Back in '95 or so, free SSH clients were more or less non-existant. About all there was for the PC was F-Secure's SSH client, which they allowed people to download 30 day trial copies for. When my copy expired, I poked around in the EULA that came with it. I found that the EULA stated that the copy was valid for exactly 60 days. =)
So I emailed them asking them for a copy with the correct number of days enabled. They wrote back, instead of making the programmers go to the effort of recompiling, how about just a free copy of the client? Which was exactly what I was hoping to get by asking for the extra 30 days.;)
To this day, I still use my free copy of F-Secure SSH.
I had a long conversation about this topic with a friend of mine at Microsoft.
It's great that you could get caught. (And it's debatable in such a case, because how do you track town which of the Starbucks you connected to a "T-Mobile" WAP at was the spoofed one?) But the person's already had access to your bank account, and possibly your computer (if you download any executables), so you've already lost.
Best thing to do is to not sign up for any wireless service in public at all (registering for T-mobile at home is the correct thing to do from a security standpoint), though this process defeats the idea of the oncoming era of wireless being available on every street corner... what's the use in having ubiquitous wireless when anyone with any sense is too paranoid to use it? (Unless it's free, of course.)
I recently spent half my holiday returning screens to Fry's until I got one that was defect free.
I had two always-on green subpixels on the last one, and it became to distracting for me to work with. I'd be writing some code, and then my eye would catch on the green-where-there-should-not-be-green.
Anything that breaks concentration while working is not worth owning. Maybe I could tolerate defects on a second screen on the same system, but never on the primary.
Just out of curiosity, I tried to fill out the claim form (wow, $5 off Excel. The gov't really stuck it to them...:p) and couldn't get it to give me the form. Filled out my name, clicked "Create Form"... clicked "Get Form"... took me back to where I had to fill out my name.
Perhaps the web site is not Mozilla compatible. Which is just ironic on a number of levels.
And because companies use deceptive practices. No small text saying that you can only have one from the same address, but they refuse to grant the rebate if you buy more than one from the same location. Some companies ignore them entirely. Some only send rebates if you pester them about it.
Speaking from experience, I was responsible for rebates for my company (yay, much more important than software engineering:p), I filled out about 50 last year, and learned to keep very detailed records on rebates. Also, I learned Fry's will help you go after the company if they're refusing to honor rebates.
And I'm sure the CCTV that kept rotating to follow me as I ate an ice cream in a public square in Bath was all part of my imagination.
There was a cop, somewhere, that had decided that a tall american eating a scoop of chocolate was a threat of some sort.
A british woman we were talking with said, "Don't look at the cameras, they don't like that."
Fortunately, in America we take liberties a bit more seriously. While I can see us heading in the same direction as the UK (red light cameras, especially, are springing up everywhere) at least for now an attempt to go in the same direction would cause rioting by people from both the left and the right.
I am amazed the British people tolerate it at all.
I was about to post a story on my Cutlass going crazy on me, then I saw this.
83 Cutlasses going crazy, trying to kill people? And there wasn't a recall? WTF??
I was on campus at UCSD, when my car decided to floor itself, running through a 4-way intersection and then aiming itself at a crowd of people crossing the street. I was stunned, holding down the brake as hard as I could (which did nothing! damn the naysayers in this thread which claim that the brake would work -- the brake pedal felt soft, and felt like it was slowing the car only a litte). The car then went tearing down a 25MPH zone at about 50MPH. The whole time, I was desperately trying to steer while reaching down with my hand trying to pull out the pedal (which I had (incorrectly) assumed had gotten stuck). When tugging on the accelerator did nothing I had a few more seconds of panic, then realized I could just cut power to the engine and turned the car off.
I coasted to a stop and just sat there for five minutes...
Unbelieveable this was a common problem and they didn't recall the cars. I mean, they have recalled cars that have had minor defects in seat belts... why the hell wouldn't they recall a car that goes crazy and tries to kill people?
The sad part is, I was fairly poor at the time and kept driving the damn thing. I had other problems, like the pipe that you put the oil dipstick in (whatever it is calld) falling off, but it never went crazy again.
Communism is a system based ultimately on a misunderstanding of human nature.
But I'm sure you know about this problem already, being a member of the Communist party and all. I won't belabor the point. I more want to respond to this: >Yes, he's facing militant opposition from exile >Cubans, many of whom fled for all the wrong >reasons
Well, I'd actually say they were fleeing for all the right reasons. Protecting your family, and moving from an oppressed to a free country seem to be about the best motives you can ask for to risk a dangerous ocean crossing.
