No one else is posting a translation. I'm not Dutch, but I'll take a crack at it (I spent a year in Belgium (Leuven) learning Dutch... and I like to practice it every now and then, so here's my chance). I have to get back to work, but I did do the first paragraph (Dutch speakers please feel free to correct me).
The first photos of AMD's Socket F have emerged on our Gathering of Tweakers forum. In May we wrote that AMD had a new processor socket on its roadmap. The new footprint should have 1207 pins and is intended for multi-Opteron servers. To make possible a processor with support for DDR memory on a DDR2-footprint and vice-versa a new socket was needed. The extra pins that are available are according to reports for an integrated PCI-Express controller on the processors. Noticeable in the photos is the clear separation in the middle of the socket. This seems to indicate that each core of the dual-core Opteron has its own group of pins, and so works as two processors.
Where to begin... data clustering is very much an AI technique. I did my PhD dissertation on it, in an AI laboratory at a major research institution. That some techniques may be simple (such as k-means) does not mean that they fall outside of AI.
Clustering is also known as unsupervised learning, and is an essential tool for many other higher-level AI technologies, including machine vision.
The field of AI is very broad, broader than you probably think. It encompasses clustering. Pick up any good AI book or look at top conferences in the field and find out for yourself. AI has always had the problem that as soon as someone understands part of it, that part is no longer "intelligent" and is now just a "technique". By that logic, AI would not exist.
Come on now, don't be so one-sided and misleading. In the same document you reference, look at page 24 (the page before the graph you point out). The total dollar amount given by the USA has ranked first for most of the last several decades. In 2002, the USA gave more than twice what other nations gave (besides Japan). And this does not count private contributions, which are quite common in the USA. I am an American and have donated to the aid efforts for the Tsunami.
To see how accurate the suggestion is for the number of hits, I typed in "principal component analysis", and the suggestion box said that should yield 3 hits. When you actually do the search, it yields over 3 million hits.
I tried the same thing again, 5 minutes later, and suddenly Google suggest did not even know anything about "principal component analysis". It would not auto-complete the phrase (it did earlier), and would not predict the number of hits.
My parents wouldn't let me buy any games for our computer when I was young. But when I was about 12, my father told me I could play any game -- that I programmed myself, and then bought me a Pascal compiler and taught me how to use it. And I wrote games.
Thus, even if you can't purchase games to play, you can create them (given sufficient interest).
I'm curious where you found such a school. What city do you live in, if you don't mind me asking? I am presuming here that you live in the USA, but I could of course be wrong.
I just finished installing FC3 on a laptop using a firewire cd drive to do the install, and it worked perfectly. I havent tested it further, but if it installs on firewire (which is better than FC1/2 could do without help), it seems to be doing well.
As a side note, I did manage to install FC1 last year on this same box with the same firewire drive, but it was a huge pain -- required patching the ISOs.
According to the nature article, the researchers were from UC San Francisco, not UC San Diego as the post claims. No one else seems to have noticed, so I had to post.
Your post is a bit misleading. It's true that the words are all considered together, but it's not true that they are considered "in context" in the sense that phrases are considered. The thing that makes Naive Bayes classifiers viable for most applications is that they are "naive", and do not consider phrases. Instead, each word is considered conditionally independent of every other word (conditioned on the class label, in this case spam or not spam). The "spamminess" of each word has an additive effect, and the phrase "Joe wants to buy viagra" (in a non-spam) is about equally spammy as "You should buy viagra" (in a spam).
Just wanted to clear that up. It may have been what you meant all along, but that's not what came through.
At the ACM 2000 world finals contest (the one I originally posted about), it was allowed to use an assert() to intentionally crash your program. It's interesting that they may be disabling this in future contests.
Contrary to the parent, you can use assert to help you. You can be *almost* 100% sure that your assert is crashing the program by doing this: submit the program twice, once with the assert, once without. If only the program with the assert (and no other changes) crashes, then your assert has given you some useful information about the guard on the assert.
Hate to reply to my own post... but I need to make a correction to the dates.
I meant the world finals that occurred in spring 2000 and spring 2001. The regional competitions were in 1999 and 2000, respectively, which were the dates I wrote above... blargh.
I was a contestant at the ACM world finals in 1999 and 2000, and have been a coach since.
In 1999, we had a similar problem to what you just described -- the test data did not match the problem description. Many teams that followed the description exactly were not getting their programs accepted, while many other that simply accomodated for the flaw (without knowing that they were doing it) got their programs accepted.
