If you can prove that you are the inventor of an idea, you can sue any corporation for infringement. A great way to prove you're the inventor is by having a patent. I say, the larger the corporation, the deeper the pockets,...
Yep, all true. Unfortunately, it works both ways.
The larger the corporation, the bigger its pool of lawyers, and the bigger its stable of patents. A corporation with sufficiently deep pockets can stop you dead in the water simply by alleging that your patented method infringes on their patent. Their patent might be totally unrelated, or might be invalidated by your earlier patent. It won't matter, though; the simple fact that they are threatening legal action will effectively cut you off from venture capital. If they never sue, you're probably stuck. If they do sue, they can bleed you to death, whether there is any merit to your case or not. You might be right, and you might even win, about the time pigs fly, but it might not do you much good.
Yes, patents offer us little guys some protection against the big bullies. The problem is that they can use the system against us just as easily. Bare knuckles on both sides is fair, unless one of us is a pro boxer and the other is a toddler.
I think that the patent system has become a dance of elephants, and us little guys can only hope to stay out of the way and not get squashed, while everything around us is flattened.
I'm not sure what we can do about it. If we make patents non-transferable, then the inventor would have a job for life (the life of the patent). This might make it harder for companies to justify applying for patents, since they could walk out the door with the inventor if he quit. I suspect that would also make it harder for you to try to sell your patent to a corp. Companies would have a lot of incentive to try to bring back indentured servitude. They probably couldn't make that fly, but they could do some harm trying. Just look at non-compete agreements.
We might try making patents holdable by natural persons only. The consequence of that would be probably be that the management types would be the patent holders for their corporations or divisions, and thus would have an even stronger bargaining position than now. This wouldn't help the inventors much. Again, this might make it harder to sell patents to corporations.
I suppose that if I did get a good idea for dealing with the problem, I could try to patent it?
It seems to me that most folks support the present patent system because they fantasize about ``inventing something'' and getting rich. This fantasy is promoted by the people who are getting richer by exploiting the patent system; they realize that popular support for it would evaporate if the reality became evident. The fantasy is also stoutly defended by folks who are dreaming the dream. The fact that a few people DO manage to go from rags [1] to riches via a patent lends an aura of legitimacy to the system, and a false appearance of reality to the dreams.
I haven't seen any evidence that our present system actually has the potential to help those daydreamers. What it would take is a system which couldn't be exploited by the rich and powerful. Good luck.
[1] I would offer Bill Gates as an example of someone who climbed out of relative poverty by exploiting our intellectual ``property'' system. It's my understanding that his trust fund was providing only a paltry few millions a year, before he developed his monopoly. This proves that any millionaire can become a billionaire under our system.
Yes, I know that Gates' empire depends on copyright, rather than patents.
THis proposal is a kludge on a kludge on a bad idea.
We need to dump the x86, and go to an architecture with many GP registers (say 2K or so) and a flat address space. The Alpha or the PowerPC are closer to the ideal. This proposal is just another ugly hack which tries to get around the fundamentally stupid limitations of the x86 architecture, and makes it still more confusing and harder to use. Just when I thought that segmented memory was the ultimate in futile stupidity, we get the Registermap and Registermapcontrol registers. Just say no to Intel!
The x86 is the ugly mess that it is because it has tried to maintain backwards compatibility with the 8080. Each step of the way, that backwards compatibility could be justified, sort of. But when you must justify, you're wrong, and Intel's CPUs are a prime example of crap outselling technically superior product.
I've programmed 8080's, and I can tell you it left a lot to be desired, in comparison to its contemporaries. Everyone says that assembly is hard, but they're talking about Intel assembly when they say that. Vax assembly was a breeze, in comparison. The 6502 wasn't bad, compared to the Intel and Zilog chips.
I do know something about credit data. I've worked with data from one of the big credit bureaus, developing the sort of models they use for credit scoring. You're right, there aren't a lot of jobs for people who do research work, and that's where the education is invaluable. Because of that, odds are you've never worked with researchers, so the idea of not being a commodity isn't familiar. The managers and the coders can be replaced. The guys who have the education and the experience to understand the problems involved in risk management, or credit scoring, or marketing credit cards, and the ability to solve those problems better than the competition, are probably able to get jobs today. If you're working for a credit card company, your company probably has folks like that working for them. You might never know about it.
I guess that post came across sounding a little meaner than I intended. I do think it needs to be said that working with computers isn't necessarily ``high-tech''. At the beginning of the industrial revolution, mechanics were ``high-tech'', and well paid. By the middle of this century, they were poor, uneducated louts. That same change is already starting to happen in the computer/IT field. It sounds as if you are a manager, and that's certainly not ditch digging, but it really doesn't take much more education. That's what my rich uncle tells me, anyway, and he should know. He did it for years, without education. But he's very bright. It sounds as if you are, too, and you'll be fine, as long as you never have to look for a similar job with no credentials. My uncle got some education after he retired, and he's very happy he did. It made his life better, even though it never made him a dime. I've gotten some education, and I think it's made me poorer but happier.
I aimed that free clue at the folks who think that Java (or tomorrow's trendy language), alone, is a meal ticket. They're the ones who are going to be replaced by a teenager, or an Indian contractor. Easy come, easy go. I'm certainly glad that you have a better job than that. Folks who rely only on coding skills are in trouble, and that will never change, boom or bust.
