True, true... I would much more likely connect to a red (factory setting) "linksys" network than a blue (cloaked) one with a random SSID in Kismet. Mostly because the people with the factory configuration would never know I used it, and it wouldn't make a difference to them.
I went warwalking the other day around my block, and it's like 90% unencrypted, and 25% factory defaults. Can you think of any way to "get the word out" to my neighbors (just to improve everyone's security and make the world a better place) without looking like an idiot or an evil cracker?
Fortunately, MAC filtering and turning off the SSID makes it LESS likely that someone is going to set up outside their house and use their connection
It doesn't make it less likely that someone will go out of their way to use it, because those people have things like Kismet on hand. It only prevents the people who have naïve Windows XP boxen from accidentally connecting.
It's really a shame, because (/me just can't force himself to write "in this post 9/11 era")... well, because it is good to have a long-range, backup data network that can be driven by battery-powered portable devices and can route emergency position information across the nation (with VHF-HF digipeaters). It would be cool if packet had evolved into a parallel Internet, and hams could basically use the Internet (albeit at 9600 baud) anywhere near a digipeater.
Yes. It's find to realize that Windows has improved, but Microsoft is still so far from a non-dysfunctional software company that it is not deserving of anyone's love.
It is? I grepped the kernel for soundmodem and only found one reference to it, in a file called hdlcdrv.c. The page for soundmodem suggests that it is a usermode program that integrates with the kernel AX25 stack. I'm confused...
Even in Linux this service is available, as part of the standard kernel.
Really?! The Linux kernel has a built-in soft modem? I thought it just understood the higher-level packet protocols like AX.25. All I can find is: Device Drivers > Networking Support > Amateur Radio support > Packet Radio protocols > Amateur Radio AX.25 Level 2 protocol. If it's not in the kernel, is there a software TNC/modem package for Linux?
My brother explained to me that this mouse cannot in fact tell which button you are pressing. If, like everyone on earth, you rest both your fingers on the mouse when you click, it doesn't know which button you're pushing. It's really lame.
I wish there was a company that made an OS that was better than Windows, hardware that was cheaper than Apple, and software that is better supported (e.g., Linux + better driver support from manufacturers) than Linux. (Not that Linux isn't well-supported, but obviously Windows, by way of market share, has better driver support.) Windows is too crappy, Apple/Mac OS X is too expensive, cult-powered, and toy-like, and Linux doesn't have enough drivers.
It's not that we're criticizing them for making a multiple-button mouse. We're criticizing them because they have an operating system that has clearly needed multiple buttons (unless you want to go all the way up to the menus, you use option-click to pop up the context menu) for a while and that they refuse to just accept a multiple button mouse. This mouse still keeps with the idea that "one button is easy", when it really isn't. I use a PC most of the time, and when I use a Mac, I find that things like making Command-Tab (a.k.a. Alt-Tab) switch applications instead of windows, using modifier keys instead of mouse buttons, and having applications in different places in the Dock depending on what I'm doing with them (closed or visible apps on the left, minimized windows on the right... it should have a launcher menu, and then just all the open apps, visible or not. it would be much easier to switch windows than Command-Tab, digging through them, or Exposé), are all annoying. All the little "simplifications" get in the way. Yes, you do need to do multiple things to one on-screen object. No, the solution is not to put everything else in the menu. No, the solution is not to remove functionality. It's to have multiple, consistent, well-defined buttons.
They can only sense the number of fingers, not which one. Synaptics touchpads, for example, return x, y, z (pressure), f (number of fingers), and w (width). Apple's driver translates 2 fingers down into scrolling; the Linux (maybe Windows too) driver translates one finger = left click, two fingers = right click, three fingers = middle click, and uses the right and bottom for scrolling.
I think they should just put one flat piece of plastic at the bottom, and attach two microswitches to it. In grandma mode, the switches do the same thing. In geek mode, the switches are separated into 2 buttons. (This might also make the button easier to push, maybe...)
