Well, what about buffer overflows in glibc, zlib or one of the other system libraries? Even if you're keeping Apache current (as the parent poster mentioned), you've still got other issues, potentially.
Better, IMHO, to try to stay current, and use a relatively secure OS, like OpenBSD.
Besides, upgrading regularly forces you you keep documentation on what you did for the install and setup. These docos also help with new administrators and disaster recovery.
I've never gotten a chance to implement one of these (my firewalls have to do other things too), but the thought of a stealthy firewall (which is nearly invisible) is so cool.
I think that OpenBSD's pf is more featureful and easier to use than anything on Linux.
Yup. We have to remember that science is the search for useful knowledge. Useful shortcuts for making estimates for things that can happen.
For example, you can use the law of gravitation to estimate where a planet will be, based on its mass, speed, and the mass of the sun. However, this calculation can never be completely accurate. Because you'd also need to take into account the other planets. And asteroids. And the galaxy. And you'd need the exact speed of the planet in relation to the sun. And you'd have to take into account the sun's mass changes second by second. As does the planet's. As does everything else. And then there's intersellar dust...
Sure, you can say most of that is down in the noise. Even the relatively simple estimate is good enough to get you to Mars, or send a probe to Neptune.
However, can something be completely true if it is off by 0.0000001 percent? The 'laws of nature' is just the set of estimates we have which work really well, and are relatively easy to use. Nothing more.
Yup. The airplane dope they used back then (to strengthen and waterproof the fabric) was pretty flamable itself.
And then if you coat it with powdered aluminum...
Yikes.
The hubub about hydrogen is all kind of silly, really. Sure, it is explosive if well mixed with oxygen, so you need to be a little more careful. But it's not that dangerous. It is also not that much more efficient than helium, which isn't so expensive anymore. If helium makes you feel safer, so be it.
Yup. The SCO case seems to rest on the assumption that because IBM worked with SCO previously, the only way to they could have improved Linux (for enterprise scalability) was to steal SCO's patented ideas.
Um, yeah, right. That's like saying there is only one way to create a process scheduler. Or VM subsystem. The complaint by SCO mostly consists of quotes from press releases. I didn't see anything like "IBM used such-and-such algorithm for the NUMA implementation on Linux 2.5.x". If they ever publish some specifics, then we can see if their case holds any water.
I have my doubts, just from my own experiences with IBM's handling of IP. The tend to be very anal-retentive about all of that, and very cautious about what they'll release pubically, and how. I think it is unlikely that they release (intentionally or not) anything for Linux that was patent-encumbered.
Most people walk around happy to sport logos everywhere: their t-shirts, shoes, cars, computers (or computer components). They actually pay for the privilege.
I still don't understand this myself. It is not as if the logo-adorned products are cheap either. They're usually more expensive.
Maybe I'm just a socially-innept geek, but I still can't figure the appeal of having 'Tommy Hilfinger' blazoned across my back.
I've got a few logo t-shirts. Most of them I got for free. The only ones I paid for are Tux the penguin, the Debian swirl, and OpenBSD. Oh yeah, and a Akira t-shirt.
Heck, I usually cut the leather "placard" off the back of my jeans. The 'Intel Inside' swirl is the first thing to go on a new PC (well, one that doesn't have an Althon in it).
At least the yahoo bloggers are moving in the right direction, getting paid to advertise.
We don't need manned missions either, just some good robots.
I'd like to see a couple sample return missions. One of the most intriguing ideas recently is the suggestion that there may have been life on Mars at one point.
Finding out if there was (or wasn't) life on Mars could tell us a lot about how likely there is life on other planets. Let's get some probes on there, and roam around a bit, dig up some stuff, and bring it back!
Until launch costs get much cheaper (and that's a whole 'nother rant), let's just do some good, meaningful science. We have the technology. NASA's existing budget (if we weren't building the ISS) is good for a dozen missions per year to the rest of the solar system, plus another spiffy space telescope.