I slept through school, I did really bad, all because I felt it was worthless. I did feel that my business class, business law, and basic Algebra has been useful. But overall, it wasn't worth my time.
You're doing the modern-day equivalent of ditch digging. Ditch diggers have never needed much education. If you're bright, you'll get by. That hasn't changed in 1000 years.
I'm not saying this to pick on you; I'm trying to make a point that a lot of folks regularly overlook: the nature of unskilled work has changed over the years.
75 years ago writing algorithms to be implemented by machines was a big deal, and it took a great deal of education to be able to deal with the sort of abstraction required. People who did it were able to command quite a bit of money.
Today, children can do many of the basic ``web development'' and ``system administration'' tasks by the time they're old enough to be legally employed. Do you really think that's skilled work? 100 years ago, those same kids would have been bobbin-tenders in a Manchester spinning mill.
I think that getting some more education would probably go a long way to make your life better, and might make you a more interesting person. It probably won't help you in your present job, but it might help you get a better one.
Here's a free hint for code monkeys: If your only value to your employer is your skill at the currently trendy programming language, you can be replaced by a 16 year old, and probably will be when the programming language of the day changes. If you want to have some respect, and some security, become skilled in some field such as physics or engineering (or chemistry, or mathematical genetics, or anything which is HARD). Let the coding be the tool which lets you do what makes you valuable to your employer, rather than the only value you can provide.
... I realized I don't know a damn thing about sine and cosine.
A free trig lesson: Draw a right triangle, with the right angle on the right, and one point up, and the third point to the left. Look at the angle on the left. For that angle, the length of the side opposite to it divided by the length of the side opposite the right angle is a constant, HOWEVER BIG THE TRIANGLE IS. That's right, the ratio is determined only by the angle, not the lengths of the sides. If you know the length of one side and two angles, you know all there is to know about the triangle.
... you need the incompe produced by two to even approach the style of living that took one (income) then.
I hear this a lot, and it's crap. My mother never worked while we kids were still at home. We had one car at most, and never had a tv.
I drove a cab for five years, and managed to support my wife (who didn't work) and kid, and saved enough money to start graduate school. No tv, no stereo, little eating out, no going to movies, et cetera. That was the style of living that took one income in the 50's, and it still takes just one income today. I now have a much better job, and we can afford to eat out and have a computer, and a (single) reliable car. We're still living cheaply, happily doing without things that a lot of people seem to think are essential. My wife still doesn't work, except for taking care of our two kids.
The style of living that takes two incomes involves paying a lot for entertainment, living in a house that costs five or more years income (for two), having two expensive new cars, having lots of electronic toys, flying someplace expensive for a vacation every year, and on and on. That sort of life would have taken a lot more than just two average incomes in the 50's.
Up until the early seventies, Social Security was sufficent for at least a reasonably comfortable retirement.
There's some truth in this. Social Security is a Ponzi scheme, and today it is a drag on the savings of the young. On the other hand, if you live frugally (as previous generations had to), you can save money faster than any previous generation, for any given level of deprivation. That is, if you only do without what your father did without in the 1950's (and that's a lot more than I'm doing without), you will save money far faster than he could have back then.
If I'm worse off than my parents, I want my kids to be worse off than I am, in the same way, and I think that they'd agree.
No generation since the depression? Quit whining
on
Generation Wrecked
·
· Score: 2, Insightful
My parents were born in the early 20's. When the great depression started in '29, my father was 7, my grandfather was 41.
Grandpa had a highschool education (I think... maybe not.) and was in debt, running a small business. When the depression ended, his peak earning years were definitely behind him. I don't think it ever occurred to him that he'd been set up for failure, or that he had failed.
This is pretty silly stuff, I think. Everyone has some rough patches, and for some of us they last pretty long. The big oil price bust in the early 80's was hard on the folks in Alaska and Texas. The generation which fought WWI was called ``the lost generation'' (This was the backgroung for Hemmingway's ``Big, something-hearted River''). If you ever find a generation which was NOT set up for failure in some way, let us know.
First, ``patent it'' and ``give it to the world'' aren't mutually exclusive. You can patent, and then give all users a free, non-revocable license. This is probably a good idea, to avoid being abused by holders of other patents. Or, you could give such a license for use only in software issued under your favorite license(s) (GPL, maybe?).
You say that it is ``... proof (i.e. unbreakable) against brute force attacks and known-plaintext attacks.... Can you prove that? Can you prove it well enough that a mathematician won't laugh at you? If you haven't gotten this reviewed by some competent cryptographers, the whole issue is probably moot anyway.
As for your explicit question: `` Could I sell enough $10 shareware GPG extensions...'' I suspect that the answer is ``probably not''. PGP doesn't seem to have sold very well, and cryptography doesn't seem to be a hot seller right now. Patent or not, this may not be a big money maker. A better way to have phrased your question might have been: ``Is this invention likely to make enough money that I could come out ahead by patenting it?''
A better place to have asked your question might have been a forum where cryptographers hang out. I'm not sure that a lot of them will see this here on slashdot. If you have some sort of credentials as a cryptographer or mathematician, you might try sending emails to some patent-holding cryptographers, and ask about their opinions on your algorithm, and their experiences with patents.
