Click the right link. And no, it's not hacked, I simply have no particular reason for people looking at the index of my home directory, and "pwn3d" seemed more appropriate at the time than "This is the default page for James Longstreet."
I'd like to see you work your ass off for an entire semester, bury yourself in other people's C code for hundreds of programs, understand all the material, get As on the exams, and then fail because you weren't lucky enough -- and not be just a teeny bit pissed about it.
The reason is that we were instructed to look for "low hanging fruit," like sprintf(buffer_on_the_stack, "%s", untrusted_input), or while (ch = getc()) { buffer_on_the_stack[i] = ch; i++; }.
DJB's software doesn't have these kind of holes. If it has any, we weren't about to spend our time analyzing every little atom of it. $500 isn't enough for me to spend that much time on it.
Deployed Unix software, as defined for the purposes of this class, is something that the professor can put into Google and find references to people using it. Not just it's Sourceforge or Freshmeat page, but people actually using it.
Perhaps it's not as big of a deal in lower math classes, but I never would have learned (and, as a direct result, passed) Calculus 3 if it weren't for the exercises we did with Maple. I have trouble visualizing graphs, especially 3D graphs. I don't think visually, so this is incredibly difficult for me. Maple, Mathematica, Matlab, and probably open-source counterparts -- gnuplot frontends and so forth, are the easiest, most logical way for me to understand visual aspects of math for me.
Yes, but how many people who didn't know what Linux was fully understood that ad? I would say that most of them either decided that Linux was some new tech company they didn't care about, or to ask there local geek what it was, thereby spreading word of mouth.
Yes, but let's say I really like Lenner's sausages. You would throw me, your friend, out of your house for saying, "here, try these sausages, I think you'll like them?"
For me, the most effective advertising is *real* people promoting a product. If people I trust (or enough people I don't necessarily trust) say "hey, product x is really good for reason x,y, and z," I'm going to be interested in that product.
This is why it's called "open-source." Do you think people are putting "Get Firefox" buttons on their websites because it gives them more credit in the SpreadFirefox street team? I certainly hope not. They're doing it because they like Firefox.
I come from UIC, sister campus of the honorable UIUC. We got down to Chambana at about 5:30 Friday, and didn't do much that night, as all the Friday events were over. Saturday we heard Charles Leiserson from MIT talk about shared-memory multiprocessing in the Cilk language, that the MIT AI lab came up with. It was very interesting, especially the charts and raw numbers... the best performance they got was from the 8 queens problem. Running on one processor, it took T1 seconds (I forget the exact figure). Running on 8 processors, it took exactly T1/8 seconds. Absolute parallelism...
After lunch, we saw Jimmy Wales, the founder of Wikipedia. He seems like a really great guy... a millionaire from the futures and options market deciding to devote his life and fortune to amassing the sum of human knowledge in one central, but not controlling place.
Later Saturday night, we went over to the HOL (House of Leet) for a conference party they were throwing. I think I spotted Brad Kuhn, the executive director of the FSF, and I know for a fact that Ari, the maintainer of SourceForge, was DJing the party. root@sf.net spinning for a bunch of nerds... what a sight.
Sunday we saw Phil Zimmerman's talk. He started by saying that there was no reason for him to explain the technical basis behind PGP, because everyone pretty much knew what it was. So the whole talk was mostly about his legal battles getting PGP out in the world, getting it through customs without getting put in jail, etc. I thought it was specifically interesting that in order to get PGP 2.6.2 distributed worldwide, they had MIT Press print the source code in a very OCR-readable font, with checksums on every line, a rolling checksum for all previous lines in the page, and an MD5 for every file to simplify error correction. Books, apparently, can be exported, disks with source code cannot.