Communists love to equate America with Cuba and other despotic regimes, and as such equate Castro imprisoning 75 people for 20 years for independant thought (google for it) with America throwing protestors in jail for a night when they block streets.
But there's an objective difference there, if you care to think about it for a bit. As much as it's not cool to admit it, America is not a police state.
Re:What police/intelligence agencies have learned.
on
Blackhat/Defcon Report
·
· Score: 1
>I'm afraid we don't need Black Hat/Defcon to tell us this. Just yesterday we had major terrorism alerts about specific targets and today we find out the information was all years old.
Years old... and yet the last modified date on the file was in January.
Makes you feel real appreciative to the New York Times for assuring us all we're safe, eh?
>For highly tuned code in numerical libraries people > even worry about L1 and L2 cache utilization.
It's not "even worry" -- optimizing the use of the cache is about the biggest issue in developing high performance codes. =)
I worked in the high performance lab at UC San Diego for years, and we'd typically run hundreds of benchmarks every time we moved to a new machine so that we'd know what sort of parameters would give the best performance based on the real-life (not the published theoretical speeds that are so useless:) performance of cache, network bandwidth, network latency and computation speed.
The trouble with spam is the forged return addresses. If spammers were forced to use real email addresses: 1) It would be much easier to block spam 2) It would be much easier to get their accounts revoked.
A friend of mine runs a script which ensures every email he reads is a real address. Essentially, he's got a cure for spam.
He has a script running on his mail that replies to every email he gets with a confirmation code. When the end user replies with that confirmation code (all it takes is hitting ctrl-r and ctrl-enter) that email address is adding to his "verified email address" list, and the original email goes through.
He doesn't even look at emails that aren't confirmed yet.
If we could get this implemented on a systematic level (such as via confirmaiton reciepts automatically & transparently handed by the Mozilla mail client) it would essentially end free for all spam as we know it. And it doesn't require rewriting the RFCs or adding new headers, or whatever. It would work with any mail reader... though adding in transparency would require updating people's mail clients.
The downsides: -Two extra emails for every one original email are sent... but only the first time. After the email address is verified, it doesn't need another confirmation. If this is implemented system wide, the savings in the reduction of spam messages would greatly outweigh the extra cost on the network. -People who do not confirm don't let their email get through. This happened to me the first time I mailed him after he installed his system. I send him an email, and went home for the day. Didn't see he didn't recieve it until I checked my mail again. Mail clients that handle confirmation transparently would (nearly) solve this problem.
As someone who has experience writing spam filters (I wrote a pretty good neural net spam filter way before that Graham fellow wrote his bayesian filter, that publicity hog!;) (Kidding... his is way eaiser to update than mine) I think that implementing something like this on every ISP in America would immediately kill spam as we know it.
Shame they move so slowly... and never can agree on how to implement anything...
I (a big BG player back in the day) was stoked when I read about NWN. The day it came out, I pressured my roommate to buy the game at the same time, so we could run through the game together.
Then, two days later, we were both repeating: "Never again." "Never again."
It will be a cold day in hell before Bioware gets another hundred dollars out of us for a sequel to a product so inferior in every possible way.
I think the brightest memory I have of playing NWN was running around getting chased by my cohort, who had decided he wanted to kill me. "Run!" "Run, Bill, he's crazy!"... sign, nothing better than being killed by a crazy cohort. And if I'd raise him, he'd immediately start trying to backstab me again. And I couldn't get a new cohort.
So I made a new character, this time the half-orc fighter, which in retrospect, was a bad mistake, since a couple hours later when this cohort decided to kill me as well, he did a lot more damage than the crappy thief.
And NWN had glowing reviews in the press, too. Underdark 4.5 stars my ass.
Microsoft security is better? Hmm, I've been sitting here listening to my roommate curse for two days straight as he's been reformatting his computer, reinstalling windows and installing security patches... only to have the torrents of worms on the net infect his computer before he can download the patches.
It might be amusing if I wasn't trying to sleep right now.
So, ok, the guy in this article says that games journalism needs to be more thoughtful. Ok, whatever. Sure. That's a good idea.
What I think is the most significant problem with games journalism, though, the guy seems to have completely missed -- people who review games almost universally give a game a good rating when they have a preview copy of it.
This makes games reviews completely irrelevant.
It derives from ego, and their ability to say that they have a spiffy new game before anyone else. When they say this in a preview/review, they can't follow it up with "...but the game has pretty major problems with it" because then their coolness for having the game early is negated.