I went and checked what Ballmer is reported to have said; it looks as if you're right. That makes better sense than what I was remembering, anyway. Of course, lower TCO is only one dimension on which we (or they) can add value.
Back to the point I was making: any value which MS adds to Windows, they can extract from the customer, up to the point that its TCO equals that of the substitute good (Linux, in this forum). If the TCO of Windows is less than that of Linux, MS is leaving money on the table.
Thus, in the long term, leaving aside loss-leader deals, Windows TCO must be >= Linux TCO.
Not true, monopoly pricing relates to the monopolist's costs not his customers.
I know what you're saying; I used to teach this to the undergrads. Pricing relates to the producer's cost curves, and to the customers' demand curve. If MS really adds value, that increases quantity demanded (shifts demand curve up or to the right), and thus ups the monopolist's optimal price.
MS's cost curves are probably about flat, meaning that selling a million units more or less won't change average or marginal cost. Furthermore, average cost is above marginal. Given that, MS's problem is to maximize revenue, without regard to the (constant per unit) costs. MS must therefore price where elasticity of demand is one (i.e., a dollar gained by raising the price is offset by a dollar of lost sales). That price will be higher, because the demand curve will be farther to the right, if they add more value. In so far as there is no substitute for their product, the demand curve will be steeper, and again the price will be higher.
It all adds up to what I said in the first place: a monopolist can extract from his customers any economic rents which may accrue to the use of his product.
Support for OSS is minimal at best, and that's to be expected.
If you want support, buy a boxed set. I think RH and Mandrake run $50 - $90 or so, and include some (maybe 90 days?) support in that price. There's your ``... someone is payed to answer phone calls, to write thorough docs.. because it is their JOB.'' I've never done that, because I never felt the need. Perhaps you do have the need, and should pay for support. You will pay for support with Widows, need it or not. That's a (doubtless small) part of the high cost of Windows.
If something isn't done though, OSS software will always take more time to setup than commercial software.
You've never installed Windows, have you? Getting the right drivers, and getting them installed, can be a nightmare. The Windows install process itself seems to be beyond most Windows users, and it can certainly be worse than a Linux install. By the way, when you time the Windows install, you need also to time the installation of all the software you'll be using on the box. After all, that's what you're doing when you time the Linux install.
Most Windows users begin with Linux by comparing the familiar to the unfamiliar. Try to remember how you struggled the first time you had to install Windows, or the first time you ever saw a word processor. It was hard then, just as Linux is now. Windows is easy for you now, just as Linux will be after you've used it for a few months.
So, if you're not excited by the thought of learning something new, why switch? Cost savings, and the fact that what you learn about unix today will still be valid in 20 years. Your knowledge of MS-anything will have a much shorter shelf life. You may well do a LOT more relearning during your life if you stick to Windows.
Does this TCO report factor in the cost of training users? Purchasing support contracts?
Users need training anyway. Most users of MSOffice products will either have enough understanding to generalize their knowlege to Staroffice (no training needed for either path), or they will have to be told what buttons to push for each task (same amount of training required for either path).
Support contracts cost money. Period. If you buy MS, you buy your support there, at their prices, or you do without. If you go with Libre software, you can buy support from several third party suppliers, or from the vendor. You even have the option to develope your own, in-house support. The point here is that these extra choices should enable one to get more cost-effective (and perhaps cheaper) support than is available for MS-anything.
... what costs more: a $10 potted plant or a free kitten?
If you pay [1] someone to set up a debian box, with a cron job which runs apt-get dist-upgrade every night, you get a free kitten which will probably run as intended longer and cheaper than the potted plant. That's no way to run a server, of course, but it makes the point that if you insist on taking the penny-wise-and-pound-foolish approach, you still shouldn't rule out linux.
[1] Think of this as the equivalent to going out and getting the kitten.
The cost of the software license is very small in proportion to the recurring costs of owning the system.
Yep. Since MS has a monopoly on Windows, if you save anything by being able to hire an untrained monkey as sysadmin, they can raise the license fees to capture some or all of that saving. If they fail to keep their prices high enough to capture most of the savings, they have failed to use their monopoly power. The stockholders should probably kick out the management.
The only way that MS can justify its current stock price is to keep on gouging out of their customers any economic rents which may result from using MS products. If Windows' TCO is lower, MS has failed. Even MS is publicly admitting that their TCO is higher than the non-monopoly products, and that they must compete on value. Of course, they will actually compete on lobbyists, but that's another rant.
I've always thought something was wrong with those TCO studies that say MS has a lower TCO than Linux.
Yep, something's wrong.
It just seems blindingly obvious that free software would give you a much lower TCO than something that comes with massive license fees,...
This is where you begin to miss the point, I think.
Redhat doesn't have a monopoly on Redhat, so they cannot extract any of the economic rents you may obtain from using their product. MS, on the other hand, does (famously) have a monopoly. They can and must extract most of the economic rents from their users, in order to justify their stock price.
The upshot of this is that if MS-anything has a lower TCO than any non-monopoly product, MS isn't using their monopoly wisely, and the stockholders should throw out the current managment. If MS products really are cheaper to run, MS should be raising the price to make up the difference (assuming that the products really are comparable, of course). That's where those ``massive license fees'' come in. The point here is that if MS actually gets the long-run TCO for Windows lower than the TCO for Linux, they've failed; they aren't making use of their monopoly power. This is the point I said you were missing.