But you have a recourse in this contest, which I remember the Waterloo team (sitting across from us) used: you can intentionally crash your program (using assert()) if an assumption you make about the program fails. If it crashes, you know that your assumptions are not holding. This is what Waterloo did, and they found out that the their program was failing because of a mis-specification.
As others have said, the ACM contest is about time-management, teamwork, good decision making, and clever and fault-tolerant programming. I understand your frustration (I'm right there with you), but I think the lesson to be learned is how to work together solve the problem, not how to hold a grudge.
I use Ampache (a web-based PHP application) to stream my MP3s from the western USA to Europe this year, and it works very well. Sometimes a song stops in the middle, but I have diagnosed that it's a webserver problem, not Ampache, but haven't had time to fix that.
Oh yeah, and once you have all the files on your server and in Ampache, you can keep a local cache of the URLs to all the songs. I do this so I never have to use the web interface unless I want to.
I believe you have boxing & unboxing reversed. I didn't know the definitions before this, but the ones you gave didn't seem intuitive to me, so I searched google for "autoboxing", and found several definitions that are the reverse of yours.
You forgot about the "." (as in ".jp"), which Sun used to claim to have ("we're the dot in..."). Maybe they still do; I don't kept up with who runs the root DNS servers.
If we figure them in, they've been around for about 22 years. So we have (8*6+2*1600+1*22)/13=251 years old. Quite a bit younger than your proposition.
I made a a scope from these plans -- pretty much the same scope that you want to build (8"). I built it in summer 2002.
You can see my scope (including pics) here:
http://www.cs.ucsd.edu/~ghamerly/personal/astro/te lescopes.html
I have some advice on this page, which I will repeat/expand on here:
For me, making the telescope was all about the experience of learning & having a project -- not saving money. I believe I could've bought a better telescope for less money than I spent on making mine (not including the cost of time spent).
The plans are quite good and complete -- though I still had to improvise quite a bit when I didn't have exactly the same materials. However, improvising turned out to be some of the most fun, since it was problem-solving.
On a more technical note, one of my improvisations forced me to place the secondary mirror closer to the main mirror than it really should have been, therefore I lost about 3/4" of the main mirror. This is something to be aware of. This was mostly due to my naivete, and the plans don't really mention this issue. When placing the secondary mirror, take time to really consider your main mirror, focal length, and *the size of your secondary mirror*.
The mirror collimation design (3 bolts on a board) works really well, I think, for how simple the design is.
I bought my mirrors from E-scopes, and was happy with them. I bought teflon from Rob Teeter at Teeter's Telescopes, and I recommend them.
The tube can be had from a local construction surplus store -- it's called Sonotube (brand name) in the industry, and it's used to pour concrete pillars. I was afraid that 1/8"-thick would be too thin, and I considered going with 1/4", but 1/8" turned out to be fine (and much lighter).
I'm extremely happy with my telescope, and while it isn't perfect, I had so much fun making it, and it's so easy to use. I had no sooner finished it than I wanted to make a larger one.:)
Here are 6 pictures taken of the moon with this telescope (just holding my camera up to the eyepiece with my hand):
123456
In an interesting correlation, both of these pictures were taken on trips for the ACM World Programming Contest (different years), which made them even more relevant, since it leads me to think about good problem solving techniques.
Thanks for posting a link to those pics. I'm in Europe and my parents live about 3/4 of a mile from that intersection where the pictures were taken, and I have been searching for pictures close to home. Very scary to see your own neighborhood. Didn't really expect to find these pictures on slashdot.
Back on topic, I have found the US forest service webpages useful (that other posters have given), but I agree that it was generally hard to find information about where the fires actually were. I know, I'm in Europe, so how would I know that there isn't enough information? But I was looking at all the San Diego news websites, and listening to KFMB 760 AM over the internet. It was hard to find specifics about where the fires actually were.
The International Dark Sky Association is an organization that has been looking at these issues for some time, if you want to get involved. They have some interesting information and data about night-time lighting and its relationship to crime, as well as other things.
Your morals didn't change, but you chose to go against them -- to rationalize that it would be ok to cheat.
You said that you still felt it was morally wrong, but you still did it, because otherwise you *may* lose. That's a rationalization. How can you say "Cheating is wrong. If I don't cheat I may get a B in this class. Everyone else is cheating. Therefore, I will cheat, and it is somehow ok."