Given what your average person does with a computer, they are in fact running many background tasks at any given time. On a Windows XP box being run by anyone normal, you'll probably find some sort of system software (anti-virus, Windows Update, Windows Firewall and/or ZoneAlarm), some sort of P2P software (hopefully Shareaza;-), maybe some email software, and a game running over all of it. If they're media nuts, it's possible they're ripping a CD, encoding some files, and/or recording a TV show off a tuner card.
Basically Dell and co. should advertise dual-core not as "Run a few specially designed apps faster, and everything else the same!" but as "Do two things at once without skipping a beat!", e.g. you can rip and encode music, run P2P, etc... without having games lag.
Of course, just like with CPU speed and hard drive space, programmers with über-fast boxen will ensure that their software expands to fill the available resources, and in 2 years we will be looking for quad-core chips so we can run 2 dual-threaded apps at once!;-)
(This is a bit offtopic, but still related to Trusted Computing...) You know, I think the whole Trusted Computing thing has gone on long enough with the industry spreading FUD about how it will enable "an enhanced, market-driven dynamic media distribution paradigm shift" (e.g., lots of fancy buzzwords meaning that we aren't allowed to share media anymore). We should start spreading equal FUD, saying that "Foobar chipset computers will keep you from sharing media with your friends." Realisticly, when used by M$, TC will extend to include banning P2P programs as best as they can and enforcing blanket restrictions on what you can do with media. People will hear this and think, "I'm not buying *that* piece of shit! It won't run KaZaA and all those lame spyware-ridden P2P apps! How will I *ever* get music without downloading it from some *other* person who *did* buy the CD?!" In other words, they will call BS like we have all been doing. Any thoughts?
You know, I was actually slightly thinking about buying a Mac. I figured, if the prices drop enough when they switch to Intel, I can buy one and run Mac OS X and Linux (probably Gentoo) on it. But I am absolutely not ever buying an Intel Mac for myself, just because the next logical step after restricting the OS from booting on "non-trusted" BIOSes is restricting the BIOS from booting "non-trusted" OSes, like, oh, say, Linux.
I will build my own CPU out of TTL chips and wire-wrap if it comes down to it, but I'm never accepting a computer where Trusted Computing is used against me. (Frankly, it's an awesome idea to include the chip, but only if it is used for user-requested cryptography. Using it against the user is BAD!)
*This is based on the assumption that the average Windows user is, in fact, a chimpanzee.
This is a valid assumption. The other day, I asked my mom to go to http://www.whatismyip.com/ because she had just reconnected the Internet (dog kicked out the phone cord) and I didn't want to wait for ddclient to update my IP, and she gave me an IP. I pinged it, and it said "PING crawl12-34-56-78.googlebot.com (12.34.56.78)...". She put the address in to google and read me the search result instead of going to the site.
Depressingly, you're probably right that pasting into the address bar is far, far beyond the level of knowledge of the average user.
What about black and white printers? I have an hp LaserJet 1320, and I'm curious if it might have one of these. I would guess they don't because the microdots are probably designed to catch counterfeiters. Can you provide any insight? OH, and don't forget, Slashdot has repeatedly refused lawyer-powered requests for the usernames of anonymous posters.;-)
Agreed... he's trying to make installers work so you don't have to "bootstrap" off a Windows installation to get programs installed, and people complain anyway. It's like "Geez, the Linux kernel sucks--I heard they're adding $REALLY_COOL_FEATURE but it sux because my $CRAPPY_WINDOWS_ONLY_HARDWARE doesn't work." Be happy for what they are doing. It'd be great if "Full Windows Compatibility" or close to it could be a selling point of e.g., Ubuntu. Then you can say to someone: "You can buy Windows for $200, or you can buy Ubuntu for $5 (media cost)--Ubuntu is cheaper, better, and it runs Windows programs too!