Finally, we got to see the final showdown of MechMania X. The teams had been coding for 24 hours, with only a 5 hour break from the labs. Despite some problems with the visualizations (they were testing an OpenGL one that looked really cool but didn't work very well), it was pulled off quite well. Some of the teams did nothing, or had bad logic that ended up killing themselves, but some were really interesting. Once a team won a match, the ships would inevitably do some sort of crazy dance, not having any idea what to do, since there was no longer anything to do...
It was a fun weekend. I would highly suggest attending next year's conference. Check http://acm.uiuc.edu for more information about the conference and the ACM.
In 2099, there is no difference between computers and humans. Molly and George have become the same "person." She no longer has a body, or if she does, she's not aware of it. The lines between virtual space and real space have been blurred.
At the end of the book, Ray and Molly have a discussion about how the world works in 2099, and the few people who are not "enhanced", referred to as MOSHs, Mostly Ordinary Substrate Humans. They have a discussion about music:
Molly: I'm really just dabbling but creating music is a great way for me to stay close with Jeremy and Emily
Ray: Creating music sounds like a good thing to do with your kids even if they are almost 90 years old. So could I hear it?
Molly: Uh I'm afraid you wouldn't understand it
Ray: So it requires enhancement to understand?
Molly: Yes most of it does. For starters, the symphonies use frequencies that a mosh can't hear and it has much too fast a tempo and it uses musical structures that a mosh could never follow
Ray: Can't you create art for non-augmented humans? I mean there's still a lot of depth possible. Consider Beethoven, he wrote almost two centuries ago and we still find his music exhilarating
Molly: Yes there is a genre of music - all the arts actually, where we create music and art that a mosh is capable of understanding
Ray: And then you play mosh music for moshes?
Molly: hmm now that's an interesting idea. I suppose we could try that, although moshes are not that easy to find anymore. It's really not necessary though, we can certainly understand what a mosh is capable of understanding. The point though is to use the mosh limitations as an added constraint.
Ray: Sort of like composing new music for old instruments?
Molly: Yeah, new music for old minds
After this, the appendix includes a complete history of the universe from the Big Bang to 2099. Extremely interesting, you should check it out.
CISC in general is designed for hand-coded assembly. Having a much larger instruction set, variable-width instructions, etc, is a performance hit. Compilers normally use a subset of the instructions available on a CISC chip. Why have an instruction to move the contents of one register to another? OR works for the same purpose... and the assembler can have macro instructions so that MOVE dest, src maps to OR dest, src, src. The chip has to do more work to decode instructions in CISC. Same thing with variable instructions... instead of reading a byte, figuring out if you need to read another one, repeat, why not just read 4 bytes and be done with it?
Like it or not, a lot of the vulnerabilities in Windows are due to the highly exploitable nature of the x86 architecture. Need a payload without NULLs? Okay. Need a payload that passes isalpha()? Okay.
x86, at this point in time, is a dirty hack. A 16-bit real mode BIOS is a dirty hack. I see no reason why CISC should be used on modern systems. How much code on your system is handcoded assembly? The x86 is designed for handcoded assembly. The PPC and other RISC chips are designed for compiled code.
It might be libel, but until it's proven libelous, it's speech. Cybersquatting is usually defined as buying a domain name in order to sell it back at huge cost. As far as I've heard, this guy isn't trying to make money off of the domain name.
2600 registers domain names like www.fuckcbs.com. They do this partly as a political statement, partly cause they can, and partly cause they're kind of jackasses and like to get cease and desists to show they're not afraid of the big bad man. But it's definitely free speech.
The problem here is that iPhoto isn't anywhere near the killer app that iTunes is, even with an iPod. I think it would be smarter for the iPod to be able to display, say, album art and images in a "Pictures" folder on the root level of the disk, and let people use whatever program (or not) they wanted. iPhoto doesn't have much in the way of features. If it wasn't bundled with the OS, I don't think most Mac users would use it.
Of course, I forgot. Music didn't exist until there was a record industry... it's absolutely incapable of existing without it.