Now's the chance to take the money from something that isn't nearly as useful (the shuttle and ISS) and put it into answering some questions about life, the universe, and everything.
I think the movie failed because it wasn't satisfying, nor it it seem to make much sense.
And no, I'm not talking the dumber-then-hell physics nonsense that usually permeates ST. I'm talking about the plot.
Cuts. Check IMDB for the movie's quotes. How many of those were actually in the movie? The cut out a lot of background stuff that explained why this stuff was happening, and what the character's motivations were.
For instance: That Romulan commander (woman). At first she appears she's going to be a toady for Shinzan. She was asking that Romulan admiral some pointed questions after the big ship goes to zap Earth, like she was trying to see if the admiral was going to betray Shinzan. Then, all of a sudden, she's betraying Shinzan? What? Why did she change her mind? Did we learn enough about her character to understand why she might change her mind?
Over and over the point of the movie is that people are good when they aspire to be better than themselves. That's what is supposed to make Data better than B4. It's a fine sentiment, but where is it actually shown in the movie? Saying it is fine for a book or something. But you've got to have it be a central part of the movie, or it is just a plattitude.
And how did Deanna learn to fight back against Shinzan and the Chancellor? Why didn't she do that the first time?
Oh yeah, and as for physics... Ships with impulse drive can go like 0.99c. If you decide to ram another ship, you're going to end up with a big cloud of plasma and debris, not some lame "crunch".
And the cloaking... why couldn't the Enterprise's brilliant engineers program the weapons to shoot right back at anything shooting at the ship? It doesn't matter if the Romulan ship is cloaked, it's shooting you right now! Right over there! Sheesh.
Well, there's different kinds of costs. And time has different value to different people and organizations.
At most schools the labor is cheap. Students are usually close to free, and even the teachers (or grad students) aren't very expensive.
In a typical school, they have a lot more labor available than hard dollars (which must come out of a shrinking budget somewhere).
As for complex documents: In my opinion you shouldn't be using a word processor anyway. Desktop publishing, research papers, and other things like that are best done on different applications. Word processing is fine for letters and memos, but that's about it.
Actually the mail server is OBSD 3.1. However, the Perl stuff on 3.1 doesn't include all the necessary libraries for SpamAssassin. When I tried installing them using CPAN, there was significant breakage. It could be I was doing something wrong.
Haven't had the time to look into that again.
I'd switch to OBSD 3.2, but the software RAID doesn't seem to work... Maybe now just wait for 3.3. and see how that goes.
I was running qmail, but then needed an IMAP server. Courier-IMAP was the best for Maildirs. Then I needed a web mail server. Sqwebmail (part of Courier) was the best for Maildirs, and used the same subfolder format.
Then I needed mail filtering, and wanted to use Maildrop because it works well with Maildirs. So I just bit the bullet and installed Courier.
Had some initial issues with configuration setup, but after that it's run pretty smooth. Now I just need to get SpamAssassin installed (need a new version of Perl on the mailserver).
Sendmail scared me off in the mid '90s. Haven't touched a.cf file since.
What's funny is this is also exactly what happens with the new Justice League show on Cartoon Network. Have a look at the normal and widescreen versions. They just cut off the top and botton, and viola, it is 'widescreen'.
Re:Well, Brin's article was mostly successful
on
David Brin On LOTR
·
· Score: 3, Interesting
The fact most of you also vented plenty of vitriol at Brin in the process is entirely beside the point; from what I know of the man he would consider that a success as well.
Well, he's not doing himself any favors in my book. I have quite enjoyed some of his work. I really liked, and will occasionally re-read "Startide Rising" and "Earth" along with his short fiction. The rest of the Uplift saga I didn't enjoy nearly as much (too wacky and preachey, IMHO).
However, I've been less impressed with his analysis of Star Wars (though he does make some good points) and rather disappointed with his recent analysis of LOTR. I doubt he'll be crying over lost royalties (such an interesting word, in this context) he might have gotten from me. Because of his off-target remarks on LOTR I am even less likely to try anything else written by him.