The fan is certainly spinning (it's pretty loud), and there's no dust severely clogging it (there's some on the center motor part, but none where it shouldn't be). It does spit hot air out the back, so I assume something's working. It does get a little hot to the touch (not where you will burn yourself by touching it, but where if you leave the laptop on your bare legs for a while your skin will be a little red.) I do wonder sometimes if the sensor is a bit screwed up, since it rarely goes below 46-48 degrees even after cooling down for a long time. (Interestingly, it seems to be working a little better today--the temperature is changing a lot more quickly and it was slowing down as it reached 85.) The hard drive is actually hotter externally, since it's not explicitly cooled and is closer to the bottom panel of the laptop. It's only been up to 50 itself, and is normally around 45. I'm running a script to record speed and temperature over a long period, if I manage to get it graphed with gnuplot I'll reply to myself and you can take a look.
My Intel Pentium M 735 (Dothan, 1.7 GHz, 400 MHz FSB, 2 MB L2 cache) runs at about 50C idle and 96C full load. It seems absolutely insanely high, but it works--what can I say?
Is it true that Pentium M's can run up to insane temperatures? (I heard the limit is higher on the Pentium M than the Pentium 4...)
It helps you feel more comfortable because you start out with something that is completely working, and you aren't always doing system administration using it. First, it's good because you aren't depending on your shell skills to run the system. It's a comfortable way to experiment because, if something isn't working or you can't figure it out, you can always go back to the GUI. Second, it's good because system administration doesn't have to be done in the shell. I run Gentoo, and I've b0rked my system a couple times by doing something dumb like downgrading glibc. On Mac OS X, that's not really a problem because administration tasks like software updates are done in the GUI. In other words, using a shell is not essential to administrating the system. What Mac OS X provides you with is a fully-functional GUI layered on top of a Unix core that you can directly access at your own leisure. It lets you tinker without asking anything in return.
He's not suggesting that you should update production machines without testing, but rather that when you use open source software, you don't need a second license for everything on the test box.
Oh, I wasn't suggesting it would be hard to implement, but that it would get swallowed up in Microsoft's patch process.
Slightly offtopic, but it's a real PITA when running Windows Update on a very out-of-date computer to have to install.Net 1.0, then upgrade to.Net 1.1, then download a security patch. Can't *Microsoft* slipstream their *own* patches for even *one* component?
Probably within a week there will be updates to the timezone definition files (/usr/share/zoneinfo/...). Not a big deal for Linux users. But Windows users will probably have to wait for the next semi-millennial patch cycle.
The point was that Creative Commons gives you an easy way to express your wishes as to what rights to grant people. Without it, you have to figure out the legalese to write yourself, and chances are you'll leave a loophole and won't have the time or money to fight someone if they take advantage of it. With Creative Commons, you are sure that your choices are expressed properly, both in legalese and in English. Basically, if you want to selectively share your works, Creative Commons is like those "Living-Will-and-Testament-in-a-box" thingies--it gives you pre-written legalese so you don't have to hire a lawyer.
They're striking a deal that gives the students DRM-free music! On Linux! No more "oh, your operating system isn't controlled by SuperMegaEvilCorp, so we can't support it", just "I guess we'll have to trust them not to copy the OGG's, just like we have to trust them not to copy CD's they buy". Be happy!
This is already done. It's called graylisting. Here's a website about it. Basically, you examine the unique combination of sender, recipient, and IP. The first time they connect, you return a "temporary failure" message. You continue doing this for a period of time (maybe about one hour), and then you accept the mail. The idea is that spammers, who use bulk-mailing programs, won't have the time or reason to resend a message, but that normal, well-behaved mail servers will. (This also means graylisting has to be employed on the mail server where the mail gets in. Once a "real" mail server receives the message, graylisting can't help.) I use Sneakemail, which is similar to Spamgourmet but a little more featureful, and it offers optional graylisting of addresses. I've used it on the (not spam armored) address posted on my website/blog, and it has filtered every piece of spam so far.