If it seems like an artist is only making music in order to make money from it, I don't want to listen to them -- it usually means the music is bad. Just look at Metallica in the last few years. Even the most hardcore Metallica fans I know think their last few albums sucked.
This is not to say I won't support artists. I just don't support artists who don't seem to want to make music anymore than I want to help users when their email "stops working."
Or anyone can overflow the keyword buffer when it's parsing RTF messages.
"Oh, but GroupWise uses SSL so it's not vulnerable to attack..."
But gaim doesn't check the cert.
"Oh, but you can plainly see, in the unchecked for loop that overflows keyword, it only allows alphabetic characters, so you couldn't put shellcode in it."
But the x86 has several instructions that pass isalpha(). Use GroupWise on gaim and ride my 'AI' NOP sled.
Gaim can overflow a buffer when parsing RTF for the Novell GroupWise protocol, allowing any user with OpenSSL and a network connection to execute arbitrary code as the user running Gaim! Yay Gaim!
Record companies (not the RIAA) and artists (not Lars Ulrich) coming out against the DMCA and the restrictions against fair use and P2P. Get the artists to say that they make money off of filesharing. This is an old argument, but a true one... I first heard Modest Mouse when a friend of mine burned me a CD of theirs. I fell in love with the band, and bought that and their four other albums. I've also spread the word that Modest Mouse rocks my socks, and gotten several other people into them as well.
Nobody has ever brought together the world of documents, media and structured information in giving you one simple set of verbs that lets you richly find, move around and replicate those things.
And this is a FAT/NTFS issue... my 68k Mac from 1992 can find a file faster than a 3.4 GHz P4 with a gig of ram, if it's running Windows. Some filesystems are simply superior to others. The mistake MS made when making NTFS was to not provide it with any sort of indexing, making it impossible to search the directory tree without traversing each node.
HFS/+ has never had this problem. Hit Cmd-F on a System 7.0 box, type a partial filename, and bam... it's there. It's that simple.
That said, WinFS is a really cool idea, since we see hard drives getting bigger than anyone needs them for (read: room for metadata) and systems getting faster and faster (read: easier to parse through metadata). I do, however, wish it was an open implementation. This could be a chance for MS to gain some credibility with the F/OSS world.
Slashdot Mirror
Click the right link. And no, it's not hacked, I simply have no particular reason for people looking at the index of my home directory, and "pwn3d" seemed more appropriate at the time than "This is the default page for James Longstreet."
I'd like to see you work your ass off for an entire semester, bury yourself in other people's C code for hundreds of programs, understand all the material, get As on the exams, and then fail because you weren't lucky enough -- and not be just a teeny bit pissed about it.
60%. This assignment is worth 60% of the FINAL SEMESTER GRADE. I suppose I should have put that in the summary.
DJB's software doesn't have these kind of holes. If it has any, we weren't about to spend our time analyzing every little atom of it. $500 isn't enough for me to spend that much time on it.
Deployed Unix software, as defined for the purposes of this class, is something that the professor can put into Google and find references to people using it. Not just it's Sourceforge or Freshmeat page, but people actually using it.
to Kris Kubicki's mirror is here.
Perhaps it's not as big of a deal in lower math classes, but I never would have learned (and, as a direct result, passed) Calculus 3 if it weren't for the exercises we did with Maple. I have trouble visualizing graphs, especially 3D graphs. I don't think visually, so this is incredibly difficult for me. Maple, Mathematica, Matlab, and probably open-source counterparts -- gnuplot frontends and so forth, are the easiest, most logical way for me to understand visual aspects of math for me.
Someone in a previous comment stated that Encarta Atlas is very good, and that he knows of people buying Encarta solely for the Atlas.
Yes, but how many people who didn't know what Linux was fully understood that ad? I would say that most of them either decided that Linux was some new tech company they didn't care about, or to ask there local geek what it was, thereby spreading word of mouth.
