This result required the use of *many* computers to solve *many* positions (approximately four million billion positions), and each found solution was 20 moves or less.
Yes, in some terms it was brute force, but consider how big a number four million billion is, and how long it takes to solve just a single position in 20 moves or less.
You don't understand virtual memory and processes. Go back to Unix 101. It's not tough to use tens of gigabytes of swap on a 32-bit, 4GB RAM system.
Also, linux supports swap partitions larger than 2G for some time now; I almost always run a single 20GB swap partition and it works just fine (and yes, frequently I do use tens of gigabytes of swap). Of course, I do unusual things with my system.
While it's not in-place, it most assuredly does not use n log n memory in the typical case (although for a bad initial input, it *can* use n^2 memory). This implementation, for random input, uses O(n) memory. The n^2 memory problem can be solved by choosing a random pivot element.
18 isn't rare. I've had a Celeron solving random positions for 154 days now (using God's algorithm); here is the solution length distribution (half-turn metric):
15 27
16 400
17 4343
18 11020
19 623
So 18 is by far the most common as well as being the median.
Best flash-based players are the Frontier Labs NexII and the Diva; see http://www.frontierlabs.com/ and http://mydivaplayer.com/. They both use standard CompactFlash (512MB is about $100 these days and the capacity keeps going up), don't have any stupid protection (so you can just copy mp3 files to the flash card using your computer), you can switch the cards, they are small as heck, have no moving parts, and the batteries last a long time.
I've been very happy with these. Much smaller and lighter than an IPod, and with 512MB, I fill them up about once a month for that month's running.
Just as a counterpoint, she's one of the few things I cannot stand about Dr. Dobbs. She thinks she's hilarious but I find it all so predictable, derivative, and self-indulgent. Luckily it's easy for me to turn the page.
I learned all about a live chassis when using an old tube HAM amp in a trailer one day. I was shocked (*shocked*!) to learn that not everyone grounds their chassis.
No; my topology is still better for the snoop traffic because this traffic can be balanced across the links better.
Consider the top two processors in their topology. There is a total of two links connected between these two processors and all other processors. This is a weak bisection and these two links will carry a lot of traffic.
In my topology the worst-case two-processor bisection has three links.
Can't WMP invoke IE5, so IE5 has got its own process etc., and communicate with IE5 over any arbitrary medium (such as raw Windows sockets) instructing IE5 to go out and fetch whatever bits WMP needs? Since ZoneAlarm only ties into the IP stack, and WMP is controlling IE5 through a different medium, all ZoneAlarm sees is IE5 starting up and accessing the net. Essentially, WMP is just using IE5 as a big wget.
Do we know that Microsoft isn't just using IE and HTTP to update your system? Most people tell ZoneAlarm and other firewalls to let port 80 requests from IE5 through . . .
This is yet another way MS can use their monopoly to control your system.
This does seem obvious; it's certainly used in many other electronics. Simply determine the frequency response of the line and load and build a filter (could even be a simple digital filter in this case) to `overdrive' attenuated frequencies. Sure you need a tiny bit of storage [essentially driving about 1M digital filters all in parallel] but it does seem straightforward.
I hope this wasn't patented, or we'll be in a situation like with the Trinitron (where Sony had a monopoly on that technology for a long time and thus could charge a lot . ..)
-tom
Re:What happened to safe.millennium.berkeley.edu?
on
More On Tragedy
·
· Score: 1
Just came up. Maybe safe.millennium.berkeley.edu was a new DNS entry and didn't propogate quickly enough? I've not been able to reach it all day, and neither could the aggregator at http://www.elbnet.com/wtc/.
What happened to safe.millennium.berkeley.edu?
on
More On Tragedy
·
· Score: 1, Offtopic
Anyone know what went wrong here? They've got fat pipes and lots of fast machines, and Berkeley in general is fully reachable.
This seems incredible to me, that anyone playing someone of the caliber of Short can move all his pawns one square forward (giving Short 8 moves to develop his attack, essentially)---and still win!
