Its true about exploits, even though the distributions test their OS as a "whole", because the packages are from different locations, its possible that somewhere, the code written is not up to the mark at a perticular time. This is true even for Linux (the kernel), because of which some debian systems were compromised (even though debian is known to be stable). In my earlier post, I was simply saying that because Linux Systems (note, systems) follow the *nix file system heirarchy and permissions, a virus cannot just "execute and destroy" the system. Simply because the system is not writable. Its a different thing when you're logged in as root, but then why would you log in as root just for doing "user-stuff"? Having a firewall is about as normal as anything to people like me, why do you need those symantec guys anyway? "The debate goes on forever about the *nix virus. Is it because it's "better" than windows or is it just because the hackers/crackers/script kiddies have not yet targeted the *nix systems." The latter may be true to some extent, but it is also a lot more difficult to do something like that on a *nix system. The user's home directory is writable to any script/program he runs (quite obvous) and so the only danger according to me, from "scripts" is that you can loose data on your home dir. But its not like windows where the whole system may be left unbootable. Show me a record/link/anything where a Linux system was compromised because of a "virus", which was a stand-alone executable and did the damage. I've never heard of such a thing. On micro* products however, you hear it everywhere, all the time.

0

Audiophile, I'm sure you didn't mean this as an insult: "gmoney, you should first find out more about how a linux system actually works, and then give your "proofs". ...but I've been familar with the concept of UNIX file permissions and security since I worked on SPARC systems around 1995 and I totally agree with you that even if viruses were common in Linux it would be difficult to cause system-wide damage because a root exploit would be necessary. I guess I'm only trying to say with my "proofs" that it's incorrect to say they are an *impossibility* but rather an improbability. And to be fair to MS Windows, in the 9x/ME versions, everyone user has root-level priveldges but in NT/2000/XP/2003 things are getting a little better because you can set up limited users which can't install programs or alter critical things in the OS. The problem is by default Windows XP Home sets up new users as Admins/Superusers. Maybe you've heard of the Linux distribution called Lindows in which the default priveldges are root. Unfortunately, for convenience, they are ignoring some of the great security aspects of Linux and UNIX systems that you described . One person's commentary on this is here. More sophisticated Linux users probably wouldn't use Lindows or if they did, they would choose the more secure options during the install but the concern about Lindows' policy is if Lindows becomes more mainstream because a certain large retailers' online store pre-loads Lindows on some of their PCs.

0

Sorry if I was rude, but I really don't understand why anyone should search on symantec for "linux" viruses, or install their patches. The firewall is handles by iptables, the distribution is from your distributor, theres no need for symantec.

0

>theres no need for symantec. I agree (in non-Windows environments of course ;) I only searched symantec as a sort of "Google" for viruses/worms/trojans - to only find information if Linux viruses had been documented at all, not to endose their products. They sometimes have good descriptions of Windows viruses (for example, when a company I worked for got hit by the W32.Blaster worm this summer and they initially thought the problems were caused by downtown messing up their Active Directory rollout ;) Rick mentioned installing their [symantec?] patches... I don't think he means to patch Linux security holes but maybe symantec has products that run on Linux file and e-mail servers that scan for Windows viruses to prevent/cleanup infections of a company's Windows clients?

0

Just a little technical response to your real world example, gmoney,.... The Blaster worm was based on a MS specific vulnerability (MS-RPC-DCOM Buffer Overflow), and most likely it was an MS patch that needed to be loaded. To the rest of the group, I have been working with *nix systems exclusively for the last 3+ years. Once product that my organization used was called vFind. The AV software RAN on UNIX/Linux systems but mostly held definitions for Windows viruses. In my humble opinion, I think the biggest thing to be concerned about is host and network security. If you are employing good security practices with your *nix system(s), there is most likely no need to worry about an antivirus product for Linux. The ironic thing is that most hackers probably use some form of *nix to do their hacking! But nothing has really come out against them. Maybe even hackers have respect..... ;)

0