Or am I missing something?
It's pretty standard practice to use self-administered tests, and just because others may cheat doesn't justify cheating. My fiancee went through a her PhD qualifying exam, where she had to write four papers in about 5 hours, timing herself, and only using certain resources. She didn't cheat, even though it meant she might not pass, and get kicked out of the program she had been working on for two years already. Thankfully, she passed with a clear conscience, knowing that she didn't cheat.
Slashdot Mirror
Or, they could be converting passwords to lowercase before hashing them.
Thanks for the feedback! I appreciate it. Helps me learn.
No one else is posting a translation. I'm not Dutch, but I'll take a crack at it (I spent a year in Belgium (Leuven) learning Dutch... and I like to practice it every now and then, so here's my chance). I have to get back to work, but I did do the first paragraph (Dutch speakers please feel free to correct me).
The first photos of AMD's Socket F have emerged on our Gathering of Tweakers
forum. In May we wrote that AMD had a new processor socket on its roadmap. The
new footprint should have 1207 pins and is intended for multi-Opteron servers.
To make possible a processor with support for DDR memory on a DDR2-footprint
and vice-versa a new socket was needed. The extra pins that are available are
according to reports for an integrated PCI-Express controller on the
processors. Noticeable in the photos is the clear separation in the middle of
the socket. This seems to indicate that each core of the dual-core Opteron has
its own group of pins, and so works as two processors.
Where to begin... data clustering is very much an AI technique. I did my PhD dissertation on it, in an AI laboratory at a major research institution. That some techniques may be simple (such as k-means) does not mean that they fall outside of AI.
Clustering is also known as unsupervised learning, and is an essential tool for many other higher-level AI technologies, including machine vision.
The field of AI is very broad, broader than you probably think. It encompasses clustering. Pick up any good AI book or look at top conferences in the field and find out for yourself. AI has always had the problem that as soon as someone understands part of it, that part is no longer "intelligent" and is now just a "technique". By that logic, AI would not exist.
Come on now, don't be so one-sided and misleading. In the same document you reference, look at page 24 (the page before the graph you point out). The total dollar amount given by the USA has ranked first for most of the last several decades. In 2002, the USA gave more than twice what other nations gave (besides Japan). And this does not count private contributions, which are quite common in the USA. I am an American and have donated to the aid efforts for the Tsunami.
To see how accurate the suggestion is for the number of hits, I typed in "principal component analysis", and the suggestion box said that should yield 3 hits. When you actually do the search, it yields over 3 million hits.
I tried the same thing again, 5 minutes later, and suddenly Google suggest did not even know anything about "principal component analysis". It would not auto-complete the phrase (it did earlier), and would not predict the number of hits.
I guess they are changing things on the fly.
My parents wouldn't let me buy any games for our computer when I was young. But when I was about 12, my father told me I could play any game -- that I programmed myself, and then bought me a Pascal compiler and taught me how to use it. And I wrote games.
Thus, even if you can't purchase games to play, you can create them (given sufficient interest).
I'm curious where you found such a school. What city do you live in, if you don't mind me asking? I am presuming here that you live in the USA, but I could of course be wrong.
I just finished installing FC3 on a laptop using a firewire cd drive to do the install, and it worked perfectly. I havent tested it further, but if it installs on firewire (which is better than FC1/2 could do without help), it seems to be doing well.
As a side note, I did manage to install FC1 last year on this same box with the same firewire drive, but it was a huge pain -- required patching the ISOs.
According to the nature article, the researchers were from UC San Francisco, not UC San Diego as the post claims. No one else seems to have noticed, so I had to post.
- UCSD alum
Easier calculation: 103 square miles is about 10 miles * 10 miles, which concurs with your 5-mile radius.
Your post is a bit misleading. It's true that the words are all considered together, but it's not true that they are considered "in context" in the sense that phrases are considered. The thing that makes Naive Bayes classifiers viable for most applications is that they are "naive", and do not consider phrases. Instead, each word is considered conditionally independent of every other word (conditioned on the class label, in this case spam or not spam). The "spamminess" of each word has an additive effect, and the phrase "Joe wants to buy viagra" (in a non-spam) is about equally spammy as "You should buy viagra" (in a spam).
Just wanted to clear that up. It may have been what you meant all along, but that's not what came through.
At the ACM 2000 world finals contest (the one I originally posted about), it was allowed to use an assert() to intentionally crash your program. It's interesting that they may be disabling this in future contests.
