesn't tend to matter much, but for Macintosh machines, this typically means problems like drawing current beyond the power budget on wake-from-sleep (and causing the system to be unstable) or being unable to step-up up from a stepped-down clock rate in the necessary time window when the rest of the system is stepping-up, etc. - things that cause the contents of the memory to end up unreliable, or cause the memory bus to be clocked down, and the system run slower.
Unsurprisingly, IBM Power architecture systems tend to have the same issues, given the way multiple simultaneous in-flight transactions are allowed on the memory bus, after the lines and latches have been timed out by the system. When you tune a system that closely, then the electrical characteristics of the components start to matter more and more.
I can't tell you how many customer reported "kernel bugs" I've personally tracked down to actually being faulty third party RAM.
Short of building in a hardware memory tester, which would be prohibitively expensive (including over/under clocking and under/over voltage/current testing, slow refresh cycles, etc.), there's no way to tell if a memory problem exists, short of running into a problem, and throwing the sticks into a real hardware memory tester (it's an NP incomplete problem if you are running test software in the RAM being tested, since it can malfunction and give you a false "no problem" result).
Certified memory is memory that has been thrown into a hardware memory tester, and certified to be in spec. before it's sold to the consumer.
Other peripherals and add-ons are much less sensitive to close-to-spec/out-of-spec problems, compared to RAM. Feel free to take chances with them long before you take chances with third party RAM.
-- Terry
Answer to "What's hard about pointers?"...
There are people who cannot grasp the difference between "a thing" and "a respresentation of a thing". In fact, there was an entire century where almost the entire school of French philosophers were unable to grasp that there was a difference.
For pointers, it comes down to realizing that a pointer to memory is not the memory, it's a representation, and the real memory lives somewhere else: at the dereference of the pointer. And at that location it takes up real space. And so does the representation (but it's comparatively tiny).
The ability to make this disctinction comes naturally when you've written assembly code of any complexity, since you've had to manage the memory yourself. But if you've never done that, or, worse, you've only ever programmed in languages which eschew pointers and try to pretend that their own internal implementations don't use them, even if they won't let you use them, well, then, you've got a conceptual problem. An that conceptual problem is going to be very hard for you to overcome.
It also makes it very, very hard for someone to understand row-major vs. column major languages, or how you would link a program in one against a library in another, and be able to usefully communicate your data between the two. Or it makes it hard to understand the difference between "call by value" and "call by reference.
If I'm interviewing someone, and it becomes clear that they need a "code interview" because some things don't add up between their banter and their resume, the first thing I whip out is a pointer problem.
-- Terry
"If things worked out the way you are supposing, then BSD or equiv. would be the preferred license rather than the GPL."
I'm not the guy you were replying to, but I did get something back from the BSD license... I got an Internet.
The reason TCP/IP is so widespread isn't because AT&T and Novell didn't try to strike a deal to make sure the Internet backbone ran SPX and IPX, or because Microsoft didn't try to get everyone to use NetBEUI, it's because a DARPA grant covered some research, and because the code that resulted from that was under a BSD license, everyone could grab it and use it for free. And when everyone did exactly that, things interoperated.
Just to point out the obvious, but the vast majority of inventors are corporate employees who assign the rights for anything they invent over to the company that's paying their paycheck before they invent anything. If they are lucky, then they work for someone who gives them a plaque or a small bonus when they get a patent issued. So the benefit to the actual inventor is minimal. If they are really lucky, their company collects patents for defensive purposes, rather than beating people over the head with the patent bat to stifle competition. Such defensive portfolios tend to FUD the small guy out of business before they start.
Even with a defensive portdfolio, the companies actually producing product are effectively stripped of their defensive capability by patent holding companies, who don't counter-infringe, and so are not open to negotiation because they produce *absolutely nothing of value*.
So as for reform, you would need to start by disallowing third party assignment (which is how these patent holding companies get the vast majority of their patents).
Then rather than going off half-cocked with a "throw it all away" approach. see what happens after that and maybe one or two other small adjustments, before turning the patent system into your own personal experiment in Second System Syndrome.
Just MHO...
-- Terry
So basically, if you use this, then there's only one entity to sue to shutdown a project now?
How handy for people who want to use litigation to shut down a competing Open Source project... just think how much more convenient it would have been to have one small group of people with no funding to sue, so they could shut down Linux without having to take on IBM.
(If you can't tall from the above, I'm currently having problems typing at the same time I'm doing the "bad idea dance"...)
-- Terry
``What's wrong with licensing privately-owned, competitive "ID Verification Entities"?''
Use your "Discover" card much?
When I worked for IBM, due to the fallout of cobranding and comarketing agreements that I could not possily map out and decode for you, if you had an employee ID number ending in an odd digit, the corprorate credit card you were issues was American Express, and if you had one ending in an even digit, you got a Discover card.
So if you needed to travel, if your employeed ID was an odd number, you made a phone call. If it was an even number, you had to go get a cash advance on the card, and then use it to pay for the travel in person with a cachiers check.
The moral to this story is that, no matter how you try to level the playing field, you will end up with natural monopolies forming, so long as there's an inequality of trust.
As an experiment to prove this to yourself, I suggest you get copies of the top 6 Internet broswers, get the list of certificate authorities each of them recognize as providing valid root certificates by default, and then throw out all the ones that are not supported by all 6 browsers. When you are done doing this, post:
(1) The list of certificate authorities whose root certificates are by default known by all 6 browsers
(2) The count of root CAs for the browser with the most root CAs
I guarantee you that if you do this, and you are ever needing to start an Internet commerce site that needed secure transactions for payment, you will be picking a particular root CA to buy that certificate from.
