I never used FreeBSD, but your/bin/cp problem is simple: you're bumping into the command line length limit. It exists in Linux too, although from what you say it's larger. Just work around it, for example:
Is that a really weird progress bar, or they really offer no indication at all of how much has installation progressed? Sometimes I don't have the whole day, so a good indication of how much has been done and how much is left is very valuable to me. An estimation of how long will it take would be nice too.
Now that I think about it, this is really cool. I would be able to avoid being tracked by switching to random IPs inside that 48 bits of address space.
That looks pretty silly to me too. Why would you want to have 2^68 addresses for, to give an IP number to every piece of dust in your house? The fact that there's a lot of address space available doesn't mean you have to waste it like that. I'm sure that 256 or 65536 addresses would have been enough. Hell, you've got more address space than MIT! They have a class A network (16777216 addresses).
I've got a Duron 850 with 768MB RAM, opening Konqueror takes about 3 seconds. Opening a new window for this reply took less than I can reliably measure though, maybe 0.5 seconds or so.
Well, I'm using KDE 3.0.4 on Debian and it works great. Stuff doesn't load immediately though, but what's the problem with that anyway? It's not like you're going to start a program just to close it in 5 seconds. When I open a link in Konqueror in a new window it appears immediately.
This is an interesting question, will Linux ever become public domain? I don't think so. Linux is not really Linus' anymore. It has lots of code from hundreds of people, and I think every of those people have copyright in their code. I'm pretty sure 2.4 is nothing like the first version of the kernel and there's nothing of the original code in it.
Compiling stuff is actually an advantage Linux has. If you don't want to you don't have to, just use any distribution. It's an advantage because it lets you try unfinished or unsupported stuff. When was the last time you could try an experimental file system or driver on Windows? I'm pretty sure you never did that. On Windows most of the time you can only try software when the developer lets you, and I've never heard of experimental Windows features being available with source code.
If you don't like compiling it's just fine, you can just use Linux just as if it was Windows and install only compiled software. You don't need to compile the kernel either, just use the one that comes with your distribution
I'm sure you could adapt quicky, just like I could have if I spent a few days with it. But are you sure that the first time you saw a Mac you could immediately find how to write a document, how to print or how to change the mouse's sensitivity? I'm sure it took you at least a few hours to figure where things are.
What about tastes, my favourite GUI is KDE 3, which curiously enough took less time to get used to than the Mac. But it took weeks to really learn all its features.
Heh, that's very funny. I remember I tried an iMac once. I can't do absolutely anything useful with it, and I've been using computers and Windows for years. I see lots of people talking about how intuitive is $GUI. That's a plain lie. Any GUI requires getting used to it. If a really intuitive one is ever made it will work by reading your mind.
It'd be tough, I'm sure. A boot virus would kill LILO or whatever is used, making it pretty useless. And even if it infected it I'm not sure it'd survive the switch to protected mode and kernel initialization. If the virus infected the kernel then I don't think it could be called "boot virus" anymore.
Re:Well, there's one way to deal with them....
on
Dealing with the RIAA?
·
· Score: 4, Funny
I wonder, what if a few hundred people did a complete mirror of the RIAA server every day? If somebody asked they could always say that they *love* the RIAA and are mirroring the site because it's so unstable lately and they don't want to miss anything;-)
For example. Unlike Windows, Linux usually works everywhere unless you have really different hardware. A Debian kernel should have everything compiled as modules, so it should be fine. If you use your own kernel make sure it supports all the hardware on all the computers you have, plus some common hardware you might get in the future just in case.
Things you could do are boot a computer with a Linux floppy, connect to the net, get the image from there and overwrite the disk. Or write the files to a CD and do the same. You could then use rsync to keep machines up to date.
Nah, it's just that lately switches are getting cheaper. My old 5 port hub sent packets everywhere so you could sit on any computer and sniff the whole network. With a switch you can only see the traffic from and to your computer (unless it's a big expensive switch that lets you sniff everything). That's pretty much the definition of a switch, it switches traffic instead of just sending it to every cable.
Since now small switches got pretty cheap I wouldn't be surprised there's some confusion. Besides, they both do the same function, just in slightly different ways.
Okay, how would you predict the time between my keystrokes? You couldn't get it right with the precision to a nanosecond, and can't predict when I'll take a pause to go to the bathroom or decide to scratch my nose.
I don't think you can predict random radio noise either, as it's a combination of many things: environment (cars, computers, microwave), the microphone, the cable, my computer, the time and location... if you record random noise anywhere else it'll be different
First you have to know where in the encrypted data is whatever you want. XORing it with the victim's account number won't give you the location of that account in the ciphertext. In most cases it's probably complicated. I doubt banks exchange information in clearly delimited chunks. Over one TCP/IP connection they could pass thousands of transaction that can perfectly have a different length.
Second, that's trivially thwarted by the following protocol: 1. Bank A calculates all the data to send and its hash. 2. Bank A sends hash followed by the cyphertext 3. You do your substitution 4. Bank B checks the checksum and sees something changed.
You can't break this because you only know how to change the account number - you know it's say bytes 100 to 104. But you don't know the rest of the information so you can't replace the hash as well.
