A computer would *destroy* a human in scrabble, since with a reasonably powerful computer it's possible to try all words, and always make the highest-scoring one. For scrabble, this greedy strategy does really, really well (a bit of a heuristic about blocking TWS and similar might be the only thing worth bothering with as far as lookahead).
For checkers, computers can look ahead so deep that I'd guess (but I don't actually know) they'd also cream the human opponent. (Or, against sufficiently advanced players, tie.)
Go is a good one, though. Word has it that the best go computers can't beat even novice players with a week of training.
Giving the human the source code is a great idea. (In practice, I doubt the human will be able to take much advantage from this, but it will certainly make it more embarassing when he loses!)
In fact, it would be even cooler if the computer players each could read each other's source code* and/or memories. Now *that* would be an interesting program to write...
* It should hopefully be possible to force the programmers to use total functions, or some other interesting but not turing-interesting way of writing their heuristics.
Of course, this is nowhere near the level of play found at Schizophrenic Internet Chess Online, where most people think to a depth of 0. Chess without remorse!
Well actually, more nodes will mean more space for all types of freenet content. That's a Good Thing. I think they are hoping for an MP3 rush to drastically increase the number of nodes, to tell you the truth...
I was able to download MP3s pretty fast (say, as fast as downloading from a moderately loaded cable user on Napster), provided that they were actually still in the network. (Freenet's biggest problem seems to be a lack of permanence). As the number of nodes increases, the speed will increase (as opposed to napster and gnutella), as hopefully will the lifetime of data.
#1: The connections between nodes aren't encrypted (yet)
#2: From what I understand, it's possible to listen to key searches and trial-decrypt the data (which is encrypted) until a match is found. (node-to-node encryption should keep third parties from being able to do this).
Anyway, freenet is starting to actually get good. I think anyone with a permanent net connection should set up a node and help build the network!
> Bounds checking and GC are best done during the
> debug phase...
Bounds checking, probably, but then you don't catch security holes. It's usually difficult or impossible to make a program do all the things it can ever do on any possible input in any order.
Anyway, my point is that if we have spare computing cycles, why not put them to this task? Most people (myself included) would take slower, more stable software over faster, buggier software.
We can use extra speed to bring value to people, rather than make software "bloated"?
One thing we might do is write software which has checking to make sure certain kinds of common errors can't happen, and the programs don't crash. Specifically, we could do bounds checking on arrays (getting rid of some high percentage of security holes, and potentially catching memory corruption heisenbugs earlier) and automatic memory management (getting rid of memory leaks).
I'm not saying that we necessarily need to pay the efficiency price to do this kind of thing, but if the speed is there, why not?
> The United States is a union of states, and it's these states that choose the president, not the people.
Would you then take the position that the current "weighting" of states based on their population is also broken? Should alaska have as big a say as california?
"A couple big cities, and that's it. A majority of the US, from Florida to Alaska and points in between, wanted Bush. Bos-Wash, Chicago, and San-Angeles voted for Gore."
Huh? What do you mean, a majority of the US, based on land area?
If you want to make an argument against letting the popular vote decide using your argument above, you should explain why people who live in cities count less than people who live in rural areas.
BTW, Gore won several New England states which don't have any big cities.
This is a great thing to know, but it doesn't change the fact that Microsoft is deliberately weakening MP3 in order to promote their own proprietary format. I can't expect my younger music-pirate brother to figure this out (or find it on the internet).
WMA may be a better sounding format, but it's dirty ball to do what they are doing to consumers with this move.
MP3 is effectively an open standard. It is not truly free in the sense of Ogg Vorbis, but in practice people make encoders and decoders without paying a cent to Fraunhaufer. It is, or will become, the GIF of music.
If in fact MP3 is under fire from The Man, we don't need more fragmentation in the scene. Concentrate on strengthening the MP3 format. IMO, the risk of (effectively) losing freedom regarding the distribution of music is not worth the small gain in freedom from using a free format.
I think these guys should have waited a bit longer before appearing on slashdot et al... The site looks like ass.
If this proposes to be the hub of stable, polished software, then the site itself should reflect that desired level of quality. Right now it looks like the home page for a shareware app.
Gosh, these licenses sure are hard to keep track of!
Oh I know... what if there was a way for me to pay for and use my software online -- it wouldn't even be stored on my computer! Then I'd have no worries about licensing!
If only Microsoft had some kind of product for me...
Slashdot Mirror
A computer would *destroy* a human in scrabble, since with a reasonably powerful computer it's possible to try all words, and always make the highest-scoring one. For scrabble, this greedy strategy does really, really well (a bit of a heuristic about blocking TWS and similar might be the only thing worth bothering with as far as lookahead).
For checkers, computers can look ahead so deep that I'd guess (but I don't actually know) they'd also cream the human opponent. (Or, against sufficiently advanced players, tie.)
Go is a good one, though. Word has it that the best go computers can't beat even novice players with a week of training.
