"Another kind of lack of silver bullet are declarative languages...it's vastly easier to declare what is needed than it is to specify procedurally how to achieve the goal. However, no one has ever invented a Turing complete declarative language, and there is good reason to think that such a thing is impossible (infinite potential problem domains)."
XSLT and Haskell are declarative, Turing complete languages.
How much media attention did the February 15th protests get? How much did the protests yesterday get? Now, what was the difference between them? Oh yeah, the February 15th ones were BIGGER, but these involved CIVIL DISOBEDIENCE. The lesson is: if you want attention, disrupt business-as-usual.
I assume you aren't opposed to patents on physical objects. Let's take the light bulb. Say I'm Edison. I have a choice between patenting "a device for converting electricity to light" or "a method for producing said device". Which of those is more likely to stop all progress in the lighting industry for 20 years? Which one is a patent on an algorithm? That's right, we'd rather someone patent the algorithm so that someone can design around the patent.
Any method, process, procedure, etc. IS an algorithm, just not specified as well. What makes an algorithm unpatentable just because it's implemented on a computer? In order to fight software patents, we have to find a real answer to that question. Just saying "an algorithm is a theorem" doesn't work.
I don't see any features that would enable the use of functional programming techniques. I don't see any mention of function pointers or overloading (), so how does he expect to write, for example, a general sorting function? (i.e. Sort array a of strings case insensitively)
AFAIK, No. And it probably wouldn't save much time in kernel code. Modern processors juggle instructions around so much anyway, that the overhead of keeping track of the delay slots would cost more than any benefit of having them. Also, how many slots would you need? the original 8086 probably only needed 1 (or none, which would explain not having any). The P4 could use like 15, which is impossible for other reasons.
First, if "a few tens of thousands wouldn't even make Disney blink" then it's trivial.;)
Second, allowing someone to pay to extend their copyright is actually contrary to the justification of copyright. Copyright exists to ensure that stuff is created and eventually released to the public. If a creation is still valuable enough to justify paying for a copyright extension, then the creator has already received compensation for it, and it's valuable enough to require release to the public.
The research project I'm working on uses a Graph library called LEDA which I don't fully understand and which won't compile when our source files have a.cpp extension. *.c works fine (using g++ as the compiler) A couple days ago, I started adding a polymorphic class heirarchy to the project, and the first time I tried to compile it, I got a bunch of strange linker errors telling me "undefined reference to [classname] type_info function". I, of course, blamed LEDA and made another test program to see if I could isolate "LEDA's bug." Of course, I couldn't get the same thing to break in another program. "When in doubt, turn to Google": A search turned up that this error meant I hadn't put the "= 0" in a pure virtual function's declaration. I didn't want any pure virtual functions, but it turned out that I had forgotten the [classname]:: in front of one of my functions. Talk about dumb.
Sure AMD is better than Intel in terms of price, but was it wise of them to stay with old, clunky x86? Pro: Maintain backwards compatibility -- irrelevant with Linux apps as long as the compiler is upgraded. Con: You still have to deal with only 8 general registers and CISC
I believe that RSA relies on the intractability of undoing an exponentiation in the integers mod some prime. In other words:
Pick two probably-prime numbers and call them e and p. Compute (message^e) mod p, and send that along the wire. There's a third number you can pick if you already know e and p, that'll decrypt it. No factoring involved, except for picking your probably-primes in the first place.
Slashdot Mirror
How much media attention did the February 15th protests get? How much did the protests yesterday get? Now, what was the difference between them? Oh yeah, the February 15th ones were BIGGER, but these involved CIVIL DISOBEDIENCE. The lesson is: if you want attention, disrupt business-as-usual.
What a time to be without mod points!
A sci-fi author is the perfect person (maybe other than a materials engineer) to ask about the space elevator and the future of space exploration.
I assume you aren't opposed to patents on physical objects. Let's take the light bulb. Say I'm Edison. I have a choice between patenting "a device for converting electricity to light" or "a method for producing said device". Which of those is more likely to stop all progress in the lighting industry for 20 years? Which one is a patent on an algorithm? That's right, we'd rather someone patent the algorithm so that someone can design around the patent.
Any method, process, procedure, etc. IS an algorithm, just not specified as well. What makes an algorithm unpatentable just because it's implemented on a computer? In order to fight software patents, we have to find a real answer to that question. Just saying "an algorithm is a theorem" doesn't work.
I don't see any features that would enable the use of functional programming techniques. I don't see any mention of function pointers or overloading (), so how does he expect to write, for example, a general sorting function? (i.e. Sort array a of strings case insensitively)
AFAIK, No. And it probably wouldn't save much time in kernel code. Modern processors juggle instructions around so much anyway, that the overhead of keeping track of the delay slots would cost more than any benefit of having them.
Also, how many slots would you need? the original 8086 probably only needed 1 (or none, which would explain not having any). The P4 could use like 15, which is impossible for other reasons.
*sigh* You'd think a petition would be check fur grammar. It looks like these too were run threw a spell-checker set too auto-accept it's guesses.
(btw: 6 intentional mistakes)
An oxymoron doesn't have to be false. It's just something that sounds contradictory at first, which exactly describes "Secure, Efficient and Easy."
First, if "a few tens of thousands wouldn't even make Disney blink" then it's trivial. ;)
Second, allowing someone to pay to extend their copyright is actually contrary to the justification of copyright. Copyright exists to ensure that stuff is created and eventually released to the public. If a creation is still valuable enough to justify paying for a copyright extension, then the creator has already received compensation for it, and it's valuable enough to require release to the public.
The research project I'm working on uses a Graph library called LEDA which I don't fully understand and which won't compile when our source files have a .cpp extension. *.c works fine (using g++ as the compiler)
A couple days ago, I started adding a polymorphic class heirarchy to the project, and the first time I tried to compile it, I got a bunch of strange linker errors telling me "undefined reference to [classname] type_info function". I, of course, blamed LEDA and made another test program to see if I could isolate "LEDA's bug." Of course, I couldn't get the same thing to break in another program.
"When in doubt, turn to Google": A search turned up that this error meant I hadn't put the "= 0" in a pure virtual function's declaration. I didn't want any pure virtual functions, but it turned out that I had forgotten the [classname]:: in front of one of my functions.
Talk about dumb.
So, how long before everything is done on wirelessly connected notepad-shaped flat touch screens?
Upshot doesn't mean what you think it means.
upshot -- 1. The final result; the outcome; 2. The central idea or point; gist.
-- dictionary.com
Sure AMD is better than Intel in terms of price, but was it wise of them to stay with old, clunky x86?
Pro: Maintain backwards compatibility -- irrelevant with Linux apps as long as the compiler is upgraded.
Con: You still have to deal with only 8 general registers and CISC
IA-64 just seems like a better ISA
Hmm, now where's that manual for M$ Office? WinXP? I guess all Microsoft products are pirated.
I believe that RSA relies on the intractability of undoing an exponentiation in the integers mod some prime. In other words:
Pick two probably-prime numbers and call them e and p. Compute (message^e) mod p, and send that along the wire. There's a third number you can pick if you already know e and p, that'll decrypt it. No factoring involved, except for picking your probably-primes in the first place.
Any cryptologists are welcome to correct me.
http://www.acm.org/classics/sep95/