I don't know how I've wasted on games which weren't worth my time & money, simply because all the reviews I've read rave about how great it is, simply because there's a big buzz about the game and they got a copy first.
What I'd like is for game journalists to review their best ratings for games that are real pearls, like Deus Ex, Quake, or Daggerfall, and not waste America's money with their faux critical reviews.
Like a lot of people here, I run two boxes, one Win2k, one Linux. You gotta play to their strengths.
Windows is great for: Return to Castle Wolfenstein Warcraft III A UI that, sadly, is more mature than KDE|Gnome Inertia (My windows box is still using the 2.5GB hard drive I bought in '96, and I don't really feel up to porting all the cruft that has accumulated on it to Linux.)
But on the other hand, I would never consider using my windows box to run: MySQL daemon File Serving Remote interactive prompt (Have you *seen* windows terminal server???) Web Serving Or anything else that requires the least modicum of stability Or anything that would slow down my aforementioned RtCW or Warcraft III if it was run in the background.;)
There's nothing inherently wrong with using Windows over Linux. You just have to play to each of their strengths. Linux has stability, speed and power. Windows has lots of games.
UC San Diego has been a leader in research on hyperthreading. We used to have the Tera MTA, which kinda pioneered the whole field, and we have Dr. Dean Tullsen (and his lab of students), whose hyperthreading architecture was used in the new, now-cancelled, alpha chip.
References: The Tera: http://www.cs.ucsd.edu/users/carter/Tera/tera.html Dean Tullsen: http://charlotte.ucsd.edu/users/tullsen/
I was one of the first five students to use the Tera after it came out of development. I decided to take a different approach in evaluating its performance. I didn't like what the Tera corporate benchmarkers were doing. Which was taking applications with known parallelism, writing a serial version of the code, and then post with glowing reviews the results of the Tera automatically finding parallelism, ignoring that the number of pragmas they had to put into the code to allow the compiler to discover parallelism was more work that just writing a parallel code oneself.
I instead called them on their advertising that their compiler could discover latent parallelism in any computation-heavy code. I noticed John Carmack's.plan file at the time openly questioned the same claim, so I took the single threaded, computation-intensive utility for Quake2 (BSP; LIGHT & VIS are multithreaded) and ran them on the Tera. Nutshell: it couldn't find parallelism. The 300Mhz Tera supercomputer ran at the equivalent speed of a 600Mhz Pentium. Which is crap considering the incredible memory bandwidth and number of computational units it had available.
When I reported the results to Carmack, his response was, "I have never been a big believer in magically parallizing dusty deck codes. I don't mind specifying explicitly parallel activities and threads, especially with the large payoffs involved."
We took the attributes that had been collected from 7,000 pieces of spam from the UCI repository, so that we wouldn't have to collect our own spam and look for common attributes.
Collecting at least 7,000 spam letters off my hotmail account would take at least, oh, say three days, so that saved me a lot of time.;)
There is a trainer included in the package so you can choose whatever attributes you want to look for. The email parser would have to be rewritten a little bit, too, but its definitely viable.
Automatic identification of attributes is outside of its scope.
At UCSD, Bob Boyer and I wrote a neural net spam filter. Neural Nets, as everyone knows, are not really like biological brains, but really just statistical engines similar to the approach the guy above claimed to do.
Our approach worked pretty well (95-97% accuracy), and we had to deal with the same issues that the above "Bayesian" approach did. I.e., weighing the neurons so that false positives occur much less frequently than false negatives, etc. We built it using data on spam collected from the UCI machine learning repository.
It ties in with procmail. I'm not really a windows guy, so if anyone knows how to put a filter between an IMAP server and Microsoft Outlook/Netscape Communicator, I'd be interested in hearing how it's done.
The README for it is at: http://www-cse.ucsd.edu/~wkerney/spamfilter.README And you can download it at: http://www-cse.ucsd.edu/~wkerney/spamfilter.t ar.gz
As a recently graduated grad student at UCSD, I'm fairly familiar with Bill Griswold's project. It's huge -- they have about 10 or 11 students working for them, and more volunteering. I've considered writing source code for them, but other projects, such as work (www-cse.ucsd.edu/~baden) and girlfriend (no innate web address) have taken priority.