Slashdot Mirror
True, true... I would much more likely connect to a red (factory setting) "linksys" network than a blue (cloaked) one with a random SSID in Kismet. Mostly because the people with the factory configuration would never know I used it, and it wouldn't make a difference to them.
I went warwalking the other day around my block, and it's like 90% unencrypted, and 25% factory defaults. Can you think of any way to "get the word out" to my neighbors (just to improve everyone's security and make the world a better place) without looking like an idiot or an evil cracker?
It doesn't make it less likely that someone will go out of their way to use it, because those people have things like Kismet on hand. It only prevents the people who have naïve Windows XP boxen from accidentally connecting.
It's really a shame, because (/me just can't force himself to write "in this post 9/11 era")... well, because it is good to have a long-range, backup data network that can be driven by battery-powered portable devices and can route emergency position information across the nation (with VHF-HF digipeaters). It would be cool if packet had evolved into a parallel Internet, and hams could basically use the Internet (albeit at 9600 baud) anywhere near a digipeater.
Yes. It's find to realize that Windows has improved, but Microsoft is still so far from a non-dysfunctional software company that it is not deserving of anyone's love.
It is? I grepped the kernel for soundmodem and only found one reference to it, in a file called hdlcdrv.c. The page for soundmodem suggests that it is a usermode program that integrates with the kernel AX25 stack. I'm confused...
Really?! The Linux kernel has a built-in soft modem? I thought it just understood the higher-level packet protocols like AX.25. All I can find is: Device Drivers > Networking Support > Amateur Radio support > Packet Radio protocols > Amateur Radio AX.25 Level 2 protocol . If it's not in the kernel, is there a software TNC/modem package for Linux?
You know, I'm starting to agree.
My brother explained to me that this mouse cannot in fact tell which button you are pressing. If, like everyone on earth, you rest both your fingers on the mouse when you click, it doesn't know which button you're pushing. It's really lame.
I wish there was a company that made an OS that was better than Windows, hardware that was cheaper than Apple, and software that is better supported (e.g., Linux + better driver support from manufacturers) than Linux. (Not that Linux isn't well-supported, but obviously Windows, by way of market share, has better driver support.) Windows is too crappy, Apple/Mac OS X is too expensive, cult-powered, and toy-like, and Linux doesn't have enough drivers.
It's not that we're criticizing them for making a multiple-button mouse. We're criticizing them because they have an operating system that has clearly needed multiple buttons (unless you want to go all the way up to the menus, you use option-click to pop up the context menu) for a while and that they refuse to just accept a multiple button mouse. This mouse still keeps with the idea that "one button is easy", when it really isn't. I use a PC most of the time, and when I use a Mac, I find that things like making Command-Tab (a.k.a. Alt-Tab) switch applications instead of windows, using modifier keys instead of mouse buttons, and having applications in different places in the Dock depending on what I'm doing with them (closed or visible apps on the left, minimized windows on the right... it should have a launcher menu, and then just all the open apps, visible or not. it would be much easier to switch windows than Command-Tab, digging through them, or Exposé), are all annoying. All the little "simplifications" get in the way. Yes, you do need to do multiple things to one on-screen object. No, the solution is not to put everything else in the menu. No, the solution is not to remove functionality. It's to have multiple, consistent, well-defined buttons.
They can only sense the number of fingers, not which one. Synaptics touchpads, for example, return x, y, z (pressure), f (number of fingers), and w (width). Apple's driver translates 2 fingers down into scrolling; the Linux (maybe Windows too) driver translates one finger = left click, two fingers = right click, three fingers = middle click, and uses the right and bottom for scrolling.
I think they should just put one flat piece of plastic at the bottom, and attach two microswitches to it. In grandma mode, the switches do the same thing. In geek mode, the switches are separated into 2 buttons. (This might also make the button easier to push, maybe...)