Even MS is starting to admit this. They are now saying that they compete on value. That would be dangerous ground for them. I'm sure that they will really be competing on lobbyists and lies.
Let's say that Windows really DOES add more value to your hardware than does, say, Redhat Linux (I'll pause for a moment while we all stop giggling. It might be true, after all...). You still wouldn't want to use Windows unless the long-term value added (above that from Linux or BSD) was worth more than the extra cost. To justify a new installation, Windows has to be a lot better than Linux. Enough better to more than make up for the extra cost.
Use Knoppix, or some other live-on-Cd distribution. Customize it to suit your needs (probably need to put the networking information on the CD, and a script to get it to automagically hook up to the network), then take out the HD and put in a second CD drive for the Knoppix CD. The second CD is optional, of course, but it would be nice.
This has the following advantages: the users CAN'T mess it up beyond the power of a reboot to fix. Upgrades are safe and easy: just hand out new CDs. Easy to implement, easy to roll back. Even if the box gets cracked, it only stays cracked until the next boot. Unlike the xterm/compute server solution (which I also like), this will work over a dialup line.
The two disadvantages I see would be that this will have to pause to read from the CD periodically, and it will take more RAM than a HD-based distribution. The first should be acceptable for this sort of use. The second could be ameliorated by leaving in the HD and using it for swap.
What if they scrap the technology and simply post a cop with a stopwatch at every mile marker and overpass?
That would be fine. First, the cops are capable of acting to stop a crime in progress, rather than simply recording it for the evening news. Second, the cops can observe, but they can't RECORD. Tell me, exactly what was the first person you saw on 11 July 1991 wearing? Cops can't record the way cameras can. Third, if you think that you would find that level of constant scrutiny bothersome, rest easy: it would cost too much to keep up for any length of time.
There are both quantitative and qualitative differences between cops and cameras. Cops are OK, cameras aren't.
Remember, the RIAA is well-funded by all the CDs they overprice, and all the artists they rip off.
It's silly nitpicking, I suppose, but you're wrong. The RIAA is well-funded by all the fools who buy the over-priced CDs. The distinction seems important, because it shows where to attack their funding: not by going after the CDs or the ``artists'', but by educating the fools.
Xterms running from a compute server eliminate all those hassles. That cuts direct admin costs, and cuts WAY down on the indirect admin costs which occur when your highly paid workers spend hours each day putzing about trying to get their PC's working.
When everyone is running the same program from the same machine, the admin can make it work right, once, and then never worry about it again. When everyone is running their own copy, on their own machine loaded with other crap, the admin (the many admins and helpdesk people, in this setting) has to troubleshoot each of them, separately, over and over and over...
People can bring in their laptops from home if they want games.
You're telling me that if a high school kid who messes around with with Photoshop occasionally downloads a pirated copy off IRC, that Adobe loses 500 bucks?
So, if this same kid downloads three times, is Adobe out $1500? How about if he keeps all three copies?
A better question is: ``Is that three thefts, or one?'' Surely it's theft, but putting a dollar amount on it doesn't look easy. If I (who haven't a clue how to use Photoshop, nor any desire to learn) downloaded a warez copy of Photoshop, I can assure you that Adobe wouldn't be out any money. I'm not going to buy it, ever, because I'm not going to use it, ever. It would be theft, but a theft that deprives Adobe of nothing.
I think that a better analogy than theft for a case like this is the following: if you disregard the deceased's wishes and cremate instead of bury the remains, you've done no harm, but have shown disrespect. Generally, I don't think that's a crime. Perhaps this is how we should regard the sort of pointless copying that you were talking about.
... what experience I do have is that fortran is excessively difficult to write and lacking in some of the basic features I'd expect a programming language...
You've missed the point of Fortran: Fortran is SIMPLE. FORTRAN is for FORmula TRANslation. If you have an equation, it is easy to translate it into Fortran. You do have functions and subroutines, so you can make your code readable, but you don't have any of the goodies that make C powerful, and nasty. It is difficult to shoot yourself in the foot with Fortran.
If you have something that is best solved by recursion, you probably have a problem that isn't suited for Fortran. Or, more likely, you have a problem which you represent via recursion, but your program (after compilation) solves with a loop.
If you stop pretending that Fortran is C, you'll find that for the set of problems it is meant for (number crunching, ideally with complex numbers), F77 is easier than C, and probably faster, as many other posts have explained already. My F77 program will be written, and run, while you're still debugging your arrays and pointers in C.
Having said that, numerical python and R let you use the good old libraries (Atlas, Blas, Linpac, et al) without giving up the modern conveniences.
Also, last I checked (admitedly a while ago), FORTRAN under Linux gcc was translated to C for compilation.
No. All the frontends, for Aida, C, Fortran, Java, et cetera, compile to a common object code (or assembler? Now my memory is failing.), which is then compiled by a single back end. So, different front ends, same back ends. Nothing gets ``compiled to C''. Unless, of course, you're using gcl, which compiles Lisp to C, which it then passes to gcc. But that's a whole 'nother kettle of eels.
Their gesture-based system is nothing like a keyboard, but I'm still comparing it to my old IBM model M, which has the wonderful, mechanical, click which I can both feel and hear. That feedback works wonders for me; I can type faster on this than on the modern, squishy, low-force keyboards.