If I am an isolated case, it doesn't matter. But am I the only one who feels this way?
It has been mentioned in magazines like EE Times that the smaller the process size (0.18 micron, etc.) the shorter-lived the component will be. Thermal cycling (heating and cooling), and electron migration (where the current erodes the metal interconnects) mean that stuff these days will only last a few years of continuous power-on time.
I guess that's not so bad, it is just that some of the stuff made in the late 1980's and early 1990's will last just about forever after they make it out of the initial failure zone. Or at least the silicon will, the wire bonds may not, however...
It tends to rely on blocklists, many of which have demonstrated unfair practices in the past.
True. Spamassassin does use block lists as part of the score, but you can lower the scores for those, not use them at all. The scores aren't high enough to kill a message by itself, I believe the highest score for a block list is 3.0 with the default threshold being 5.0.
The more SpamAssassin is used, the more spammers will specifically avoid doing things SpamAssassin checks for.
And if spammers decide not to send me pr0n or other crap, that's a bad thing?
The only real problem I've had with SpamAssassin lately is that I'm stuck on version 2.20. My ISP needs to upgrade Perl before I can run more recent versions.:-(
I don't run stuff like SETI@Home, but lots of people do. Processing blocks probably shouldn't have much priority when you're doing stuff on the desktop.
I ran some numbers on this recently. I was looking just at DLT vs. VXA. All prices US Dollars. This doesn't include the price of the drive, because that is relatively minor.
For VXA-1, tape costs about $2/GB, retail price (you may be able to do better).
For DLT-IV, tape costs about $1.4/GB.
For VXA-2, tape costs about $1/GB. About the same for AIT-3.
If you can find decent and not too expensive hot-swap drive carriers, those 320GB drives at $300 USD almost start looking good for backup media themselves! They could be close to $1/GB if the carriers aren't too expensive.
All that above was uncompressed storage. Compression can cut those prices in half if you can use it with your data.
HDs can backup data real fast, especially if you're using rsync. The problem is the drives themselves are more fragile than tapes. Though you can easily damage a tape by dropping it too (especially DLTs). Tapes are a bit better in terms of temperature range.
Dunno about long-term archival storage. CDs or some other kind of optical would be a better bet than any kind of magnetic media for long-term.
I guess you're not familiar with the Purple one. He's been doing that for a long, long time, way before it was 733t. Go read the lyrics to the Purple Rain album.
O'Reilly makes a *very* important point about forcing governments to use Open Source software: it's morally reprehensible.
Feh. All the time companies make rules about how IT will be implemented. Why should the government be any different?
At a company I used to work for, there was a corporate mandate handed down that all laser printers needed to come from HP. They also mandated MS Exchange for e-mail accounts, which I thought was a bad idea. They also mandated a whole bunch of other things, like security standards. Stuff like every corporate laptop should have an encrypted harddrive, so that if it was stolen, the files couldn't be (easily) read from it. I thought that was a very good idea.
So how is a memo from the CIO's office in a private company any different that a law passed by the state legislature in a government agency? How is that morally reprehensible, huh?
It would be morally reprehensible if they passed a law that said everyone had to use OSS. But that's not even being suggested here.
I'm a CIO now (though of a small company). There is no 'right to choose' for everyone in the company as far as IT decisions go. I decide. If there's something you need that can't be done with the existing infrastructure, fine, we'll discuss it and come up with a solution. But there will always be times when someone says 'I want to use X', and I'll reply 'make it work with Y instead', because that's what fits in with my company's IT policies.
Slashdot Mirror
Well, what about buffer overflows in glibc, zlib or one of the other system libraries? Even if you're keeping Apache current (as the parent poster mentioned), you've still got other issues, potentially.
Better, IMHO, to try to stay current, and use a relatively secure OS, like OpenBSD.
Besides, upgrading regularly forces you you keep documentation on what you did for the install and setup. These docos also help with new administrators and disaster recovery.
Which is why the firms I've worked with NEVER bid fixed-price, always time-and-materials.