Yes, but let's say I really like Lenner's sausages. You would throw me, your friend, out of your house for saying, "here, try these sausages, I think you'll like them?"
For me, the most effective advertising is *real* people promoting a product. If people I trust (or enough people I don't necessarily trust) say "hey, product x is really good for reason x,y, and z," I'm going to be interested in that product.
This is why it's called "open-source." Do you think people are putting "Get Firefox" buttons on their websites because it gives them more credit in the SpreadFirefox street team? I certainly hope not. They're doing it because they like Firefox.
Homer: Let the bears pay the bear tax! I pay the Homer tax!
Lisa: Dad, that's homeowner's tax.
Homer: Well, anyway, I'm still outraged.
So is this, according to Wikipedia. Show me spim in the OED and I'll call it a word. For now it's slang. Still not a misspelling though.
After lunch, we saw Jimmy Wales, the founder of Wikipedia. He seems like a really great guy... a millionaire from the futures and options market deciding to devote his life and fortune to amassing the sum of human knowledge in one central, but not controlling place.
Later Saturday night, we went over to the HOL (House of Leet) for a conference party they were throwing. I think I spotted Brad Kuhn, the executive director of the FSF, and I know for a fact that Ari, the maintainer of SourceForge, was DJing the party. root@sf.net spinning for a bunch of nerds... what a sight.
Sunday we saw Phil Zimmerman's talk. He started by saying that there was no reason for him to explain the technical basis behind PGP, because everyone pretty much knew what it was. So the whole talk was mostly about his legal battles getting PGP out in the world, getting it through customs without getting put in jail, etc. I thought it was specifically interesting that in order to get PGP 2.6.2 distributed worldwide, they had MIT Press print the source code in a very OCR-readable font, with checksums on every line, a rolling checksum for all previous lines in the page, and an MD5 for every file to simplify error correction. Books, apparently, can be exported, disks with source code cannot.
Finally, we got to see the final showdown of MechMania X. The teams had been coding for 24 hours, with only a 5 hour break from the labs. Despite some problems with the visualizations (they were testing an OpenGL one that looked really cool but didn't work very well), it was pulled off quite well. Some of the teams did nothing, or had bad logic that ended up killing themselves, but some were really interesting. Once a team won a match, the ships would inevitably do some sort of crazy dance, not having any idea what to do, since there was no longer anything to do...
It was a fun weekend. I would highly suggest attending next year's conference. Check http://acm.uiuc.edu for more information about the conference and the ACM.
In 2099, there is no difference between computers and humans. Molly and George have become the same "person." She no longer has a body, or if she does, she's not aware of it. The lines between virtual space and real space have been blurred.
At the end of the book, Ray and Molly have a discussion about how the world works in 2099, and the few people who are not "enhanced", referred to as MOSHs, Mostly Ordinary Substrate Humans. They have a discussion about music:
Molly: I'm really just dabbling but creating music is a great way for me to stay close with Jeremy and Emily
Ray: Creating music sounds like a good thing to do with your kids even if they are almost 90 years old. So could I hear it?
Molly: Uh I'm afraid you wouldn't understand it
Ray: So it requires enhancement to understand?
Molly: Yes most of it does. For starters, the symphonies use frequencies that a mosh can't hear and it has much too fast a tempo and it uses musical structures that a mosh could never follow
Ray: Can't you create art for non-augmented humans? I mean there's still a lot of depth possible. Consider Beethoven, he wrote almost two centuries ago and we still find his music exhilarating
Molly: Yes there is a genre of music - all the arts actually, where we create music and art that a mosh is capable of understanding
Ray: And then you play mosh music for moshes?
Molly: hmm now that's an interesting idea. I suppose we could try that, although moshes are not that easy to find anymore. It's really not necessary though, we can certainly understand what a mosh is capable of understanding. The point though is to use the mosh limitations as an added constraint.
Ray: Sort of like composing new music for old instruments?