The system is intended to replace the keyboard AND the mouse. I like the sound of that part. If you try to use a mouse, you waste a lot of typing time moving back and forth from the keyboard to the mouse. This would really help out there. Of course, keyboard shortcuts accomplish the same thing. They say:
... the communication power of their system is "thousands of times greater" than that of a mouse, which uses just a single moving point as the main input. Using this new technology, two human hands provide 10 points of contact, with a wide range of motion for each, thus providing thousands of different patterns, each of which can mean something different to the computer.
That all sounds a lot like emacs and its key-chords.
They say that it will reduce repetitive stress problems, but I wonder. Is tapping your fingers on a pad, or twisting your wrist, really that different than typing? If you have to do the same operations over and over, aren't you going to eventually get stressed?
... i try to abstain from being a smart-ass but outages are normal?
Verity Slob, PR flack for Worldcom, confirmed that Worldcom customers are indeed offline. Asked why the company's website said ``Outages: Normal '', she replied: ``Well, outages are normal if you are a Worldcom sucker. Oh, did I say `sucker'? I meant victim! I mean, I really meant...'' At this point, Ms. Slob was interrupted by several aides, who sedated and removed her, concluding the press confrence.
Several Worldcom customers were able to confirm that Ms. Slob's statements had been quite accurate.
>>crontab entry to do updates from security.debian.org can do them unassisted
>You've never run an important box then, because no sysadmin worth his salt would ever trust something as critical as security updates to an automated process.
With the stable distribution, using apt-get upgrade, you'll probably never get screwed up doing this. It's STABLE, which means that things don't change. You're still right about the wisdom of testing, but you're very unlikely to get burned on anything with stable.
Slashdot Mirror
Yep, all true. Unfortunately, it works both ways.
The larger the corporation, the bigger its pool of lawyers, and the bigger its stable of patents. A corporation with sufficiently deep pockets can stop you dead in the water simply by alleging that your patented method infringes on their patent. Their patent might be totally unrelated, or might be invalidated by your earlier patent. It won't matter, though; the simple fact that they are threatening legal action will effectively cut you off from venture capital. If they never sue, you're probably stuck. If they do sue, they can bleed you to death, whether there is any merit to your case or not. You might be right, and you might even win, about the time pigs fly, but it might not do you much good.
Yes, patents offer us little guys some protection against the big bullies. The problem is that they can use the system against us just as easily. Bare knuckles on both sides is fair, unless one of us is a pro boxer and the other is a toddler.
I think that the patent system has become a dance of elephants, and us little guys can only hope to stay out of the way and not get squashed, while everything around us is flattened.
I'm not sure what we can do about it. If we make patents non-transferable, then the inventor would have a job for life (the life of the patent). This might make it harder for companies to justify applying for patents, since they could walk out the door with the inventor if he quit. I suspect that would also make it harder for you to try to sell your patent to a corp. Companies would have a lot of incentive to try to bring back indentured servitude. They probably couldn't make that fly, but they could do some harm trying. Just look at non-compete agreements.
We might try making patents holdable by natural persons only. The consequence of that would be probably be that the management types would be the patent holders for their corporations or divisions, and thus would have an even stronger bargaining position than now. This wouldn't help the inventors much. Again, this might make it harder to sell patents to corporations.
I suppose that if I did get a good idea for dealing with the problem, I could try to patent it?
It seems to me that most folks support the present patent system because they fantasize about ``inventing something'' and getting rich. This fantasy is promoted by the people who are getting richer by exploiting the patent system; they realize that popular support for it would evaporate if the reality became evident. The fantasy is also stoutly defended by folks who are dreaming the dream. The fact that a few people DO manage to go from rags [1] to riches via a patent lends an aura of legitimacy to the system, and a false appearance of reality to the dreams.
I haven't seen any evidence that our present system actually has the potential to help those daydreamers. What it would take is a system which couldn't be exploited by the rich and powerful.
Good luck.
[1] I would offer Bill Gates as an example of someone who climbed out of relative poverty by exploiting our intellectual ``property'' system. It's my understanding that his trust fund was providing only a paltry few millions a year, before he developed his monopoly. This proves that any millionaire can become a billionaire under our system.
Yes, I know that Gates' empire depends on copyright, rather than patents.
We need to dump the x86, and go to an architecture with many GP registers (say 2K or so) and a flat address space. The Alpha or the PowerPC are closer to the ideal. This proposal is just another ugly hack which tries to get around the fundamentally stupid limitations of the x86 architecture, and makes it still more confusing and harder to use. Just when I thought that segmented memory was the ultimate in futile stupidity, we get the Registermap and Registermapcontrol registers. Just say no to Intel!
The x86 is the ugly mess that it is because it has tried to maintain backwards compatibility with the 8080. Each step of the way, that backwards compatibility could be justified, sort of. But when you must justify, you're wrong, and Intel's CPUs are a prime example of crap outselling technically superior product.
I've programmed 8080's, and I can tell you it left a lot to be desired, in comparison to its contemporaries. Everyone says that assembly is hard, but they're talking about Intel assembly when they say that. Vax assembly was a breeze, in comparison. The 6502 wasn't bad, compared to the Intel and Zilog chips.