Basically, IMO, this article is full of it. These aren't homing devices planted in the shoulders of all the students on campus, used by campus police to track potential hooligans. These are PDAs which can be turned off, whose location tracker service is only enabled by people who *want* to be found. Professors, for example, can turn on the service when they have office hours, but want students to be able to find them while visiting another professor a few doors down. Study partners can use it to find the location of the person they thought they were going to meet at the library. The argument that people can be tracked without their knowledge basically devolves down to an access-control issue... which can be entirely bypassed by turning off the PDA itself.
Anyhow. I think the Salon article misses the point of ActiveCampus entirely. Salon is focusing on privacy, and students' (presumed) inherent desire to stay apart from one another. Active Campus' goals are the opposite: it allows people to get a form communities on a gigantic, alienating campus. You can walk to a coffee shop on campus and invite your friends to meet you there. You can sit in a classroom and discuss the lecture with all the other ActiveCampus participants, allowing you to meet people you'd never talk with before. You can quiz a TA in real-time about the professor's lecture, without becoming that "irritating guy who always disrupts the lectures".:)
Active Campus, by the way, won the top award at the campus Research Review. As well it should... it's one of those rare pieces of technology that is actually good for society.
Cheers, Bill Kerney
Deux Ex "Special Limited Edition"
on
Worst Buy
·
· Score: 1
Heh,
Recently had a bad experience with the local Best Buy in San Diego. The only thing I like about Best Buys are their gigantic bargain bin section, where they put two old (~1999) games in a new box and sell em for 10$. I've bought a couple cheap games that way and have been generally happy with the practice.
Then, over Spring Break, (looking for something to do while all my friends were off doing fun stuff) I buy Deus Ex off the bargain bin section for 10$. It's marked as a "Special Limited Edition", with no explaination of what a special limited edition is, precisely. (I looked.) I figured it meant it didn't come with a manual, or maybe multiplayer.
So I figure 10$ for a game that was supposedly pretty good is a good deal, so I buy it. Take it home. Play it for a while. Get halfway through the game. Splash screen comes up: "Buy Deus Ex, game of the year for 2000! If you buy the real version, you will get to see all these nifty features..."
Bah. Fricken Best Buy sold me a DEMO for 10 dollars!
Maybe I should have guessed that "Special Limited Edition" meant "demo", but all the other games I've bought from their bargain bin have been the real thing. Deceptive as hell.
My mistake, the java source should be in the.jar, but I did a *.class instead of *.[class,java], ah well.
You're right, we use strcat and strcpy, but with a fixed buffer size, fgets() and a terminating condition on reading too many lines, I can't see a buffer exploit in there, but you may disagree.
System() uses nothing read in from the email, it simply chdirs into the directory, and runs./jmail.script,./test.script and./nn.
I've been running it on my main mail account for the last couple days, and with a 95% accuracy rate, 5 of my 100 good messages are misclassified, including one from a friend that I hadn't heard from in a year.
I could definately do better using attributes from my own mailboxes instead of a generic one collected from 4800 sample emails.
A friend and I wrote a neural net spam filter using the UCI Machine Learning Database (on spam), that gets 90-95% accuracy on classifying a message as a spam or not-spam. It's integrated with the mail delivery system via procmail, so you can set it up to deliver all Good mails to one mailbox and Spam mails to another.
It is available (free + open source) at:
http://www-cse.ucsd.edu/~wkerney/spamfilter.tar. gz
The README is at:
http://www-cse.ucsd.edu/~wkerney/spamfilter.READ ME
For Solarius/Linux only, Windows users need not apply. The system works pretty darn well for a simple neural net, and can be greatly improved, so if anyone is interested in modifying our code, drop us a line!:)
There were three arcade games I was in love with in 1992-1994. I need help tracking two of them down:
1) Cdashus -- Wonderful arcade RPG game, four classes to play. I've found it for the TG-16.
2) Time Warriors -- A top view scroller like Ikari Warriors, but with like 40 different characters to play. What platform was this on? I haven't been able to locate it.
3) -- A side scrolling RPG/action game where you had to go through forty levels to get to the top of a mountain. The main thing I remember about it is that you could get an NPC to follow you around, who would attack when you would attack -- there were like 6 different ones, from a ninja that tossed stars to an archer, to a wizard. Anyone know the name of this? About 1990-1991 when it came out.
Slashdot Mirror
Back in '95 or so, free SSH clients were more or less non-existant. About all there was for the PC was F-Secure's SSH client, which they allowed people to download 30 day trial copies for. When my copy expired, I poked around in the EULA that came with it. I found that the EULA stated that the copy was valid for exactly 60 days. =)
;)
So I emailed them asking them for a copy with the correct number of days enabled. They wrote back, instead of making the programmers go to the effort of recompiling, how about just a free copy of the client? Which was exactly what I was hoping to get by asking for the extra 30 days.
