I started out programming systems under DOS (DBase, TAS, etc.) and then later under Windows (C, C++, Delphi, Visual nasty Basic). I only started developing under Linux when I heard about it 3-4 years ago.
Until then, I'd never heard of Linux. I knew about the existence of UNIX in general, but the small firms I worked for couldn't afford it and then when I worked for larger software houses, I was pigeon-holed into the Windows development area so got no exposure to UNIX. I just recall bigoted sysadmins complaining about these "Windows toys".
The first time I used Linux was installing Mandrake 6 (or something like that) from a magazine cover CD, where I immediately fell in love with the wonderful power of the command line again, along with the marvelous tenet of programs that "do one job, and do it well".
So yes, I fall into the ex-Windows (badoom-tisch!) bracket - how many other Slashdotters do?
One example that they have is "Text adventure". Well, obviously, people are not interested text games any more but it doesn't mean that the niche is gone, it has just been filled with games with more technical capabilities but which still satisfy the same needs and appeal to the same types of people.
There is still a very much alive and kicking text adventure scene. Even the Infocom Z machine is still around and you can write your own games for it (I even did a multiplayer IF project myself a while back).
You can use the keyword filters to fairly effectively reduce spam on evolution.
By looking for subjects starting with "FREE" I managed to get rid of about 80% - further checks for "viagra", "penis" and "teens" pretty much got rid of the rest.
Occasionally I get an odd one, like "chicks and donkeys" or something but I just add new rules as they come. Generally one gets through about once a month.
Err... I think the routers have quite enough to do shunting mind boggling numbers of packets around without trying to analyse packets containing SMTP protocol email messages - the amount of processing power required (on top of what the backbone already suffers) would be absolutely phenomenal.
It is theoretically possible (aside from the fact that most routers are dumb hardware devices dedicated to shuffling packets around and unable to analyse content which means hardware changes), but given how many data packets flow through these bottlenecks, the net would be a much slower experience.
Oh, and who do you contact when legitimate mail gets dropped as spam?
What I meant to say is that the government has not requested any help from ISPs in this particular initiative - probably because it would require some serious monitoring and storage from the ISP end that (most ISPs) are simply not in a position to do.
I didn't mean to imply at all that UK ISPs wouldn't help the police!
I too have had to track down users from their IPs to report DoS activities and failed attempts to sabotage my web and mail server.
Err.. If you spoof the IP, how can peers download off you since they don't know where you are?
In theory you could come up with some kind of encoding scheme for the IP (spoof the IP on the packet header to some rubbish) and encrypt the real IP in the body of the packet somewhere.
Problem is that manipulation of packet headers is not particularly easy since socket libraries generally don't allow access to the header.
Oh, and you could STILL just use a simple packet monitor to watch outbound packets on your connected machine to get the real IP! If you can do it, so can anyone else...
People can still connect to your IRC server using a standard SSH transport the same way that you are - no matter where it is in the world.
All IRC servers allow you to easily access the IP addresses of the people you're talking to (It's part of the protocol - WHOIS). Once they have the IP, they can look up your ISP from the netblock and lean on them to reveal who you are!
Also, whilst you encrypt your normal IRC traffic, DCCs will not be similarly encrypted, since they are simply a TCP socket opened between the two machines with the raw data of the file dumped down.
You'd be better using your encrypted IRC in conjunction perhaps with FTP over SSH. Got to make sure then that you only give out passwords to people you know - but who can you trust?
Something like Freenet may work a little better (http://freenetproject.org). Ok, your ISP could probably tell that some kind of encrypted traffic is going on, but that's it (I believe Freenet uses a standard 128-bit cipher).
I think there may be some flaws still in this due to TCP itself - any decent packet monitoring program could still determine the destination IP address on packets going to particular nodes, so you could still discover who was hosting a freenet site (ie. government agencies/RIAA etc. could simply join freenet as other people do with a packet sniffer on their machines, find a site they don't like, get the IP, etc. etc.)
That's really what it comes down to - if you are hosting something for others, all it takes is for someone to get your IP and they can find out who you are (eventually).
I believe ISPs should not be allowed to reveal who its customers are to third parties. Perhaps some kind of privacy law should be passed to deliberately prohibit ISPs from keeping email/traffic logs (although this requires so much hardware that it is already expensive to do this - the UK government tried to get ISPs to keep email logs, but they simply can't afford it), and more importantly if they issue dynamic IPs, no logs of who held which IP at what time.
Here in the UK, a government crackdown is being run on people just accessing child pornography (a heinous crime indeed, don't get me wrong). However, they have no co-operation whatsoever from the ISPs - the only way they are tracing these people is by getting lists from credit card companies of people who paid money to known child porn sites.
Anyway, bottom line - nobody is safe. Be afraid! Very afraid!
Slashdot Mirror
Call me simple if you will, but I thought the standard *NIX type fs had this ACL thingy called a GROUP which you made USERS a member of!