I guess that post came across sounding a little meaner than I intended. I do think it needs to be said that working with computers isn't necessarily ``high-tech''. At the beginning of the industrial revolution, mechanics were ``high-tech'', and well paid. By the middle of this century, they were poor, uneducated louts. That same change is already starting to happen in the computer/IT field. It sounds as if you are a manager, and that's certainly not ditch digging, but it really doesn't take much more education. That's what my rich uncle tells me, anyway, and he should know. He did it for years, without education. But he's very bright. It sounds as if you are, too, and you'll be fine, as long as you never have to look for a similar job with no credentials. My uncle got some education after he retired, and he's very happy he did. It made his life better, even though it never made him a dime. I've gotten some education, and I think it's made me poorer but happier.
I aimed that free clue at the folks who think that Java (or tomorrow's trendy language), alone, is a meal ticket. They're the ones who are going to be replaced by a teenager, or an Indian contractor. Easy come, easy go. I'm certainly glad that you have a better job than that. Folks who rely only on coding skills are in trouble, and that will never change, boom or bust.
You're doing the modern-day equivalent of ditch digging. Ditch diggers have never needed much education. If you're bright, you'll get by. That hasn't changed in 1000 years.
I'm not saying this to pick on you; I'm trying to make a point that a lot of folks regularly overlook: the nature of unskilled work has changed over the years.
75 years ago writing algorithms to be implemented by machines was a big deal, and it took a great deal of education to be able to deal with the sort of abstraction required. People who did it were able to command quite a bit of money.
Today, children can do many of the basic ``web development'' and ``system administration'' tasks by the time they're old enough to be legally employed. Do you really think that's skilled work? 100 years ago, those same kids would have been bobbin-tenders in a Manchester spinning mill.
I think that getting some more education would probably go a long way to make your life better, and might make you a more interesting person. It probably won't help you in your present job, but it might help you get a better one.
Here's a free hint for code monkeys: If your only value to your employer is your skill at the currently trendy programming language, you can be replaced by a 16 year old, and probably will be when the programming language of the day changes. If you want to have some respect, and some security, become skilled in some field such as physics or engineering (or chemistry, or mathematical genetics, or anything which is HARD). Let the coding be the tool which lets you do what makes you valuable to your employer, rather than the only value you can provide.
A free trig lesson:
Draw a right triangle, with the right angle on the right, and one point up, and the third point to the left. Look at the angle on the left. For that angle, the length of the side opposite to it divided by the length of the side opposite the right angle is a constant, HOWEVER BIG THE TRIANGLE IS. That's right, the ratio is determined only by the angle, not the lengths of the sides. If you know the length of one side and two angles, you know all there is to know about the triangle.
I hear this a lot, and it's crap. My mother never worked while we kids were still at home. We had one car at most, and never had a tv.
I drove a cab for five years, and managed to support my wife (who didn't work) and kid, and saved enough money to start graduate school. No tv, no stereo, little eating out, no going to movies, et cetera. That was the style of living that took one income in the 50's, and it still takes just one income today. I now have a much better job, and we can afford to eat out and have a computer, and a (single) reliable car. We're still living cheaply, happily doing without things that a lot of people seem to think are essential. My wife still doesn't work, except for taking care of our two kids.
The style of living that takes two incomes involves paying a lot for entertainment, living in a house that costs five or more years income (for two), having two expensive new cars, having lots of electronic toys, flying someplace expensive for a vacation every year, and on and on. That sort of life would have taken a lot more than just two average incomes in the 50's.
Up until the early seventies, Social Security was sufficent for at least a reasonably comfortable retirement.
There's some truth in this. Social Security is a Ponzi scheme, and today it is a drag on the savings of the young. On the other hand, if you live frugally (as previous generations had to), you can save money faster than any previous generation, for any given level of deprivation. That is, if you only do without what your father did without in the 1950's (and that's a lot more than I'm doing without), you will save money far faster than he could have back then.
If I'm worse off than my parents, I want my kids to be worse off than I am, in the same way, and I think that they'd agree.
Grandpa had a highschool education (I think ... maybe not.) and was in debt, running a small business. When the depression ended, his peak earning years were definitely behind him. I don't think it ever occurred to him that he'd been set up for failure, or that he had failed.
This is pretty silly stuff, I think. Everyone has some rough patches, and for some of us they last pretty long. The big oil price bust in the early 80's was hard on the folks in Alaska and Texas. The generation which fought WWI was called ``the lost generation'' (This was the backgroung for Hemmingway's ``Big, something-hearted River''). If you ever find a generation which was NOT set up for failure in some way, let us know.
You say that it is ``... proof (i.e. unbreakable) against brute force attacks and known-plaintext attacks .... Can you prove that? Can you prove it well enough that a mathematician won't laugh at you? If you haven't gotten this reviewed by some competent cryptographers, the whole issue is probably moot anyway.
As for your explicit question: `` Could I sell enough $10 shareware GPG extensions ...'' I suspect that the answer is ``probably not''. PGP doesn't seem to have sold very well, and cryptography doesn't seem to be a hot seller right now. Patent or not, this may not be a big money maker. A better way to have phrased your question might have been: ``Is this invention likely to make enough money that I could come out ahead by patenting it?''