To this day, I still use my free copy of F-Secure SSH.
-Bill Kerney
I had a long conversation about this topic with a friend of mine at Microsoft.
It's great that you could get caught. (And it's debatable in such a case, because how do you track town which of the Starbucks you connected to a "T-Mobile" WAP at was the spoofed one?) But the person's already had access to your bank account, and possibly your computer (if you download any executables), so you've already lost.
Best thing to do is to not sign up for any wireless service in public at all (registering for T-mobile at home is the correct thing to do from a security standpoint), though this process defeats the idea of the oncoming era of wireless being available on every street corner... what's the use in having ubiquitous wireless when anyone with any sense is too paranoid to use it? (Unless it's free, of course.)
-Bill
I recently spent half my holiday returning screens to Fry's until I got one that was defect free.
I had two always-on green subpixels on the last one, and it became to distracting for me to work with. I'd be writing some code, and then my eye would catch on the green-where-there-should-not-be-green.
Anything that breaks concentration while working is not worth owning. Maybe I could tolerate defects on a second screen on the same system, but never on the primary.
Just out of curiosity, I tried to fill out the claim form (wow, $5 off Excel. The gov't really stuck it to them... :p) and couldn't get it to give me the form. Filled out my name, clicked "Create Form"... clicked "Get Form"... took me back to where I had to fill out my name.
Perhaps the web site is not Mozilla compatible. Which is just ironic on a number of levels.
And because companies use deceptive practices. No small text saying that you can only have one from the same address, but they refuse to grant the rebate if you buy more than one from the same location. Some companies ignore them entirely. Some only send rebates if you pester them about it.
:p), I filled out about 50 last year, and learned to keep very detailed records on rebates. Also, I learned Fry's will help you go after the company if they're refusing to honor rebates.
Speaking from experience, I was responsible for rebates for my company (yay, much more important than software engineering
And I'm sure the CCTV that kept rotating to follow me as I ate an ice cream in a public square in Bath was all part of my imagination.
There was a cop, somewhere, that had decided that a tall american eating a scoop of chocolate was a threat of some sort.
A british woman we were talking with said, "Don't look at the cameras, they don't like that."
Fortunately, in America we take liberties a bit more seriously. While I can see us heading in the same direction as the UK (red light cameras, especially, are springing up everywhere) at least for now an attempt to go in the same direction would cause rioting by people from both the left and the right.
I am amazed the British people tolerate it at all.
Damn,
I was about to post a story on my Cutlass going crazy on me, then I saw this.
83 Cutlasses going crazy, trying to kill people? And there wasn't a recall? WTF??
I was on campus at UCSD, when my car decided to floor itself, running through a 4-way intersection and then aiming itself at a crowd of people crossing the street. I was stunned, holding down the brake as hard as I could (which did nothing! damn the naysayers in this thread which claim that the brake would work -- the brake pedal felt soft, and felt like it was slowing the car only a litte). The car then went tearing down a 25MPH zone at about 50MPH. The whole time, I was desperately trying to steer while reaching down with my hand trying to pull out the pedal (which I had (incorrectly) assumed had gotten stuck). When tugging on the accelerator did nothing I had a few more seconds of panic, then realized I could just cut power to the engine and turned the car off.
I coasted to a stop and just sat there for five minutes...
Unbelieveable this was a common problem and they didn't recall the cars. I mean, they have recalled cars that have had minor defects in seat belts... why the hell wouldn't they recall a car that goes crazy and tries to kill people?
The sad part is, I was fairly poor at the time and kept driving the damn thing. I had other problems, like the pipe that you put the oil dipstick in (whatever it is calld) falling off, but it never went crazy again.
-Bill
Communism is a system based ultimately on a misunderstanding of human nature.
But I'm sure you know about this problem already, being a member of the Communist party and all. I won't belabor the point. I more want to respond to this:
>Yes, he's facing militant opposition from exile
>Cubans, many of whom fled for all the wrong
>reasons
Well, I'd actually say they were fleeing for all the right reasons. Protecting your family, and moving from an oppressed to a free country seem to be about the best motives you can ask for to risk a dangerous ocean crossing.
Communists love to equate America with Cuba and other despotic regimes, and as such equate Castro imprisoning 75 people for 20 years for independant thought (google for it) with America throwing protestors in jail for a night when they block streets.