-- Terry
Drivers licenses are not for identification, they are used for control and information gathering.
That should be reason enough for you. If you don't believe me, have your driver's license stolen (mine was), and try to get the stolen license invalidated. It's practicaly impossible to do, even if you have a police report in hand.
The problem is that everywhere a driver's license is used does not phone home to verify that the driver's license is in fact not stolen; so anyone who looks roughly like the picture on the front of the license (a biometric identification device intended to prevent fradulent use, BTW) can use the license to identify themselves as you, and there's no cross-check that they are in fact NOT you. This is roughly the same as if you did not do a reverse DNS check followed by a forward lookup on a contacting IP address to verify that the machine contacting you is in the domain which the claim to be from. Your SMTP server might as well be an open relay.
My personal experience ended up with them doing effectively nothing but charging me $25. I suggested that they place a sign on their desk that said "This Side Towards Enemy", since their processes were clearly not directed at the criminal(s) who stole my license.
-
As to privacy, when they swipe your license in a reader to allow you to purchase cold medicine in most large drugstore chains these days, they are in technical violation of the Patriot Act Section 711, 21 USC 830(e)(1)(A)(iv)(I)(bb), in not taking a written signature for the log book, and they are in violation of HIPPA.
The HIPPA violation is collection of more information than they are authorized by law to collect (name, address, and amount purchased); instead, they collect the entire three stripes of the license, which includes all the information in the AMMVA mandated standard ANSI D320-2003, which also includes type of license, whether or not you are a senior citizen, your age, sex, birthdate, identification number, expiration date of the license, endorsements, hair color, eye color, height, weight, restictions (handicaps relevent to driving), and the issue date.
The HIPPA violation, depending on whether the information is controlled according to HIPPA standards, could also constitute a second violation of the Patriot Act, Section 711, 21 USC 830(e)(1)(C)(ii): "prohibit accessing, using, or sharing information in the logbooks for any purpose other than to ensure compliance with this title or to facilitate a product recall to protect public health and safety" -- in other words, they better be damn careful about their information systems attached to their cach registers.
Think about that the next time someone asks you about a national ID card, or you have a cold and consider buying Sudafed.
-- Terry
Fined... or...
They will simply stop letting users in those countries purchase from iTMS in order to comply with the legislation.
That would probably do very little to hurt iPod sales anyway.
If they wanted to pick a fight, they'd have iTunes block access to iTMS and put up a "sorry, your government has legislatively disabled your access to iTMS; to email your government representative and complain, click here: [insert appropriate contact information derived from credit card billing information here]".
-- Terry
You say ``Regarding "trademark": "You keep using that word. I do not think it means what you think it means."''
a de_defin.htm
According to the USPTO: "A trademark is a word, phrase, symbol or design, or a combination of words, phrases, symbols or designs, that identifies and distinguishes the source of the goods of one party from those of others."
See also http://www.uspto.gov/web/offices/tac/doc/basic/tr
So there are many aspects of the iPhone that could be Trademarked, as well as the artwork itself being Copyright, vs. Design Patent, and you could make a case for Process Patent on menu/item navigation, if you wanted to get sticky about it.
-- Terry
Falsifiability is the measure of a sound theory.
But this probably needs to be looked at in the right context.
The point of a theory is to allow you to predict future results, based first on the current state of the universe, and future actions.
Falsifiability is what makes a sound theory; what this means is that you can predict something using it, and then measure the results of an experiment based on that prediction, and decide categorically, based on the outcome of the experiment, whether the theory is true or false. If it's false, then it's no longer a theory, and we throw it away -- or, if it still gives useful approximations, like Newtonian mechanisc, then we keep it around, but constrain the circumstances in which it should be used as a tool.
Any theory that's not falsifiable is not a theory - it's a hypothesis at best, and at worst, it's a conjecture.
So, for example, creationism isn't a sound theory, and it's not even a reasonable hypothesis, since it's not falsifiable. To falsify it, you would have to be able to come up with a repeatable laboratory experiment that could prove, one way or the other, whether or not there is a creator. Since the conjecture that there's a creator is a tautology, it's impossible to do this. So the next best thing is Occam's Razor, which, to paraphrase into plain English, states that "All things being equal, the simplest solution tends to be the best one".
-
Global warming, at this point, is a theory (based on observation, without contradiction), but it's not a very good one. It's falsifiable, but not in our lifetimes, and not under laboratory conditions.
Human activity being the root cause of the currently observed global warming is, at this point, a hypothesis.
And the movies idea of what will happen if human activity continues in the current direction is merely conjecture.
-
So to get back to your question: the more ways, and the easier, and the more controlled the conditions under which you can falsify a theory, the higher the quality of the theory.
As to soundness of a particular thory, the more ways that can (and have been attempted to) falsify it, and failed to do so, the more sound the theory.
-- Terry
You are conflating definitions to make your point. Or you are not a native English speaker.
The word "sell" has different meanings, based on its usage and context. In this particular context, it's being used as a transitive verb, and there are eight proper definitions, of which two are applicable from context.
I am using the word "sell" as in "sell reading to children", not in the sense of "sell children to the rendering plant".
You have to cause the employer to develop a belief in the truth, value, or desirability of employing you, as opposed to someone else.
Or, more readably: You have to sell the employer on employing you.
-- Terry
"perhaps the simply righteous will step up"
Yeah, and "the righteous" could code, then there wouldn't be any exploits in the first place. 8-).
-- Terry