Nope. A random number sequence is simply a sequence that can't be predicted, and you can obtain one in many ways, like listening to radio noise, having a computer record from the microphone entry with no microphone connected (to record the noise introduced by ambient and electronics), or using data from the mouse, network, keyboard and disk access like the Linux kernel does.
I said *often*. If you need to be able to get support *now* and get an immediate fix, then obviously you need a support contract. But contrary to how many people think, there's not so much stuff that has to be available 24/7. For example if you're a small company that does business only during the day you could take the database server down at night and nobody would notice.
There are tons of places with information about security, btw, like bugtraq, and general information about securing Linux is available on the net and you can always ask people on IRC.
What about phone support you might get it free too with some luck. If your requirements aren't big (you just need some help to set it up) you might get lucky and find some Linux evangelist that would do some work for free just for the sake of removing MS software from a few computers. I helped with the servers at my school and nobody paid me for it. The teacher just asked nicely.
Again, I didn't, and am not saying that this is an option for everybody.
vadim@alice:~$ traceroute batasuna.org traceroute to batasuna.org (161.58.228.92), 30 hops max, 38 byte packets
1 router (192.168.0.1) 1.010 ms 0.816 ms 0.744 ms
2 10.3.50.1 (10.3.50.1) 66.757 ms 63.716 ms 63.745 ms
3 83.Red-80-58-11.pooles.rima-tde.net (80.58.11.83) 63.761 ms 60.499 ms 61.945 ms
4 37.Red-80-58-76.pooles.rima-tde.net (80.58.76.37) 66.936 ms 71.502 ms 63.596 ms
5 17.Red-80-58-72.pooles.rima-tde.net (80.58.72.17) 65.170 ms 70.477 ms 71.765 ms
6 242.Red-80-58-73.pooles.rima-tde.net (80.58.73.242) 68.336 ms !H * 68.530 ms !H vadim@alice:~$
For those who don't know, "!H" means "Host Unreachable"
Now that MS had the great idea of charging for security I'm sure that poor countries would be even more likely to switch to Linux. Who wants to pay first for the licenses, then for the support and then for additional security when with Linux you can get all that for free? Of course you can get paid support for Linux too, but as somebody mentioned here, often the community provides more than enough of it for free.
Why not configure Debian once and then use that everywhere? Set it up to use DHCP, install packages, edit/etc and then copy to all the computers you need.
Slashdot Mirror
Is that a really weird progress bar, or they really offer no indication at all of how much has installation progressed? Sometimes I don't have the whole day, so a good indication of how much has been done and how much is left is very valuable to me. An estimation of how long will it take would be nice too.
My apologies, then.
Now that I think about it, this is really cool. I would be able to avoid being tracked by switching to random IPs inside that 48 bits of address space.
That looks pretty silly to me too. Why would you want to have 2^68 addresses for, to give an IP number to every piece of dust in your house? The fact that there's a lot of address space available doesn't mean you have to waste it like that. I'm sure that 256 or 65536 addresses would have been enough. Hell, you've got more address space than MIT! They have a class A network (16777216 addresses).
I'd be careful before believing him/her. See the history, it seems to be just a clever troll.
I've got a Duron 850 with 768MB RAM, opening Konqueror takes about 3 seconds. Opening a new window for this reply took less than I can reliably measure though, maybe 0.5 seconds or so.
Well, I'm using KDE 3.0.4 on Debian and it works great. Stuff doesn't load immediately though, but what's the problem with that anyway? It's not like you're going to start a program just to close it in 5 seconds. When I open a link in Konqueror in a new window it appears immediately.
Just another VMware VM. They might collect how much I play it, but if there's nothing else on the disk they won't get much valuable information.
This is an interesting question, will Linux ever become public domain? I don't think so. Linux is not really Linus' anymore. It has lots of code from hundreds of people, and I think every of those people have copyright in their code. I'm pretty sure 2.4 is nothing like the first version of the kernel and there's nothing of the original code in it.
Compiling stuff is actually an advantage Linux has. If you don't want to you don't have to, just use any distribution. It's an advantage because it lets you try unfinished or unsupported stuff. When was the last time you could try an experimental file system or driver on Windows? I'm pretty sure you never did that. On Windows most of the time you can only try software when the developer lets you, and I've never heard of experimental Windows features being available with source code.
If you don't like compiling it's just fine, you can just use Linux just as if it was Windows and install only compiled software. You don't need to compile the kernel either, just use the one that comes with your distribution
I'm sure you could adapt quicky, just like I could have if I spent a few days with it. But are you sure that the first time you saw a Mac you could immediately find how to write a document, how to print or how to change the mouse's sensitivity? I'm sure it took you at least a few hours to figure where things are.
What about tastes, my favourite GUI is KDE 3, which curiously enough took less time to get used to than the Mac. But it took weeks to really learn all its features.
Heh, that's very funny. I remember I tried an iMac once. I can't do absolutely anything useful with it, and I've been using computers and Windows for years. I see lots of people talking about how intuitive is $GUI. That's a plain lie. Any GUI requires getting used to it. If a really intuitive one is ever made it will work by reading your mind.