But there's an objective difference there, if you care to think about it for a bit. As much as it's not cool to admit it, America is not a police state.
>I'm afraid we don't need Black Hat/Defcon to tell us this. Just yesterday we had major terrorism alerts about specific targets and today we find out the information was all years old.
Years old... and yet the last modified date on the file was in January.
Makes you feel real appreciative to the New York Times for assuring us all we're safe, eh?
>For highly tuned code in numerical libraries people
:) performance of cache, network bandwidth, network latency and computation speed.
> even worry about L1 and L2 cache utilization.
It's not "even worry" -- optimizing the use of the cache is about the biggest issue in developing high performance codes. =)
I worked in the high performance lab at UC San Diego for years, and we'd typically run hundreds of benchmarks every time we moved to a new machine so that we'd know what sort of parameters would give the best performance based on the real-life (not the published theoretical speeds that are so useless
-Bill Kerney
The trouble with spam is the forged return addresses. If spammers were forced to use real email addresses:
;) (Kidding... his is way eaiser to update than mine) I think that implementing something like this on every ISP in America would immediately kill spam as we know it.
1) It would be much easier to block spam
2) It would be much easier to get their accounts revoked.
A friend of mine runs a script which ensures every email he reads is a real address. Essentially, he's got a cure for spam.
He has a script running on his mail that replies to every email he gets with a confirmation code. When the end user replies with that confirmation code (all it takes is hitting ctrl-r and ctrl-enter) that email address is adding to his "verified email address" list, and the original email goes through.
He doesn't even look at emails that aren't confirmed yet.
If we could get this implemented on a systematic level (such as via confirmaiton reciepts automatically & transparently handed by the Mozilla mail client) it would essentially end free for all spam as we know it. And it doesn't require rewriting the RFCs or adding new headers, or whatever. It would work with any mail reader... though adding in transparency would require updating people's mail clients.
The downsides:
-Two extra emails for every one original email are sent... but only the first time. After the email address is verified, it doesn't need another confirmation. If this is implemented system wide, the savings in the reduction of spam messages would greatly outweigh the extra cost on the network.
-People who do not confirm don't let their email get through. This happened to me the first time I mailed him after he installed his system. I send him an email, and went home for the day. Didn't see he didn't recieve it until I checked my mail again. Mail clients that handle confirmation transparently would (nearly) solve this problem.
As someone who has experience writing spam filters (I wrote a pretty good neural net spam filter way before that Graham fellow wrote his bayesian filter, that publicity hog!
Shame they move so slowly... and never can agree on how to implement anything...
-Bill Kerney
I (a big BG player back in the day) was stoked when I read about NWN. The day it came out, I pressured my roommate to buy the game at the same time, so we could run through the game together.
... sign, nothing better than being killed by a crazy cohort. And if I'd raise him, he'd immediately start trying to backstab me again. And I couldn't get a new cohort.
Then, two days later, we were both repeating:
"Never again."
"Never again."
It will be a cold day in hell before Bioware gets another hundred dollars out of us for a sequel to a product so inferior in every possible way.
I think the brightest memory I have of playing NWN was running around getting chased by my cohort, who had decided he wanted to kill me. "Run!" "Run, Bill, he's crazy!"
So I made a new character, this time the half-orc fighter, which in retrospect, was a bad mistake, since a couple hours later when this cohort decided to kill me as well, he did a lot more damage than the crappy thief.
And NWN had glowing reviews in the press, too. Underdark 4.5 stars my ass.
Never again. You hear me Bioware!
Microsoft security is better? Hmm, I've been sitting here listening to my roommate curse for two days straight as he's been reformatting his computer, reinstalling windows and installing security patches... only to have the torrents of worms on the net infect his computer before he can download the patches.
It might be amusing if I wasn't trying to sleep right now.
So, ok, the guy in this article says that games journalism needs to be more thoughtful. Ok, whatever. Sure. That's a good idea.
What I think is the most significant problem with games journalism, though, the guy seems to have completely missed -- people who review games almost universally give a game a good rating when they have a preview copy of it.
This makes games reviews completely irrelevant.
It derives from ego, and their ability to say that they have a spiffy new game before anyone else. When they say this in a preview/review, they can't follow it up with "...but the game has pretty major problems with it" because then their coolness for having the game early is negated.
I don't know how I've wasted on games which weren't worth my time & money, simply because all the reviews I've read rave about how great it is, simply because there's a big buzz about the game and they got a copy first.
