The parent is incorrect for the reason given by other replies. ie The 'validator' is not the axiom system. Hence, a proof of that the validator does what it's supposed to is not a proof that an axiom system is consistent.
I, on the other hand, have had horrible experiences with them. Consistently low power levels for my modem, frequent disconnections, horrible service. Despite never setting up an appt with me, the serviceman unexpectedly called me at work to let him into my complex. By the time I got home he was gone. Man I was pissed. Then for a billing problem, they said some supervisor would call me back instead of putting me on hold. Never happened. I've never had a complaint in any of my dsl experiences. Comcast has _always_ been a pain in the a@@.
Re:Instant Cryptanalysis
on
Cracking GSM
·
· Score: 1
Does anyone have more information on this? Typically if you decrypt something which has errors, those errors become greatly magnified, and error correcting codes would have a very hard time fixing those errors. I'm wondering if the attack is exploiting something about the equalization training sequence and not so much error correcting codes.
Ah yes, CMU. I went there. I never took a software engineering course though I was given a brief summary of the lessons learned. 1. Things are never finished on time 2. Things never work 3. There are always ppl to get pissed at.
I think that pretty much summarizes building real life software. There's really no reason for a university to offer a "real life" course, and no reason to take one. If you want that then go get a co-op or something. I certainly wouldn't pay tuition so I can sit in a bunch of meetings where I can discuss what error message should pop up on the screen...
Personally I think this teach thinking answer to be somewhat weak and vague. Basic thinking and learning skills are picked up at an early age. One of the things you do get from a "well-rounded" education is many different models of thought, which I is similar to learning to think.. but not quite.
For example, in econ you learn to think in terms of marginal utility and cost, nash equilibriums, and such. In stats, you may learn to think in terms of random variables, nice distributions, or maybe in a Baysian fashion. (Plus, as a bonus, you get to learn how so much of what newscasters and the uneducated is BS.)
These things aren't limited to their respective fields either. You can do stuff like view your successes and failures as random events and try to guess a distribution and parameter instead of getting fixated on some single failure or notice that wasting bandwidth when you pay a fixed monthly charge has zero marginal cost.
I'm (suprise) a CS major, but one of my favorite classes so far has been one on language. Not a foreign language class, but one on semantics, truth values, and cognitive models of language. It provided tools that I could use in certain contexts to cut through BS when ppl play games with semantics to try to prove something. I can't say I always thought this class was great, but that part of what's great about a well-rounded education. You don't really know when or how an appropriate context comes up, but it's cool when it does.
-A not particularly well-rounded person cause he's too lazy when his grades don't depend on it. =P
Are the all 8000 machines web servers or do some also do the number crunching for google's pagerank algo?
If you want to know more about pagerank/web indexing goto Monika Henzinger's (research director for google) webpage. http://www.henzinger.com/~monika
Interesting stuff.
8000 webservers boggles my mind. Also, they can't store an entire index on "just" 160 gigs. So do they do some xtra work if a query spans multiple servers' subset of the index or if it gets sent to the wrong server? It'd also be interesting how much ram they have in these suckers since less ram = more cache misses = more expensive disk accesses.
Google definitely needs tons of computing power to generate the numbers for pagerank (see volsung's post) though I don't know how much a distributed system helps. plus last time I heard, there weren't any algo's that can incrementally update the pagerank info, so it had to be totally recomputed every week or so.
btw does anyone know what altavista uses anymore? It used to be something like 10 alphas with gobs and gobs of ram and disk space.
Slashdot Mirror
The parent is incorrect for the reason given by other replies. ie The 'validator' is not the axiom system. Hence, a proof of that the validator does what it's supposed to is not a proof that an axiom system is consistent.
Mod PDAllen's post up.
Amen brother! Dealing with Comcast customer service has consistently been a pain in the $#@% for me
I, on the other hand, have had horrible experiences with them. Consistently low power levels for my modem, frequent disconnections, horrible service. Despite never setting up an appt with me, the serviceman unexpectedly called me at work to let him into my complex. By the time I got home he was gone. Man I was pissed. Then for a billing problem, they said some supervisor would call me back instead of putting me on hold. Never happened. I've never had a complaint in any of my dsl experiences. Comcast has _always_ been a pain in the a@@.
Does anyone have more information on this? Typically if you decrypt something which has errors, those errors become greatly magnified, and error correcting codes would have a very hard time fixing those errors. I'm wondering if the attack is exploiting something about the equalization training sequence and not so much error correcting codes.
Ah yes, CMU. I went there. I never took a software engineering course though I was given a brief summary of the lessons learned.
1. Things are never finished on time
2. Things never work
3. There are always ppl to get pissed at.
I think that pretty much summarizes building real life software. There's really no reason for a university to offer a "real life" course, and no reason to take one. If you want that then go get a co-op or something. I certainly wouldn't pay tuition so I can sit in a bunch of meetings where I can discuss what error message should pop up on the screen...
Personally I think this teach thinking answer to be somewhat weak and vague. Basic thinking and learning skills are picked up at an early age. One of the things you do get from a "well-rounded" education is many different models of thought, which I is similar to learning to think.. but not quite.
For example, in econ you learn to think in terms of marginal utility and cost, nash equilibriums, and such. In stats, you may learn to think in terms of random variables, nice distributions, or maybe in a Baysian fashion. (Plus, as a bonus, you get to learn how so much of what newscasters and the uneducated is BS.)
These things aren't limited to their respective fields either. You can do stuff like view your successes and failures as random events and try to guess a distribution and parameter instead of getting fixated on some single failure or notice that wasting bandwidth when you pay a fixed monthly charge has zero marginal cost.
I'm (suprise) a CS major, but one of my favorite classes so far has been one on language. Not a foreign language class, but one on semantics, truth values, and cognitive models of language. It provided tools that I could use in certain contexts to cut through BS when ppl play games with semantics to try to prove something. I can't say I always thought this class was great, but that part of what's great about a well-rounded education. You don't really know when or how an appropriate context comes up, but it's cool when it does.
-A not particularly well-rounded person cause he's too lazy when his grades don't depend on it. =P
Are the all 8000 machines web servers or do some also do the number crunching for google's pagerank algo?
If you want to know more about pagerank/web indexing goto Monika Henzinger's (research director for google) webpage. http://www.henzinger.com/~monika
Interesting stuff.
8000 webservers boggles my mind. Also, they can't store an entire index on "just" 160 gigs. So do they do some xtra work if a query spans multiple servers' subset of the index or if it gets sent to the wrong server? It'd also be interesting how much ram they have in these suckers since less ram = more cache misses = more expensive disk accesses.
Google definitely needs tons of computing power to generate the numbers for pagerank (see volsung's post) though I don't know how much a distributed system helps. plus last time I heard, there weren't any algo's that can incrementally update the pagerank info, so it had to be totally recomputed every week or so.
btw does anyone know what altavista uses anymore? It used to be something like 10 alphas with gobs and gobs of ram and disk space.