Contrary to the parent, you can use assert to help you. You can be *almost* 100% sure that your assert is crashing the program by doing this: submit the program twice, once with the assert, once without. If only the program with the assert (and no other changes) crashes, then your assert has given you some useful information about the guard on the assert.
Hate to reply to my own post... but I need to make a correction to the dates.
I meant the world finals that occurred in spring 2000 and spring 2001. The regional competitions were in 1999 and 2000, respectively, which were the dates I wrote above... blargh.
I was a contestant at the ACM world finals in 1999 and 2000, and have been a coach since.
In 1999, we had a similar problem to what you just described -- the test data did not match the problem description. Many teams that followed the description exactly were not getting their programs accepted, while many other that simply accomodated for the flaw (without knowing that they were doing it) got their programs accepted.
But you have a recourse in this contest, which I remember the Waterloo team (sitting across from us) used: you can intentionally crash your program (using assert()) if an assumption you make about the program fails. If it crashes, you know that your assumptions are not holding. This is what Waterloo did, and they found out that the their program was failing because of a mis-specification.
As others have said, the ACM contest is about time-management, teamwork, good decision making, and clever and fault-tolerant programming. I understand your frustration (I'm right there with you), but I think the lesson to be learned is how to work together solve the problem, not how to hold a grudge.
I use Ampache (a web-based PHP application) to stream my MP3s from the western USA to Europe this year, and it works very well. Sometimes a song stops in the middle, but I have diagnosed that it's a webserver problem, not Ampache, but haven't had time to fix that.
Anyway, see this: ampache.org
Oh yeah, and once you have all the files on your server and in Ampache, you can keep a local cache of the URLs to all the songs. I do this so I never have to use the web interface unless I want to.
Actually, he did mention that he has sent spam, in the article. But he says he doesn't do it anymore due to increasing penalties for sending spam.
I believe you have boxing & unboxing reversed. I didn't know the definitions before this, but the ones you gave didn't seem intuitive to me, so I searched google for "autoboxing", and found several definitions that are the reverse of yours.
You forgot about the "." (as in ".jp"), which Sun used to claim to have ("we're the dot in..."). Maybe they still do; I don't kept up with who runs the root DNS servers.
If we figure them in, they've been around for about 22 years. So we have (8*6+2*1600+1*22)/13=251 years old. Quite a bit younger than your proposition.
Good luck, and have fun!
We were trying to make a point with the RedHat caps: that they should run linux to avoid the BSOD. Not totally obvious, but that's what we meant.
Why did we have 3 RedHat caps? For just such a situation.
Also, it made a nicer picture because it was a contrast in colors (blue vs. red).
At a "future of technology" display at Epcot Center: the future is blue
In the San Diego airport: your flight is now... cancelled
In an interesting correlation, both of these pictures were taken on trips for the ACM World Programming Contest (different years), which made them even more relevant, since it leads me to think about good problem solving techniques.
Thanks for posting a link to those pics. I'm in Europe and my parents live about 3/4 of a mile from that intersection where the pictures were taken, and I have been searching for pictures close to home. Very scary to see your own neighborhood. Didn't really expect to find these pictures on slashdot.
Back on topic, I have found the US forest service webpages useful (that other posters have given), but I agree that it was generally hard to find information about where the fires actually were. I know, I'm in Europe, so how would I know that there isn't enough information? But I was looking at all the San Diego news websites, and listening to KFMB 760 AM over the internet. It was hard to find specifics about where the fires actually were.
The International Dark Sky Association is an organization that has been looking at these issues for some time, if you want to get involved. They have some interesting information and data about night-time lighting and its relationship to crime, as well as other things.
Your morals didn't change, but you chose to go against them -- to rationalize that it would be ok to cheat.
You said that you still felt it was morally wrong, but you still did it, because otherwise you *may* lose. That's a rationalization. How can you say "Cheating is wrong. If I don't cheat I may get a B in this class. Everyone else is cheating. Therefore, I will cheat, and it is somehow ok."
Or am I missing something?
It's pretty standard practice to use self-administered tests, and just because others may cheat doesn't justify cheating. My fiancee went through a her PhD qualifying exam, where she had to write four papers in about 5 hours, timing herself, and only using certain resources. She didn't cheat, even though it meant she might not pass, and get kicked out of the program she had been working on for two years already. Thankfully, she passed with a clear conscience, knowing that she didn't cheat.