What I'd like is for game journalists to review their best ratings for games that are real pearls, like Deus Ex, Quake, or Daggerfall, and not waste America's money with their faux critical reviews.
-Bill
Like a lot of people here, I run two boxes, one Win2k, one Linux. You gotta play to their strengths.
;)
Windows is great for:
Return to Castle Wolfenstein
Warcraft III
A UI that, sadly, is more mature than KDE|Gnome
Inertia (My windows box is still using the 2.5GB hard drive I bought in '96, and I don't really feel up to porting all the cruft that has accumulated on it to Linux.)
But on the other hand, I would never consider using my windows box to run:
MySQL daemon
File Serving
Remote interactive prompt (Have you *seen* windows terminal server???)
Web Serving
Or anything else that requires the least modicum of stability
Or anything that would slow down my aforementioned RtCW or Warcraft III if it was run in the background.
There's nothing inherently wrong with using Windows over Linux. You just have to play to each of their strengths. Linux has stability, speed and power. Windows has lots of games.
Cheers,
Bill Kerney
UC San Diego has been a leader in research on hyperthreading. We used to have the Tera MTA, which kinda pioneered the whole field, and we have Dr. Dean Tullsen (and his lab of students), whose hyperthreading architecture was used in the new, now-cancelled, alpha chip.
l
.plan file at the time openly questioned the same claim, so I took the single threaded, computation-intensive utility for Quake2 (BSP; LIGHT & VIS are multithreaded) and ran them on the Tera. Nutshell: it couldn't find parallelism. The 300Mhz Tera supercomputer ran at the equivalent speed of a 600Mhz Pentium. Which is crap considering the incredible memory bandwidth and number of computational units it had available.
References: The Tera: http://www.cs.ucsd.edu/users/carter/Tera/tera.htm
Dean Tullsen: http://charlotte.ucsd.edu/users/tullsen/
I was one of the first five students to use the Tera after it came out of development. I decided to take a different approach in evaluating its performance. I didn't like what the Tera corporate benchmarkers were doing. Which was taking applications with known parallelism, writing a serial version of the code, and then post with glowing reviews the results of the Tera automatically finding parallelism, ignoring that the number of pragmas they had to put into the code to allow the compiler to discover parallelism was more work that just writing a parallel code oneself.
I instead called them on their advertising that their compiler could discover latent parallelism in any computation-heavy code. I noticed John Carmack's
When I reported the results to Carmack, his response was, "I have never been a big believer in magically parallizing dusty deck codes. I don't mind specifying explicitly parallel activities and threads, especially with the large payoffs involved."
Cheers,
Bill Kerney
We took the attributes that had been collected from 7,000 pieces of spam from the UCI repository, so that we wouldn't have to collect our own spam and look for common attributes.
;)
Collecting at least 7,000 spam letters off my hotmail account would take at least, oh, say three days, so that saved me a lot of time.
There is a trainer included in the package so you can choose whatever attributes you want to look for. The email parser would have to be rewritten a little bit, too, but its definitely viable.
Automatic identification of attributes is outside of its scope.
-Bill
At UCSD, Bob Boyer and I wrote a neural net spam filter. Neural Nets, as everyone knows, are not really like biological brains, but really just statistical engines similar to the approach the guy above claimed to do.
E t ar.gz
Our approach worked pretty well (95-97% accuracy), and we had to deal with the same issues that the above "Bayesian" approach did. I.e., weighing the neurons so that false positives occur much less frequently than false negatives, etc. We built it using data on spam collected from the UCI machine learning repository.
It ties in with procmail. I'm not really a windows guy, so if anyone knows how to put a filter between an IMAP server and Microsoft Outlook/Netscape Communicator, I'd be interested in hearing how it's done.
The README for it is at: http://www-cse.ucsd.edu/~wkerney/spamfilter.READM
And you can download it at:
http://www-cse.ucsd.edu/~wkerney/spamfilter.
-Bill Kerney
wkerney at ucsd.edu
As a recently graduated grad student at UCSD, I'm fairly familiar with Bill Griswold's project. It's huge -- they have about 10 or 11 students working for them, and more volunteering. I've considered writing source code for them, but other projects, such as work (www-cse.ucsd.edu/~baden) and girlfriend (no innate web address) have taken priority.