It'd be tough, I'm sure. A boot virus would kill LILO or whatever is used, making it pretty useless. And even if it infected it I'm not sure it'd survive the switch to protected mode and kernel initialization. If the virus infected the kernel then I don't think it could be called "boot virus" anymore.
I wonder, what if a few hundred people did a complete mirror of the RIAA server every day? If somebody asked they could always say that they *love* the RIAA and are mirroring the site because it's so unstable lately and they don't want to miss anything ;-)
Heh, MS-DOS is a MS product, and the only OS with boot viruses I know. Try making a boot virus for Linux. I'm pretty sure it's not too easy.
Yeah, wonderful idea. What if I had to reformat the computer and need to get the full update again? Pay for another license?
For example. Unlike Windows, Linux usually works everywhere unless you have really different hardware. A Debian kernel should have everything compiled as modules, so it should be fine. If you use your own kernel make sure it supports all the hardware on all the computers you have, plus some common hardware you might get in the future just in case.
Things you could do are boot a computer with a Linux floppy, connect to the net, get the image from there and overwrite the disk. Or write the files to a CD and do the same. You could then use rsync to keep machines up to date.
Nah, it's just that lately switches are getting cheaper. My old 5 port hub sent packets everywhere so you could sit on any computer and sniff the whole network. With a switch you can only see the traffic from and to your computer (unless it's a big expensive switch that lets you sniff everything). That's pretty much the definition of a switch, it switches traffic instead of just sending it to every cable.
Since now small switches got pretty cheap I wouldn't be surprised there's some confusion. Besides, they both do the same function, just in slightly different ways.
Okay, how would you predict the time between my keystrokes? You couldn't get it right with the precision to a nanosecond, and can't predict when I'll take a pause to go to the bathroom or decide to scratch my nose.
I don't think you can predict random radio noise either, as it's a combination of many things: environment (cars, computers, microwave), the microphone, the cable, my computer, the time and location... if you record random noise anywhere else it'll be different
There are two problems with that.
First you have to know where in the encrypted data is whatever you want. XORing it with the victim's account number won't give you the location of that account in the ciphertext. In most cases it's probably complicated. I doubt banks exchange information in clearly delimited chunks. Over one TCP/IP connection they could pass thousands of transaction that can perfectly have a different length.
Second, that's trivially thwarted by the following protocol:
1. Bank A calculates all the data to send and its hash.
2. Bank A sends hash followed by the cyphertext
3. You do your substitution
4. Bank B checks the checksum and sees something changed.
You can't break this because you only know how to change the account number - you know it's say bytes 100 to 104. But you don't know the rest of the information so you can't replace the hash as well.
Nope. A random number sequence is simply a sequence that can't be predicted, and you can obtain one in many ways, like listening to radio noise, having a computer record from the microphone entry with no microphone connected (to record the noise introduced by ambient and electronics), or using data from the mouse, network, keyboard and disk access like the Linux kernel does.
I said *often*. If you need to be able to get support *now* and get an immediate fix, then obviously you need a support contract. But contrary to how many people think, there's not so much stuff that has to be available 24/7. For example if you're a small company that does business only during the day you could take the database server down at night and nobody would notice.
There are tons of places with information about security, btw, like bugtraq, and general information about securing Linux is available on the net and you can always ask people on IRC.
What about phone support you might get it free too with some luck. If your requirements aren't big (you just need some help to set it up) you might get lucky and find some Linux evangelist that would do some work for free just for the sake of removing MS software from a few computers. I helped with the servers at my school and nobody paid me for it. The teacher just asked nicely.
Again, I didn't, and am not saying that this is an option for everybody.
I have an ADSL connection with Telefonica:
vadim@alice:~$ traceroute batasuna.org
traceroute to batasuna.org (161.58.228.92), 30 hops max, 38 byte packets
1 router (192.168.0.1) 1.010 ms 0.816 ms 0.744 ms
2 10.3.50.1 (10.3.50.1) 66.757 ms 63.716 ms 63.745 ms
3 83.Red-80-58-11.pooles.rima-tde.net (80.58.11.83) 63.761 ms 60.499 ms 61.945 ms
4 37.Red-80-58-76.pooles.rima-tde.net (80.58.76.37) 66.936 ms 71.502 ms 63.596 ms
5 17.Red-80-58-72.pooles.rima-tde.net (80.58.72.17) 65.170 ms 70.477 ms 71.765 ms
6 242.Red-80-58-73.pooles.rima-tde.net (80.58.73.242) 68.336 ms !H * 68.530 ms !H
vadim@alice:~$
For those who don't know, "!H" means "Host Unreachable"
Now that MS had the great idea of charging for security I'm sure that poor countries would be even more likely to switch to Linux. Who wants to pay first for the licenses, then for the support and then for additional security when with Linux you can get all that for free? Of course you can get paid support for Linux too, but as somebody mentioned here, often the community provides more than enough of it for free.
Why not configure Debian once and then use that everywhere? Set it up to use DHCP, install packages, edit /etc and then copy to all the computers you need.