A better place to have asked your question might have been a forum where cryptographers hang out. I'm not sure that a lot of them will see this here on slashdot. If you have some sort of credentials as a cryptographer or mathematician, you might try sending emails to some patent-holding cryptographers, and ask about their opinions on your algorithm, and their experiences with patents.
I propose a Norwegian Blue Parrot. Dead. Nailed to a perch.
I went and checked what Ballmer is reported to have said; it looks as if you're right. That makes better sense than what I was remembering, anyway. Of course, lower TCO is only one dimension on which we (or they) can add value.
Back to the point I was making: any value which MS adds to Windows, they can extract from the customer, up to the point that its TCO equals that of the substitute good (Linux, in this forum). If the TCO of Windows is less than that of Linux, MS is leaving money on the table.
Thus, in the long term, leaving aside loss-leader deals, Windows TCO must be >= Linux TCO.
I know what you're saying; I used to teach this to the undergrads. Pricing relates to the producer's cost curves, and to the customers' demand curve. If MS really adds value, that increases quantity demanded (shifts demand curve up or to the right), and thus ups the monopolist's optimal price.
MS's cost curves are probably about flat, meaning that selling a million units more or less won't change average or marginal cost. Furthermore, average cost is above marginal. Given that, MS's problem is to maximize revenue, without regard to the (constant per unit) costs. MS must therefore price where elasticity of demand is one (i.e., a dollar gained by raising the price is offset by a dollar of lost sales). That price will be higher, because the demand curve will be farther to the right, if they add more value. In so far as there is no substitute for their product, the demand curve will be steeper, and again the price will be higher.
It all adds up to what I said in the first place: a monopolist can extract from his customers any economic rents which may accrue to the use of his product.
If you want support, buy a boxed set. I think RH and Mandrake run $50 - $90 or so, and include some (maybe 90 days?) support in that price. There's your ``... someone is payed to answer phone calls, to write thorough docs.. because it is their JOB.'' I've never done that, because I never felt the need. Perhaps you do have the need, and should pay for support. You will pay for support with Widows, need it or not. That's a (doubtless small) part of the high cost of Windows.
If something isn't done though, OSS software will always take more time to setup than commercial software.
You've never installed Windows, have you? Getting the right drivers, and getting them installed, can be a nightmare. The Windows install process itself seems to be beyond most Windows users, and it can certainly be worse than a Linux install. By the way, when you time the Windows install, you need also to time the installation of all the software you'll be using on the box. After all, that's what you're doing when you time the Linux install.
Most Windows users begin with Linux by comparing the familiar to the unfamiliar. Try to remember how you struggled the first time you had to install Windows, or the first time you ever saw a word processor. It was hard then, just as Linux is now. Windows is easy for you now, just as Linux will be after you've used it for a few months.
So, if you're not excited by the thought of learning something new, why switch? Cost savings, and the fact that what you learn about unix today will still be valid in 20 years. Your knowledge of MS-anything will have a much shorter shelf life. You may well do a LOT more relearning during your life if you stick to Windows.
Users need training anyway. Most users of MSOffice products will either have enough understanding to generalize their knowlege to Staroffice (no training needed for either path), or they will have to be told what buttons to push for each task (same amount of training required for either path).
Support contracts cost money. Period. If you buy MS, you buy your support there, at their prices, or you do without. If you go with Libre software, you can buy support from several third party suppliers, or from the vendor. You even have the option to develope your own, in-house support. The point here is that these extra choices should enable one to get more cost-effective (and perhaps cheaper) support than is available for MS-anything.
If you pay [1] someone to set up a debian box, with a cron job which runs apt-get dist-upgrade every night, you get a free kitten which will probably run as intended longer and cheaper than the potted plant. That's no way to run a server, of course, but it makes the point that if you insist on taking the penny-wise-and-pound-foolish approach, you still shouldn't rule out linux.
[1] Think of this as the equivalent to going out and getting the kitten.
Yep. Since MS has a monopoly on Windows, if you save anything by being able to hire an untrained monkey as sysadmin, they can raise the license fees to capture some or all of that saving. If they fail to keep their prices high enough to capture most of the savings, they have failed to use their monopoly power. The stockholders should probably kick out the management.
The only way that MS can justify its current stock price is to keep on gouging out of their customers any economic rents which may result from using MS products. If Windows' TCO is lower, MS has failed. Even MS is publicly admitting that their TCO is higher than the non-monopoly products, and that they must compete on value. Of course, they will actually compete on lobbyists, but that's another rant.
Yep, something's wrong.
It just seems blindingly obvious that free software would give you a much lower TCO than something that comes with massive license fees, ...
This is where you begin to miss the point, I think.
Redhat doesn't have a monopoly on Redhat, so they cannot extract any of the economic rents you may obtain from using their product. MS, on the other hand, does (famously) have a monopoly. They can and must extract most of the economic rents from their users, in order to justify their stock price.
The upshot of this is that if MS-anything has a lower TCO than any non-monopoly product, MS isn't using their monopoly wisely, and the stockholders should throw out the current managment. If MS products really are cheaper to run, MS should be raising the price to make up the difference (assuming that the products really are comparable, of course). That's where those ``massive license fees'' come in. The point here is that if MS actually gets the long-run TCO for Windows lower than the TCO for Linux, they've failed; they aren't making use of their monopoly power. This is the point I said you were missing.
