William said:
On 12/28/2006 10:10 AM, something possessed Alun Jones to write:
in message Not necessarily. If you're just using it to illegally download
music and videos (not program executables), and you're careful
about how you play these (I wouldn't rely on Windows to launch
them, for example, but load them in Winamp, and don't let Winamp
connect to Internet), than you're more or less safe.
Right, because Winamp has never had any vulnerabilities that can
be exploited by badly formatted data.
I didn't recommend Winamp because it was invulnerable, but simply
because its not Integrated into the OS, so that if it goes bad, the
whole OS doesn't suffer.
It does if you're running as an administrator account.
Windows (as with all computer systems I'm aware of) cannot
distinguish between the user, and programs run on that user's behalf.
If you, the user, run Winamp, and it loads a data file that causes
execution through exploiting a buffer overflow, the malware inside of
that data file can do absolutely anything to the system that you can
do, with the exception of anything that requires your actual physical
presence.
So, if you're running as an administrator, it doesn't matter if
you're loading exploits into a program that's labeled "part of the
OS", or one that's labeled "third party shovelware", the exploit can
do what it chooses.
The answer, then, is to run as a restricted user account. I do it
all the time - and when I do, my Internet Explorer runs as a
restricted user account too. Exploits in the apps I use can still do
anything I can do, but the damage is limited to my personal data, not
the entire OS.
You can even run as an administrator while forcing IE to run as a
restricted user! [Search for "SAFER" and "SRP" and "Internet
Explorer" for some articles, or see
http://blogs.msdn.com/michael_howard/archive/2005/01/31/363985.aspx]
Note, though, that once you've downloaded and run a piece of malware,
whether it's an EXE or a buffer-overflowing MP3, that malware can do
everything you can do as a user.
Additionally, while exploits may exist in Winamp when accessing
questionable media locally stored, In order for any real damage to
be done (i.e. a trojan downloader), Winamp would need to access the
Internet (or maybe that reched program Internet Explorer).
Uh... no. Remember, Winamp - and any exploit it loads, as far as the
operating system is concerned, _is_ you.
It can start another program, it can inject itself into another
program you're already running, or it can combine the two.
A good Firewall (like Kerio) should be able to prevent this from
happening.
No, no it won't. Again, if you've told Kerio, or whatever, to allow
_any_ program to access the outside world, that program can be
compromised by code you've run under any other program. So, your
Winamp exploit can infect your Internet Explorer in memory (not on
disk, unless you have rights to that), and pretend to be Internet
Explorer in order to download its exploit - or, quite honestly, it
can simply start up IE to fetch the rest of its code.
But why would it need to do even that?
How big are the media files you're "sharing"? Way bigger than most
damaging code I could imagine. If you're downloading a video, or
anything more than a few seconds of sound, you won't notice the
increase in size that you get by adding some kind of malware.
Oh. No, wait, actually it has. Several times.
This is why the trend lately is to attack applications, rather
than operating systems - the operating system vendors are getting
much better at tracking and fixing problems, but many application
vendors still have their heads in the sand - and so do many users
[snip]
I understand your sentiment. Clearly, you support Microsoft, and
that's fine. I don't agree with that sentiment, but everyone is
entitled to their own opinion.
A => Z. Welcome to today's edition of "Jumping to Conclusions".
In the abstract sense, there is no dividing line between code and
data - data tells code where to go, and so acts as pseudo-code, in
many cases.
Again, requires Internet access to download the trojan. Media
itself cannot contain the final executable code that infests a
system with malware, all it can do is exploit vulnerabilities that
allow the said malware to be installed.
If you believe that, you've got a long way to go. There really is no
other way to say it, but to note that you are completely wrong in
that assertion. Media itself can quite comfortably contain the
exploit and whatever code is going to execute after the exploit has
taken over control of your system.
Also, more than likely, the P2P proggie you used had its own
malware (like Navaccel or something like that).
Don't make the mistake of assuming that I'm talking about my own
experiences with P2P - I've simply seen too many machines infected
where the source of infection is traced to an overactive P2P
exchanger.
Which is why if someone is going to use P2P, they should be advised
(as I'm trying to do) on how to use it safely. I'm not condoning
such action, but its kind of analogous to making sure your teenager
has protection, you don't want them to have to use it until they've
matured, but they do, than it'll be there for them.
Best protection against catching malware from P2P is a membership at
Blockbuster, or a Netflix subscription.
Get your movies, and your tunes, from reputable sources who have a
little skin in the game should you get infected through them.
Finally, some P2P proggies (such as Bittorrent) can be used
safely (like for downloading Linux distros), since even though
you're downloading from other computers, the tracker is
administered by the Linux Distribution and, to my knowledge, it's
not possible yet to alter a file or set of files once the tracker
has already been posted without posting a new torrent tracker.
I'm glad you put me at ease there - after all, the main Linux
distros have never been altered maliciously by hackers.
Oh, wait, they have, haven't they.
http://www.linuxinsider.com/story/32240.html
What's this got to do with altering a bittorrent stream. The
results would have been the same rather bittorrent was used to
download the distro or if it was downloaded from the server. In
fact, in this case, the bittorrent tracker probably would have been
the safer bet, since it was the server (and not the torrent) that
was hacked.
The point is that you can only trust checksummed streams as much as
you can trust the person who created the file and the checksum in the
first place. Since most "sharing" of illegally copied material is
done by people who would like to remain anonymous, you're relying on
trusting someone whom you can't identify, and whose reputation (and
reason for maintaining that reputation) is unverifiable.
Cleaning a virus or trojan infection is only going to be effective
if you can plug whatever hole they got in through - whether it's a
hole in your behaviour, or in your apps, or in your OS. Even
flattening and restoring just means that the attacker gets another
chance to try the same thing at you, but this time on a system
that's less cluttered with the debris of other previous attacks.
Agrees with you here. So, with that, I hope that if the OP
ultimately decides to continue P2P, that he/she does so safely.
That requires only loading files with hashes generated by trusted
authorities. ("Authority" here means anyone with the right to say
what is, or isn't, a valid copy of a file.)
Downloading stolen movies and songs is not going to be safe. Not
ever.
Alun.
~~~~