Given what your average person does with a computer, they are in fact running many background tasks at any given time. On a Windows XP box being run by anyone normal, you'll probably find some sort of system software (anti-virus, Windows Update, Windows Firewall and/or ZoneAlarm), some sort of P2P software (hopefully Shareaza ;-), maybe some email software, and a game running over all of it. If they're media nuts, it's possible they're ripping a CD, encoding some files, and/or recording a TV show off a tuner card.
Basically Dell and co. should advertise dual-core not as "Run a few specially designed apps faster, and everything else the same!" but as "Do two things at once without skipping a beat!", e.g. you can rip and encode music, run P2P, etc... without having games lag.
Of course, just like with CPU speed and hard drive space, programmers with über-fast boxen will ensure that their software expands to fill the available resources, and in 2 years we will be looking for quad-core chips so we can run 2 dual-threaded apps at once! ;-)
(This is a bit offtopic, but still related to Trusted Computing...) You know, I think the whole Trusted Computing thing has gone on long enough with the industry spreading FUD about how it will enable "an enhanced, market-driven dynamic media distribution paradigm shift" (e.g., lots of fancy buzzwords meaning that we aren't allowed to share media anymore). We should start spreading equal FUD, saying that "Foobar chipset computers will keep you from sharing media with your friends." Realisticly, when used by M$, TC will extend to include banning P2P programs as best as they can and enforcing blanket restrictions on what you can do with media. People will hear this and think, "I'm not buying *that* piece of shit! It won't run KaZaA and all those lame spyware-ridden P2P apps! How will I *ever* get music without downloading it from some *other* person who *did* buy the CD?!" In other words, they will call BS like we have all been doing. Any thoughts?
You know, I was actually slightly thinking about buying a Mac. I figured, if the prices drop enough when they switch to Intel, I can buy one and run Mac OS X and Linux (probably Gentoo) on it. But I am absolutely not ever buying an Intel Mac for myself, just because the next logical step after restricting the OS from booting on "non-trusted" BIOSes is restricting the BIOS from booting "non-trusted" OSes, like, oh, say, Linux.
I will build my own CPU out of TTL chips and wire-wrap if it comes down to it, but I'm never accepting a computer where Trusted Computing is used against me. (Frankly, it's an awesome idea to include the chip, but only if it is used for user-requested cryptography. Using it against the user is BAD!)
This is a valid assumption. The other day, I asked my mom to go to http://www.whatismyip.com/ because she had just reconnected the Internet (dog kicked out the phone cord) and I didn't want to wait for ddclient to update my IP, and she gave me an IP. I pinged it, and it said "PING crawl12-34-56-78.googlebot.com (12.34.56.78) ...". She put the address in to google and read me the search result instead of going to the site.
Depressingly, you're probably right that pasting into the address bar is far, far beyond the level of knowledge of the average user.
What about black and white printers? I have an hp LaserJet 1320, and I'm curious if it might have one of these. I would guess they don't because the microdots are probably designed to catch counterfeiters. Can you provide any insight? OH, and don't forget, Slashdot has repeatedly refused lawyer-powered requests for the usernames of anonymous posters. ;-)
Agreed... he's trying to make installers work so you don't have to "bootstrap" off a Windows installation to get programs installed, and people complain anyway. It's like "Geez, the Linux kernel sucks--I heard they're adding $REALLY_COOL_FEATURE but it sux because my $CRAPPY_WINDOWS_ONLY_HARDWARE doesn't work." Be happy for what they are doing. It'd be great if "Full Windows Compatibility" or close to it could be a selling point of e.g., Ubuntu. Then you can say to someone: "You can buy Windows for $200, or you can buy Ubuntu for $5 (media cost)--Ubuntu is cheaper, better, and it runs Windows programs too!