Must be nice to live in your world. Most people / companies I've dealt with want fixed-price.
In my opinion, a transparent firewall is an even better choice:
I've never gotten a chance to implement one of these (my firewalls have to do other things too), but the thought of a stealthy firewall (which is nearly invisible) is so cool.
I think that OpenBSD's pf is more featureful and easier to use than anything on Linux.
Yup. We have to remember that science is the search for useful knowledge. Useful shortcuts for making estimates for things that can happen.
For example, you can use the law of gravitation to estimate where a planet will be, based on its mass, speed, and the mass of the sun. However, this calculation can never be completely accurate. Because you'd also need to take into account the other planets. And asteroids. And the galaxy. And you'd need the exact speed of the planet in relation to the sun. And you'd have to take into account the sun's mass changes second by second. As does the planet's. As does everything else. And then there's intersellar dust...
Sure, you can say most of that is down in the noise. Even the relatively simple estimate is good enough to get you to Mars, or send a probe to Neptune.
However, can something be completely true if it is off by 0.0000001 percent? The 'laws of nature' is just the set of estimates we have which work really well, and are relatively easy to use. Nothing more.
Yup. The airplane dope they used back then (to strengthen and waterproof the fabric) was pretty flamable itself.
And then if you coat it with powdered aluminum... Yikes.
The hubub about hydrogen is all kind of silly, really. Sure, it is explosive if well mixed with oxygen, so you need to be a little more careful. But it's not that dangerous. It is also not that much more efficient than helium, which isn't so expensive anymore. If helium makes you feel safer, so be it.
Yup. The SCO case seems to rest on the assumption that because IBM worked with SCO previously, the only way to they could have improved Linux (for enterprise scalability) was to steal SCO's patented ideas.
Um, yeah, right. That's like saying there is only one way to create a process scheduler. Or VM subsystem. The complaint by SCO mostly consists of quotes from press releases. I didn't see anything like "IBM used such-and-such algorithm for the NUMA implementation on Linux 2.5.x". If they ever publish some specifics, then we can see if their case holds any water.
I have my doubts, just from my own experiences with IBM's handling of IP. The tend to be very anal-retentive about all of that, and very cautious about what they'll release pubically, and how. I think it is unlikely that they release (intentionally or not) anything for Linux that was patent-encumbered.
Most people walk around happy to sport logos everywhere: their t-shirts, shoes, cars, computers (or computer components). They actually pay for the privilege.
I still don't understand this myself. It is not as if the logo-adorned products are cheap either. They're usually more expensive. Maybe I'm just a socially-innept geek, but I still can't figure the appeal of having 'Tommy Hilfinger' blazoned across my back.
I've got a few logo t-shirts. Most of them I got for free. The only ones I paid for are Tux the penguin, the Debian swirl, and OpenBSD. Oh yeah, and a Akira t-shirt.
Heck, I usually cut the leather "placard" off the back of my jeans. The 'Intel Inside' swirl is the first thing to go on a new PC (well, one that doesn't have an Althon in it).
At least the yahoo bloggers are moving in the right direction, getting paid to advertise.
I would like to find out more about Mars.
We don't need manned missions either, just some good robots.
I'd like to see a couple sample return missions. One of the most intriguing ideas recently is the suggestion that there may have been life on Mars at one point.
Finding out if there was (or wasn't) life on Mars could tell us a lot about how likely there is life on other planets. Let's get some probes on there, and roam around a bit, dig up some stuff, and bring it back!
Until launch costs get much cheaper (and that's a whole 'nother rant), let's just do some good, meaningful science. We have the technology. NASA's existing budget (if we weren't building the ISS) is good for a dozen missions per year to the rest of the solar system, plus another spiffy space telescope.
Now's the chance to take the money from something that isn't nearly as useful (the shuttle and ISS) and put it into answering some questions about life, the universe, and everything.
Let's do it!
I think the movie failed because it wasn't satisfying, nor it it seem to make much sense.
And no, I'm not talking the dumber-then-hell physics nonsense that usually permeates ST. I'm talking about the plot.