Any chess experts want to comment on this unconventional play?
Yeah, I cut my teeth on the old TRS-80 Color Computer, which didn't have lower case. Back in 1981-1982 I wrote a program to give it lower case (by using the graphics screen and rendering characters more like what modern machines do today), and published the article in an ancient magazine. Through the magic of E-bay, I was able to find a copy of that magazine last week (my first publication ever). I typed the code in the article back in to both an emulator and an actual machine I found at a Salvation Army store, and sure enough, it still works! What great fun.
The only *real* solution to this, that works across all applications, is to use an operating system that implements a reasonable security model appropriate for a world in which computer code from hostile adversaries is executed on local machines containing sensitive data.
There is no reason a display program should read, much less write, arbitrary files on my hard disk, and this needs to be enforced at the OS level. Fixing it application by application is foolhardy and inappropriate.
The Java security model appears to be a good start, but the solution must live in the OS or else it is too easy to bypass.
So the real question: who's doing work on such a secure OS for the mainstream community? Linux does not appear to be it, nor does Windows 2000 appear to be evolving in that direction. And I see this as *the* problem of the Internet age.
I run a pair of 466MHz Celerons on an Abit BP6 motherboard, totally saturating the CPUs with SETI-like stuff, full-time, over the past four months, running Redhat 6.1, and have never had a single crash, failure, or any other problem. Trust me, I keep these machines at 100% utilization. This doesn't mean there isn't a problem, but I certainly having seen one.
Slashdot Mirror
This result required the use of *many* computers to solve *many* positions (approximately four million billion positions), and each found solution was 20 moves or less.
Yes, in some terms it was brute force, but consider how big a number four million billion is, and how long it takes to solve just a single position in 20 moves or less.
You don't understand virtual memory and processes. Go back to Unix 101. It's not tough to use tens of gigabytes of swap on a 32-bit, 4GB RAM system.
Also, linux supports swap partitions larger than 2G for some time now; I almost always run a single 20GB swap partition and it works just fine (and yes, frequently I do use tens of gigabytes of swap). Of course, I do unusual things with my system.
While it's not in-place, it most assuredly does not use n log n memory in the typical case (although for a bad initial input, it *can* use n^2 memory). This implementation, for random input, uses O(n) memory. The n^2 memory problem can be solved by choosing a random pivot element.
Or play with this *amazing* interactive 3D model of the cube:
http://www.randelshofer.ch/rubik/rubik.html
18 isn't rare. I've had a Celeron solving random positions for 154 days now (using God's algorithm); here is the solution length distribution (half-turn metric):
15 27
16 400
17 4343
18 11020
19 623
So 18 is by far the most common as well as being the median.
There are known cube positions that require 20 moves, so it's at least 20 moves. It could be more; we just don't know.
I've had an optimal cube solver running for a while on a home box, and the distribution of best solution lengths looks like (so far):
5 15
101 16
1265 17
3334 18
194 19
So 20-move cubes are probably pretty rare.
Best flash-based players are the Frontier Labs NexII and the Diva; see http://www.frontierlabs.com/ and http://mydivaplayer.com/. They both use standard CompactFlash (512MB is about $100 these days and the capacity keeps going up), don't have any stupid protection (so you can just copy mp3 files to the flash card using your computer), you can switch the cards, they are small as heck, have no moving parts, and the batteries last a long time.
I've been very happy with these. Much smaller and lighter than an IPod, and with 512MB, I fill them up about once a month for that month's running.
I thought it was "all the power and speed of assembly with the readability and abstraction power of assembly".
Just as a counterpoint, she's one of the few things I cannot stand about Dr. Dobbs. She thinks she's hilarious but I find it all so predictable, derivative, and self-indulgent. Luckily it's easy for me to turn the page.
Same with that Swaine stuff at the end.
I learned all about a live chassis when using an old tube HAM amp in a trailer one day. I was shocked (*shocked*!) to learn that not everyone grounds their chassis.