Even MS is starting to admit this. They are now saying that they compete on value. That would be dangerous ground for them. I'm sure that they will really be competing on lobbyists and lies.
Let's say that Windows really DOES add more value to your hardware than does, say, Redhat Linux (I'll pause for a moment while we all stop giggling. It might be true, after all...). You still wouldn't want to use Windows unless the long-term value added (above that from Linux or BSD) was worth more than the extra cost. To justify a new installation, Windows has to be a lot better than Linux. Enough better to more than make up for the extra cost.
This has the following advantages:
the users CAN'T mess it up beyond the power of a reboot to fix.
Upgrades are safe and easy: just hand out new CDs. Easy to implement, easy to roll back.
Even if the box gets cracked, it only stays cracked until the next boot.
Unlike the xterm/compute server solution (which I also like), this will work over a dialup line.
The two disadvantages I see would be that this will have to pause to read from the CD periodically, and it will take more RAM than a HD-based distribution. The first should be acceptable for this sort of use. The second could be ameliorated by leaving in the HD and using it for swap.
That would be fine. First, the cops are capable of acting to stop a crime in progress, rather than simply recording it for the evening news. Second, the cops can observe, but they can't RECORD. Tell me, exactly what was the first person you saw on 11 July 1991 wearing? Cops can't record the way cameras can. Third, if you think that you would find that level of constant scrutiny bothersome, rest easy: it would cost too much to keep up for any length of time.
There are both quantitative and qualitative differences between cops and cameras. Cops are OK, cameras aren't.
It's silly nitpicking, I suppose, but you're wrong. The RIAA is well-funded by all the fools who buy the over-priced CDs. The distinction seems important, because it shows where to attack their funding: not by going after the CDs or the ``artists'', but by educating the fools.
HAH! So much for hope on that front ...
When everyone is running the same program from the same machine, the admin can make it work right, once, and then never worry about it again. When everyone is running their own copy, on their own machine loaded with other crap, the admin (the many admins and helpdesk people, in this setting) has to troubleshoot each of them, separately, over and over and over ...
People can bring in their laptops from home if they want games.
Well, rapists get to keep their tools ... it's only fair (in some sense of the word).
So, if this same kid downloads three times, is Adobe out $1500? How about if he keeps all three copies?
A better question is: ``Is that three thefts, or one?'' Surely it's theft, but putting a dollar amount on it doesn't look easy. If I (who haven't a clue how to use Photoshop, nor any desire to learn) downloaded a warez copy of Photoshop, I can assure you that Adobe wouldn't be out any money. I'm not going to buy it, ever, because I'm not going to use it, ever. It would be theft, but a theft that deprives Adobe of nothing.
I think that a better analogy than theft for a case like this is the following: if you disregard the deceased's wishes and cremate instead of bury the remains, you've done no harm, but have shown disrespect. Generally, I don't think that's a crime. Perhaps this is how we should regard the sort of pointless copying that you were talking about.
You've missed the point of Fortran: Fortran is SIMPLE. FORTRAN is for FORmula TRANslation. If you have an equation, it is easy to translate it into Fortran. You do have functions and subroutines, so you can make your code readable, but you don't have any of the goodies that make C powerful, and nasty. It is difficult to shoot yourself in the foot with Fortran.
If you have something that is best solved by recursion, you probably have a problem that isn't suited for Fortran. Or, more likely, you have a problem which you represent via recursion, but your program (after compilation) solves with a loop.
If you stop pretending that Fortran is C, you'll find that for the set of problems it is meant for (number crunching, ideally with complex numbers), F77 is easier than C, and probably faster, as many other posts have explained already. My F77 program will be written, and run, while you're still debugging your arrays and pointers in C.
Having said that, numerical python and R let you use the good old libraries (Atlas, Blas, Linpac, et al) without giving up the modern conveniences.
No. All the frontends, for Aida, C, Fortran, Java, et cetera, compile to a common object code (or assembler? Now my memory is failing.), which is then compiled by a single back end. So, different front ends, same back ends. Nothing gets ``compiled to C''. Unless, of course, you're using gcl, which compiles Lisp to C, which it then passes to gcc. But that's a whole 'nother kettle of eels.
The system is intended to replace the keyboard AND the mouse. I like the sound of that part. If you try to use a mouse, you waste a lot of typing time moving back and forth from the keyboard to the mouse. This would really help out there. Of course, keyboard shortcuts accomplish the same thing. They say:
That all sounds a lot like emacs and its key-chords.They say that it will reduce repetitive stress problems, but I wonder. Is tapping your fingers on a pad, or twisting your wrist, really that different than typing? If you have to do the same operations over and over, aren't you going to eventually get stressed?
Verity Slob, PR flack for Worldcom, confirmed that Worldcom customers are indeed offline. Asked why the company's website said ``Outages: Normal '', she replied: ``Well, outages are normal if you are a Worldcom sucker. Oh, did I say `sucker'? I meant victim! I mean, I really meant ...'' At this point, Ms. Slob was interrupted by several aides, who sedated and removed her, concluding the press confrence.
Several Worldcom customers were able to confirm that Ms. Slob's statements had been quite accurate.
With the stable distribution, using apt-get upgrade, you'll probably never get screwed up doing this. It's STABLE, which means that things don't change. You're still right about the wisdom of testing, but you're very unlikely to get burned on anything with stable.