Cuts. Check IMDB for the movie's quotes. How many of those were actually in the movie? The cut out a lot of background stuff that explained why this stuff was happening, and what the character's motivations were.
For instance: That Romulan commander (woman). At first she appears she's going to be a toady for Shinzan. She was asking that Romulan admiral some pointed questions after the big ship goes to zap Earth, like she was trying to see if the admiral was going to betray Shinzan. Then, all of a sudden, she's betraying Shinzan? What? Why did she change her mind? Did we learn enough about her character to understand why she might change her mind?
Over and over the point of the movie is that people are good when they aspire to be better than themselves. That's what is supposed to make Data better than B4. It's a fine sentiment, but where is it actually shown in the movie? Saying it is fine for a book or something. But you've got to have it be a central part of the movie, or it is just a plattitude.
And how did Deanna learn to fight back against Shinzan and the Chancellor? Why didn't she do that the first time?
Oh yeah, and as for physics... Ships with impulse drive can go like 0.99c. If you decide to ram another ship, you're going to end up with a big cloud of plasma and debris, not some lame "crunch".
And the cloaking... why couldn't the Enterprise's brilliant engineers program the weapons to shoot right back at anything shooting at the ship? It doesn't matter if the Romulan ship is cloaked, it's shooting you right now! Right over there! Sheesh.
Well, there's different kinds of costs. And time has different value to different people and organizations.
At most schools the labor is cheap. Students are usually close to free, and even the teachers (or grad students) aren't very expensive.
In a typical school, they have a lot more labor available than hard dollars (which must come out of a shrinking budget somewhere).
As for complex documents: In my opinion you shouldn't be using a word processor anyway. Desktop publishing, research papers, and other things like that are best done on different applications. Word processing is fine for letters and memos, but that's about it.
Actually the mail server is OBSD 3.1. However, the Perl stuff on 3.1 doesn't include all the necessary libraries for SpamAssassin. When I tried installing them using CPAN, there was significant breakage. It could be I was doing something wrong.
Haven't had the time to look into that again.
I'd switch to OBSD 3.2, but the software RAID doesn't seem to work... Maybe now just wait for 3.3. and see how that goes.
Yeah, I switched us to Courier recently.
I was running qmail, but then needed an IMAP server. Courier-IMAP was the best for Maildirs. Then I needed a web mail server. Sqwebmail (part of Courier) was the best for Maildirs, and used the same subfolder format.
Then I needed mail filtering, and wanted to use Maildrop because it works well with Maildirs. So I just bit the bullet and installed Courier.
Had some initial issues with configuration setup, but after that it's run pretty smooth. Now I just need to get SpamAssassin installed (need a new version of Perl on the mailserver).
Sendmail scared me off in the mid '90s. Haven't touched a .cf file since.
What's funny is this is also exactly what happens with the new Justice League show on Cartoon Network. Have a look at the normal and widescreen versions. They just cut off the top and botton, and viola, it is 'widescreen'.
The fact most of you also vented plenty of vitriol at Brin in the process is entirely beside the point; from what I know of the man he would consider that a success as well.
Well, he's not doing himself any favors in my book. I have quite enjoyed some of his work. I really liked, and will occasionally re-read "Startide Rising" and "Earth" along with his short fiction. The rest of the Uplift saga I didn't enjoy nearly as much (too wacky and preachey, IMHO).
However, I've been less impressed with his analysis of Star Wars (though he does make some good points) and rather disappointed with his recent analysis of LOTR. I doubt he'll be crying over lost royalties (such an interesting word, in this context) he might have gotten from me. Because of his off-target remarks on LOTR I am even less likely to try anything else written by him.
If I am an isolated case, it doesn't matter. But am I the only one who feels this way?
Yes the FBI can get this information, but first they have to prove to a judge that there is probable cause that you are breaking the law.
Well, they used to, anyway. You need to catch up on the Patriot Act and some of the other anti-terrorism legislation that's been passed recently.
Yup, and it's only going to get worse.