Say your group is "financial", what's wrong with
a simple:
chown root:financial [resource]
chmod 660 [resource]
Besides, NTFS ACLs are a stupid way of controlling access to resources - You can also lock out the Administrator (read root) user from a resource.
I started out programming systems under DOS (DBase, TAS, etc.) and then later under Windows (C, C++, Delphi, Visual nasty Basic). I only started developing under Linux when I heard about it 3-4 years ago.
Until then, I'd never heard of Linux. I knew about the existence of UNIX in general, but the small firms I worked for couldn't afford it and then when I worked for larger software houses, I was pigeon-holed into the Windows development area so got no exposure to UNIX. I just recall bigoted sysadmins complaining about these "Windows toys".
The first time I used Linux was installing Mandrake 6 (or something like that) from a magazine cover CD, where I immediately fell in love with the wonderful power of the command line again, along with the marvelous tenet of programs that "do one job, and do it well".
So yes, I fall into the ex-Windows (badoom-tisch!) bracket - how many other Slashdotters do?
Check out:
rec.arts.int-fiction rec.games.int-fiction Brass Lantern
Odd, I was playing Return to Castle Wolfenstein through WineX this morning on my box, which has been running 4.2.99_902 for a few weeks now.
Haven't had any problems with anything - and I had no trouble recompiling the Nvidia drivers.
Could you just be unlucky?
Heh!
You can use the keyword filters to fairly effectively reduce spam on evolution. By looking for subjects starting with "FREE" I managed to get rid of about 80% - further checks for "viagra", "penis" and "teens" pretty much got rid of the rest. Occasionally I get an odd one, like "chicks and donkeys" or something but I just add new rules as they come. Generally one gets through about once a month.
Err... I think the routers have quite enough to do shunting mind boggling numbers of packets around without trying to analyse packets containing SMTP protocol email messages - the amount of processing power required (on top of what the backbone already suffers) would be absolutely phenomenal.
It is theoretically possible (aside from the fact that most routers are dumb hardware devices dedicated to shuffling packets around and unable to analyse content which means hardware changes), but given how many data packets flow through these bottlenecks, the net would be a much slower experience.
Oh, and who do you contact when legitimate mail gets dropped as spam?
Sorry, I worded that badly.
What I meant to say is that the government has not requested any help from ISPs in this particular initiative - probably because it would require some serious monitoring and storage from the ISP end that (most ISPs) are simply not in a position to do.
I didn't mean to imply at all that UK ISPs wouldn't help the police!
I too have had to track down users from their IPs to report DoS activities and failed attempts to sabotage my web and mail server.
Err.. If you spoof the IP, how can peers download off you since they don't know where you are?
In theory you could come up with some kind of encoding scheme for the IP (spoof the IP on the packet header to some rubbish) and encrypt the real IP in the body of the packet somewhere.
Problem is that manipulation of packet headers is not particularly easy since socket libraries generally don't allow access to the header.
Oh, and you could STILL just use a simple packet monitor to watch outbound packets on your connected machine to get the real IP! If you can do it, so can anyone else...
And what's the point in that?
People can still connect to your IRC server using a standard SSH transport the same way that you are - no matter where it is in the world.
All IRC servers allow you to easily access the IP addresses of the people you're talking to (It's part of the protocol - WHOIS). Once they have the IP, they can look up your ISP from the netblock and lean on them to reveal who you are!
Also, whilst you encrypt your normal IRC traffic, DCCs will not be similarly encrypted, since they are simply a TCP socket opened between the two machines with the raw data of the file dumped down.
You'd be better using your encrypted IRC in conjunction perhaps with FTP over SSH. Got to make sure then that you only give out passwords to people you know - but who can you trust?
Something like Freenet may work a little better (http://freenetproject.org). Ok, your ISP could probably tell that some kind of encrypted traffic is going on, but that's it (I believe Freenet uses a standard 128-bit cipher).
I think there may be some flaws still in this due to TCP itself - any decent packet monitoring program could still determine the destination IP address on packets going to particular nodes, so you could still discover who was hosting a freenet site (ie. government agencies/RIAA etc. could simply join freenet as other people do with a packet sniffer on their machines, find a site they don't like, get the IP, etc. etc.)
That's really what it comes down to - if you are hosting something for others, all it takes is for someone to get your IP and they can find out who you are (eventually).
I believe ISPs should not be allowed to reveal who its customers are to third parties. Perhaps some kind of privacy law should be passed to deliberately prohibit ISPs from keeping email/traffic logs (although this requires so much hardware that it is already expensive to do this - the UK government tried to get ISPs to keep email logs, but they simply can't afford it), and more importantly if they issue dynamic IPs, no logs of who held which IP at what time.
Here in the UK, a government crackdown is being run on people just accessing child pornography (a heinous crime indeed, don't get me wrong). However, they have no co-operation whatsoever from the ISPs - the only way they are tracing these people is by getting lists from credit card companies of people who paid money to known child porn sites.
Anyway, bottom line - nobody is safe. Be afraid! Very afraid!