The fan is certainly spinning (it's pretty loud), and there's no dust severely clogging it (there's some on the center motor part, but none where it shouldn't be). It does spit hot air out the back, so I assume something's working. It does get a little hot to the touch (not where you will burn yourself by touching it, but where if you leave the laptop on your bare legs for a while your skin will be a little red.) I do wonder sometimes if the sensor is a bit screwed up, since it rarely goes below 46-48 degrees even after cooling down for a long time. (Interestingly, it seems to be working a little better today--the temperature is changing a lot more quickly and it was slowing down as it reached 85.) The hard drive is actually hotter externally, since it's not explicitly cooled and is closer to the bottom panel of the laptop. It's only been up to 50 itself, and is normally around 45. I'm running a script to record speed and temperature over a long period, if I manage to get it graphed with gnuplot I'll reply to myself and you can take a look.
My Intel Pentium M 735 (Dothan, 1.7 GHz, 400 MHz FSB, 2 MB L2 cache) runs at about 50C idle and 96C full load. It seems absolutely insanely high, but it works--what can I say?
Is it true that Pentium M's can run up to insane temperatures? (I heard the limit is higher on the Pentium M than the Pentium 4...)
It helps you feel more comfortable because you start out with something that is completely working, and you aren't always doing system administration using it. First, it's good because you aren't depending on your shell skills to run the system. It's a comfortable way to experiment because, if something isn't working or you can't figure it out, you can always go back to the GUI. Second, it's good because system administration doesn't have to be done in the shell. I run Gentoo, and I've b0rked my system a couple times by doing something dumb like downgrading glibc. On Mac OS X, that's not really a problem because administration tasks like software updates are done in the GUI. In other words, using a shell is not essential to administrating the system. What Mac OS X provides you with is a fully-functional GUI layered on top of a Unix core that you can directly access at your own leisure. It lets you tinker without asking anything in return.
Um... $500 is the price of a Mac Mini. At that point he could just buy the Mac and do it his way.
He's not suggesting that you should update production machines without testing, but rather that when you use open source software, you don't need a second license for everything on the test box.
Oh, I wasn't suggesting it would be hard to implement, but that it would get swallowed up in Microsoft's patch process.
Slightly offtopic, but it's a real PITA when running Windows Update on a very out-of-date computer to have to install .Net 1.0, then upgrade to .Net 1.1, then download a security patch. Can't *Microsoft* slipstream their *own* patches for even *one* component?
Probably within a week there will be updates to the timezone definition files (/usr/share/zoneinfo/...). Not a big deal for Linux users. But Windows users will probably have to wait for the next semi-millennial patch cycle.
The point was that Creative Commons gives you an easy way to express your wishes as to what rights to grant people. Without it, you have to figure out the legalese to write yourself, and chances are you'll leave a loophole and won't have the time or money to fight someone if they take advantage of it. With Creative Commons, you are sure that your choices are expressed properly, both in legalese and in English. Basically, if you want to selectively share your works, Creative Commons is like those "Living-Will-and-Testament-in-a-box" thingies--it gives you pre-written legalese so you don't have to hire a lawyer.
I second that!
They're striking a deal that gives the students DRM-free music! On Linux! No more "oh, your operating system isn't controlled by SuperMegaEvilCorp, so we can't support it", just "I guess we'll have to trust them not to copy the OGG's, just like we have to trust them not to copy CD's they buy". Be happy!
This is already done. It's called graylisting. Here's a website about it. Basically, you examine the unique combination of sender, recipient, and IP. The first time they connect, you return a "temporary failure" message. You continue doing this for a period of time (maybe about one hour), and then you accept the mail. The idea is that spammers, who use bulk-mailing programs, won't have the time or reason to resend a message, but that normal, well-behaved mail servers will. (This also means graylisting has to be employed on the mail server where the mail gets in. Once a "real" mail server receives the message, graylisting can't help.) I use Sneakemail, which is similar to Spamgourmet but a little more featureful, and it offers optional graylisting of addresses. I've used it on the (not spam armored) address posted on my website/blog, and it has filtered every piece of spam so far.