It has been mentioned in magazines like EE Times that the smaller the process size (0.18 micron, etc.) the shorter-lived the component will be. Thermal cycling (heating and cooling), and electron migration (where the current erodes the metal interconnects) mean that stuff these days will only last a few years of continuous power-on time.
I guess that's not so bad, it is just that some of the stuff made in the late 1980's and early 1990's will last just about forever after they make it out of the initial failure zone. Or at least the silicon will, the wire bonds may not, however...
It tends to rely on blocklists, many of which have demonstrated unfair practices in the past.
True. Spamassassin does use block lists as part of the score, but you can lower the scores for those, not use them at all. The scores aren't high enough to kill a message by itself, I believe the highest score for a block list is 3.0 with the default threshold being 5.0.
The more SpamAssassin is used, the more spammers will specifically avoid doing things SpamAssassin checks for.
And if spammers decide not to send me pr0n or other crap, that's a bad thing?
The only real problem I've had with SpamAssassin lately is that I'm stuck on version 2.20. My ISP needs to upgrade Perl before I can run more recent versions. :-(
I'm not a big fan of Perl either.
I don't run stuff like SETI@Home, but lots of people do. Processing blocks probably shouldn't have much priority when you're doing stuff on the desktop.
For more of that, check out the Rice Boy Hall of Shame
There's a few papers and other interesting bits on UWB and localizer technology at the AEther Wire & Location, Inc. Homepage
I ran some numbers on this recently. I was looking just at DLT vs. VXA. All prices US Dollars. This doesn't include the price of the drive, because that is relatively minor.
For VXA-1, tape costs about $2/GB, retail price (you may be able to do better).
For DLT-IV, tape costs about $1.4/GB.
For VXA-2, tape costs about $1/GB. About the same for AIT-3.
If you can find decent and not too expensive hot-swap drive carriers, those 320GB drives at $300 USD almost start looking good for backup media themselves! They could be close to $1/GB if the carriers aren't too expensive.
All that above was uncompressed storage. Compression can cut those prices in half if you can use it with your data.
HDs can backup data real fast, especially if you're using rsync. The problem is the drives themselves are more fragile than tapes. Though you can easily damage a tape by dropping it too (especially DLTs). Tapes are a bit better in terms of temperature range. Dunno about long-term archival storage. CDs or some other kind of optical would be a better bet than any kind of magnetic media for long-term.
I guess you're not familiar with the Purple one. He's been doing that for a long, long time, way before it was 733t. Go read the lyrics to the Purple Rain album.
Well, Taco should have phrased that:
Anyone know the proper way to dispose of a monitor, in my area?
That, sadly, is not an easy-to-answer question. It's real hit-and-miss as to if there is someplace nearby to dispose of stuff like CRTs.
O'Reilly makes a *very* important point about forcing governments to use Open Source software: it's morally reprehensible.
Feh. All the time companies make rules about how IT will be implemented. Why should the government be any different?
At a company I used to work for, there was a corporate mandate handed down that all laser printers needed to come from HP. They also mandated MS Exchange for e-mail accounts, which I thought was a bad idea. They also mandated a whole bunch of other things, like security standards. Stuff like every corporate laptop should have an encrypted harddrive, so that if it was stolen, the files couldn't be (easily) read from it. I thought that was a very good idea.
So how is a memo from the CIO's office in a private company any different that a law passed by the state legislature in a government agency? How is that morally reprehensible, huh?
It would be morally reprehensible if they passed a law that said everyone had to use OSS. But that's not even being suggested here.
I'm a CIO now (though of a small company). There is no 'right to choose' for everyone in the company as far as IT decisions go. I decide. If there's something you need that can't be done with the existing infrastructure, fine, we'll discuss it and come up with a solution. But there will always be times when someone says 'I want to use X', and I'll reply 'make it work with Y instead', because that's what fits in with my company's IT policies.
But why would they black it out?
This is not some paper-only copy, it's HTML! Just delete the offending text.
Putting in black bars is just wacky.