Giving the human the source code is a great idea. (In practice, I doubt the human will be able to take much advantage from this, but it will certainly make it more embarassing when he loses!)
In fact, it would be even cooler if the computer players each could read each other's source code* and/or memories. Now *that* would be an interesting program to write...
* It should hopefully be possible to force the programmers to use total functions, or some other interesting but not turing-interesting way of writing their heuristics.
Of course, this is nowhere near the level of play found at Schizophrenic Internet Chess Online , where most people think to a depth of 0. Chess without remorse!
Well actually, more nodes will mean more space for all types of freenet content. That's a Good Thing. I think they are hoping for an MP3 rush to drastically increase the number of nodes, to tell you the truth...
They should be, actually...
I was able to download MP3s pretty fast (say, as fast as downloading from a moderately loaded cable user on Napster), provided that they were actually still in the network. (Freenet's biggest problem seems to be a lack of permanence). As the number of nodes increases, the speed will increase (as opposed to napster and gnutella), as hopefully will the lifetime of data.
#1: The connections between nodes aren't encrypted (yet)
#2: From what I understand, it's possible to listen to key searches and trial-decrypt the data (which is encrypted) until a match is found. (node-to-node encryption should keep third parties from being able to do this).
Anyway, freenet is starting to actually get good. I think anyone with a permanent net connection should set up a node and help build the network!
Hmm..
Actually, I just walked in to SEI two days ago for a job interview. I had to sign in and get a visitor's pass; that's it.
Not to defend this strategey, but the post office is federally funded, and they charge for their services too.
I'd rather that corporations pay for it than my tax dollars. =)
That's a pretty cool reference.
Sorry, no time for the full rant.
I just needed to let some steam out, and I can spare the karma.
Object-oriented programming is way, way overrated.
While you're at it, check out the excellent Uddi Uddi font foundry. It's at http://uddiuddi.com/.
> Anybody have any numbers on instruction scheduling efficiency for the P4 on non-optimized code?
3290.21 9283409.44 881.1 29934.1610
I don't know what they mean, though.
AC writes,
...
> Bounds checking and GC are best done during the
> debug phase
Bounds checking, probably, but then you don't catch security holes. It's usually difficult or impossible to make a program do all the things it can ever do on any possible input in any order.
Anyway, my point is that if we have spare computing cycles, why not put them to this task? Most people (myself included) would take slower, more stable software over faster, buggier software.
We can use extra speed to bring value to people, rather than make software "bloated"?
One thing we might do is write software which has checking to make sure certain kinds of common errors can't happen, and the programs don't crash. Specifically, we could do bounds checking on arrays (getting rid of some high percentage of security holes, and potentially catching memory corruption heisenbugs earlier) and automatic memory management (getting rid of memory leaks).
I'm not saying that we necessarily need to pay the efficiency price to do this kind of thing, but if the speed is there, why not?
> The United States is a union of states, and it's these states that choose the president, not the people.
Would you then take the position that the current "weighting" of states based on their population is also broken? Should alaska have as big a say as california?
I agree with most of your post, but...
"A couple big cities, and that's it. A majority of the US, from Florida to Alaska and points in between, wanted Bush. Bos-Wash, Chicago, and San-Angeles voted for Gore."
Huh? What do you mean, a majority of the US, based on land area?
If you want to make an argument against letting the popular vote decide using your argument above, you should explain why people who live in cities count less than people who live in rural areas.
BTW, Gore won several New England states which don't have any big cities.
This is a great thing to know, but it doesn't change the fact that Microsoft is deliberately weakening MP3 in order to promote their own proprietary format. I can't expect my younger music-pirate brother to figure this out (or find it on the internet).
WMA may be a better sounding format, but it's dirty ball to do what they are doing to consumers with this move.
MP3 is effectively an open standard. It is not truly free in the sense of Ogg Vorbis, but in practice people make encoders and decoders without paying a cent to Fraunhaufer. It is, or will become, the GIF of music.
If in fact MP3 is under fire from The Man, we don't need more fragmentation in the scene. Concentrate on strengthening the MP3 format. IMO, the risk of (effectively) losing freedom regarding the distribution of music is not worth the small gain in freedom from using a free format.
Remember, linus thinks mach sux.
Damn dude, where are the chicks?
I think these guys should have waited a bit longer before appearing on slashdot et al... The site looks like ass.
If this proposes to be the hub of stable, polished software, then the site itself should reflect that desired level of quality. Right now it looks like the home page for a shareware app.
That's not associativity, that's symmetry.
So, what do you think?
... what if there was a way for me to pay for and use my software online -- it wouldn't even be stored on my computer! Then I'd have no worries about licensing!
Gosh, these licenses sure are hard to keep track of!
Oh I know
If only Microsoft had some kind of product for me...
Does anybody know what the Automatic Stanford Checker is (mentioned in pre7)? Sounds intriguing..