Molly: Yeah, new music for old minds
After this, the appendix includes a complete history of the universe from the Big Bang to 2099. Extremely interesting, you should check it out.
CISC in general is designed for hand-coded assembly. Having a much larger instruction set, variable-width instructions, etc, is a performance hit. Compilers normally use a subset of the instructions available on a CISC chip. Why have an instruction to move the contents of one register to another? OR works for the same purpose... and the assembler can have macro instructions so that MOVE dest, src maps to OR dest, src, src. The chip has to do more work to decode instructions in CISC. Same thing with variable instructions... instead of reading a byte, figuring out if you need to read another one, repeat, why not just read 4 bytes and be done with it?
x86, at this point in time, is a dirty hack. A 16-bit real mode BIOS is a dirty hack. I see no reason why CISC should be used on modern systems. How much code on your system is handcoded assembly? The x86 is designed for handcoded assembly. The PPC and other RISC chips are designed for compiled code.
It might be libel, but until it's proven libelous, it's speech. Cybersquatting is usually defined as buying a domain name in order to sell it back at huge cost. As far as I've heard, this guy isn't trying to make money off of the domain name.
2600 registers domain names like www.fuckcbs.com. They do this partly as a political statement, partly cause they can, and partly cause they're kind of jackasses and like to get cease and desists to show they're not afraid of the big bad man. But it's definitely free speech.
The problem here is that iPhoto isn't anywhere near the killer app that iTunes is, even with an iPod. I think it would be smarter for the iPod to be able to display, say, album art and images in a "Pictures" folder on the root level of the disk, and let people use whatever program (or not) they wanted. iPhoto doesn't have much in the way of features. If it wasn't bundled with the OS, I don't think most Mac users would use it.
If it seems like an artist is only making music in order to make money from it, I don't want to listen to them -- it usually means the music is bad. Just look at Metallica in the last few years. Even the most hardcore Metallica fans I know think their last few albums sucked.
This is not to say I won't support artists. I just don't support artists who don't seem to want to make music anymore than I want to help users when their email "stops working."
"Oh, but GroupWise uses SSL so it's not vulnerable to attack..."
But gaim doesn't check the cert.
"Oh, but you can plainly see, in the unchecked for loop that overflows keyword, it only allows alphabetic characters, so you couldn't put shellcode in it."
But the x86 has several instructions that pass isalpha(). Use GroupWise on gaim and ride my 'AI' NOP sled.
Gaim can overflow a buffer when parsing RTF for the Novell GroupWise protocol, allowing any user with OpenSSL and a network connection to execute arbitrary code as the user running Gaim! Yay Gaim!
Where is it... It's on my hard drive somewhere... I have no idea.
Apparently there's a fire at the hosting company... Slashdot to the rescue! (batman theme song)
Record companies (not the RIAA) and artists (not Lars Ulrich) coming out against the DMCA and the restrictions against fair use and P2P. Get the artists to say that they make money off of filesharing. This is an old argument, but a true one... I first heard Modest Mouse when a friend of mine burned me a CD of theirs. I fell in love with the band, and bought that and their four other albums. I've also spread the word that Modest Mouse rocks my socks, and gotten several other people into them as well.
Yes, they have.
And this is a FAT/NTFS issue... my 68k Mac from 1992 can find a file faster than a 3.4 GHz P4 with a gig of ram, if it's running Windows. Some filesystems are simply superior to others. The mistake MS made when making NTFS was to not provide it with any sort of indexing, making it impossible to search the directory tree without traversing each node.
HFS/+ has never had this problem. Hit Cmd-F on a System 7.0 box, type a partial filename, and bam... it's there. It's that simple.
That said, WinFS is a really cool idea, since we see hard drives getting bigger than anyone needs them for (read: room for metadata) and systems getting faster and faster (read: easier to parse through metadata). I do, however, wish it was an open implementation. This could be a chance for MS to gain some credibility with the F/OSS world.