Renaming a plain text file resume.txt to resume.doc results in a file which can be opened by Word.
Not sure how long this will work, though.
-tom
Consider the top two processors in their topology. There is a total of two links connected between these two processors and all other processors. This is a weak bisection and these two links will carry a lot of traffic.
In my topology the worst-case two-processor bisection has three links.
The middle P should be a + (crossover):
P--P--P
|||
P--+--P
|||
P--P--P
P--P--P
|||
P--P--P
|||
P--P--P
This increases the memory bandwidth available to the four outlying processors, and it also reduces the average path length between processors.
Those inter-processor jumps will accumulate and dramatically increase the memory latency!
Can anyone do better? Assume four of the processors have three links, and four have two links; try to decrease the average path length.
It helps if the topology can be laid out on a board, too!
I can fill my Archos Jukebox with my entire 6GB MP3 collection in about two hours over USB.
-tom
Can't WMP invoke IE5, so IE5 has got its own process etc., and communicate with IE5 over any arbitrary medium (such as raw Windows sockets) instructing IE5 to go out and fetch whatever bits WMP needs? Since ZoneAlarm only ties into the IP stack, and WMP is controlling IE5 through a different medium, all ZoneAlarm sees is IE5 starting up and accessing the net. Essentially, WMP is just using IE5 as a big wget.
-tom
Do we know that Microsoft isn't just using IE and HTTP to update your system? Most people tell ZoneAlarm and other firewalls to let port 80 requests from IE5 through . . .
This is yet another way MS can use their monopoly to control your system.
-tom
This does seem obvious; it's certainly used in many other electronics. Simply determine the frequency response of the line and load and build a filter (could even be a simple digital filter in this case) to `overdrive' attenuated frequencies. Sure you need a tiny bit of storage [essentially driving about 1M digital filters all in parallel] but it does seem straightforward.
.)
I hope this wasn't patented, or we'll be in a situation like with the Trinitron (where Sony had a monopoly on that technology for a long time and thus could charge a lot . .
-tom
Just came up. Maybe safe.millennium.berkeley.edu was a new DNS entry and didn't propogate quickly enough? I've not been able to reach it all day, and neither could the aggregator at http://www.elbnet.com/wtc/.
Anyone know what went wrong here? They've got fat pipes and lots of fast machines, and Berkeley in general is fully reachable.
Any chess experts want to comment on this unconventional play?
A simple script can, given a program which calculates the MD5 checksum for any range of the file, calculate the file itself.
Just calculate the MD5 checksum of the single bytes 0..255, and then ask for the MD5 checksum of the range of each byte, one by one.
Congratulations, you've now `read' aim5.exe from the server that `published' it!
Yeah, I cut my teeth on the old TRS-80 Color Computer, which didn't have lower case. Back in 1981-1982 I wrote a program to give it lower case (by using the graphics screen and rendering characters more like what modern machines do today), and published the article in an ancient magazine. Through the magic of E-bay, I was able to find a copy of that magazine last week (my first publication ever). I typed the code in the article back in to both an emulator and an actual machine I found at a Salvation Army store, and sure enough, it still works! What great fun.
There is no reason a display program should read, much less write, arbitrary files on my hard disk, and this needs to be enforced at the OS level. Fixing it application by application is foolhardy and inappropriate.
The Java security model appears to be a good start, but the solution must live in the OS or else it is too easy to bypass.
So the real question: who's doing work on such a secure OS for the mainstream community? Linux does not appear to be it, nor does Windows 2000 appear to be evolving in that direction. And I see this as *the* problem of the Internet age.
I run a pair of 466MHz Celerons on an Abit BP6 motherboard, totally saturating the CPUs with SETI-like stuff, full-time, over the past four months, running Redhat 6.1, and have never had a single crash, failure, or any other problem. Trust me, I keep these machines at 100% utilization.
This doesn't mean there isn't a problem, but I certainly having seen one.