:)
Basically, IMO, this article is full of it. These aren't homing devices planted in the shoulders of all the students on campus, used by campus police to track potential hooligans. These are PDAs which can be turned off, whose location tracker service is only enabled by people who *want* to be found. Professors, for example, can turn on the service when they have office hours, but want students to be able to find them while visiting another professor a few doors down. Study partners can use it to find the location of the person they thought they were going to meet at the library. The argument that people can be tracked without their knowledge basically devolves down to an access-control issue... which can be entirely bypassed by turning off the PDA itself.
Anyhow. I think the Salon article misses the point of ActiveCampus entirely. Salon is focusing on privacy, and students' (presumed) inherent desire to stay apart from one another. Active Campus' goals are the opposite: it allows people to get a form communities on a gigantic, alienating campus. You can walk to a coffee shop on campus and invite your friends to meet you there. You can sit in a classroom and discuss the lecture with all the other ActiveCampus participants, allowing you to meet people you'd never talk with before. You can quiz a TA in real-time about the professor's lecture, without becoming that "irritating guy who always disrupts the lectures".
Active Campus, by the way, won the top award at the campus Research Review. As well it should... it's one of those rare pieces of technology that is actually good for society.
Cheers,
Bill Kerney
Heh,
Recently had a bad experience with the local Best Buy in San Diego. The only thing I like about Best Buys are their gigantic bargain bin section, where they put two old (~1999) games in a new box and sell em for 10$. I've bought a couple cheap games that way and have been generally happy with the practice.
Then, over Spring Break, (looking for something to do while all my friends were off doing fun stuff) I buy Deus Ex off the bargain bin section for 10$. It's marked as a "Special Limited Edition", with no explaination of what a special limited edition is, precisely. (I looked.) I figured it meant it didn't come with a manual, or maybe multiplayer.
So I figure 10$ for a game that was supposedly pretty good is a good deal, so I buy it. Take it home. Play it for a while. Get halfway through the game. Splash screen comes up: "Buy Deus Ex, game of the year for 2000! If you buy the real version, you will get to see all these nifty features..."
Bah. Fricken Best Buy sold me a DEMO for 10 dollars!
Maybe I should have guessed that "Special Limited Edition" meant "demo", but all the other games I've bought from their bargain bin have been the real thing. Deceptive as hell.
-Bill Kerney
My mistake, the java source should be in the .jar, but I did a *.class instead of *.[class,java], ah well.
./jmail.script, ./test.script and ./nn.
You're right, we use strcat and strcpy, but with a fixed buffer size, fgets() and a terminating condition on reading too many lines, I can't see a buffer exploit in there, but you may disagree.
System() uses nothing read in from the email, it simply chdirs into the directory, and runs
If you'd like to modify it, contact me.
-Bill Kerney
Well, this isn't something I want to devote much time to, but if the interest exists, I can put it up there.
That's because 90-95% accuracy isn't very good. :)
I've been running it on my main mail account for the last couple days, and with a 95% accuracy rate, 5 of my 100 good messages are misclassified, including one from a friend that I hadn't heard from in a year.
I could definately do better using attributes from my own mailboxes instead of a generic one collected from 4800 sample emails.
RBL blocking 2 out of 900 spams is pretty bad.
. gz
D ME
:)
A friend and I wrote a neural net spam filter using the UCI Machine Learning Database (on spam), that gets 90-95% accuracy on classifying a message as a spam or not-spam. It's integrated with the mail delivery system via procmail, so you can set it up to deliver all Good mails to one mailbox and Spam mails to another.
It is available (free + open source) at:
http://www-cse.ucsd.edu/~wkerney/spamfilter.tar
The README is at:
http://www-cse.ucsd.edu/~wkerney/spamfilter.REA
For Solarius/Linux only, Windows users need not apply. The system works pretty darn well for a simple neural net, and can be greatly improved, so if anyone is interested in modifying our code, drop us a line!
William Kerney
UCSD
There were three arcade games I was in love with in 1992-1994. I need help tracking two of them down:
1) Cdashus -- Wonderful arcade RPG game, four classes to play. I've found it for the TG-16.
2) Time Warriors -- A top view scroller like Ikari Warriors, but with like 40 different characters to play. What platform was this on? I haven't been able to locate it.
3) -- A side scrolling RPG/action game where you had to go through forty levels to get to the top of a mountain. The main thing I remember about it is that you could get an NPC to follow you around, who would attack when you would attack -- there were like 6 different ones, from a ninja that tossed stars to an archer, to a wizard. Anyone know the name of this? About 1990-1991 when it came out.
Thanks,
Shaka