Computer Security 101 for the Small / Casual User, Part 1.
Last month, we discussed how and why "Security" is indeed an important
issue for ALL computer users, even those who "only" casually use E-Mail
and/or surf the 'net from their stand-alone home PCs. We closed
by pointing out that it is the
responsibility of everyone who connects (even indirectly) to the
Internet — yes, this means
YOU
— to ensure
that their systems are adequately protected against such attacks and
illicit use. Hopefully, that article gave you enough
understanding of the stakes and consequences involved to make you want to do
something about it.
So the obvious (if rhetorical) question becomes,
how
do you “ensure
that your systems are adequately protected against
such attacks and
illicit use”?
The usual generic advice that you've heard a thousand times before is
to run a good antivirus program
[1] and keep
its “definitions”
or “signatures” files
up-to-date. That's valid, as far as it goes; but it
doesn't go nearly far enough. The same holds true for
popular “spyware” scanning programs like
Ad-Aware and
Spybot Search & Destroy. These
tools can be very useful when they're needed; in fact, we use
and recommend both of the ones just mentioned. But they, like
on-demand virus scanners, are fundamentally after-the-fact
“fix-it kits”, not
preemptive protection. The fitting analogy here is to the
spare tire in your car. You definitely want to have it
available if/when you need it; but if all is right with the
world, that day will
never come. Conversely, if you're
repeatedly relying on it to get you home, you've got some other much
larger and more fundamental problem — and
that is what you need to
fix. Nobody is supposed to get three flat tires per week!
In other words, the goal should be to prevent the problem, not just mop up after it.
The best — and only truly
effective — way to accomplish this to eliminate the
attack vectors (i.e., the routes into your computer that the malware
takes), so that you cannot be infected; and hence, you cannot
subsequently infect others.. To do that, the first thing you
need to understand is that
ALL so-called
"security vulnerabilities" are created by bad software
— yes, every single one of them. And to at least some
degree, all non-trivial software is buggy — i.e., bad.
There is an old
adage which applies here:
| The only truly secure computer is one which is powered-down,
disconnected from all other computers, and locked away where no one can
physically get to it. |
Unfortunately, that also makes for a pretty useless computer. But
it nicely points up the fundamental approach that is needed, which is
two-fold:
The first step is to realize that no matter how diligently we patch,
update, tweak, and otherwise attempt to remove vulnerabilities from a
computer system, we can never be 100% confident of our success in that
pursuit. Even if we were to do a perfect job and create the
mythical “perfectly secure computer” with all known
vulnerabilities excised today, a new vulnerability can (and probably
will) be discovered tomorrow. So an additional line of defense is
required, in the form of a properly configured stand-alone outboard
firewall (often called a “hardware” firewall, although that
term is actually something of a misnomer). Suitable devices can
either be purchased “off the shelf”
[2],
usually at very modest cost; or they may be built from a spare obsolete
PC using any of several special-purpose free software packages designed
expressly for this purpose, such as
Freesco,
IPCop, or
SmoothWall Express.
We generally prefer the latter approach, as it offers greater
flexibility, upgradability and control, and at least potentially better
protection; but either can be made to suffice.
Do
not be
fooled by popular marketing myths into thinking that a so-called
“software firewall” or “personal firewall program” can
ever be an adequate substitute for a proper outboard firewall.
All such products are at least mostly snake oil, aimed squarely
at the naive user
[3];
and it can be fairly said that they are actually detrimental to most of
their users, due to the false sense of security they inspire.
They aren't even really firewalls; as can be easily realized by
considering the meaning of the term “firewall”. Much
like a moat must completely surround the castle in order to be
effective, by
definition a “firewall” is supposed to be an impenetrable
barrier standing
between
the threat and the object you're trying to protect. But these
so-called “software firewall” or “personal
firewall” products are merely application programs running on the
same machine they are putatively trying to “protect”.
Hence, their
fundamental concept is inherently flawed, because (at least part of)
the target machine remains directly connected (and thus exposed) to the
general Internet. In effect, they are swimming pools inside the castle. Sure, a
particularly clumsy Barbarian might slip and fall into it while on his
rampage through the Royal abode; but by then, he and his cohorts have already smashed the
furniture, pillaged the pantry, and done unspeakable acts to the
cooks and chambermaids. Further, because these are simply application
programs running on your general-purpose PC, it only takes one mouse
click on the wrong link on the wrong web page to instantly download a
Trojan Horse program which will completely disable that “software
firewall”, and BANG — you are “own3d”, as all
of your illusory “protection” has been nullified. By
comparison, while an outboard firewall cannot be counted on to
prevent such a user-initiated download, it also cannot be affected by
it, since it is
not an application program running on the target machine
With our properly configured outboard firewall in place, next month we
will address step two, which is securing (to the extent possible) the
target
PC itself.
Footnotes:
[1] This task is not as simple as it may at first seem. Many of
the “biggest name” anti-virus programs are actually among the poorest
possible choices. For example, current Symantec products, including
Norton AntiVirus,
come loaded with a DRM Trojan (and a
badly implemented one, at that); hence it can be reasonably said that they constitute malware in and of themselves. NAI/McAfee has a
long dismal history of spamming; which means that they are by definition untrustworthy and
patronizing them would be unethical.
Others, such as Avast, F-Prot for Windows, and PC-Cillin, remove
themselves from possible consideration by virtue of the fact that they
(at least claim) dependency on Microsoft Internet Explorer.
Still others insist on “phoning
home” (putatively for “automatic updates”; but you have no way
to ever be sure that the information exchange is limited to that)
without
your express permission or approval and/or require so-called "online
product activation" (both of which are security problems in and of
themselves). Fact is, we have yet to find an anti-virus tool that we
really
like in all respects; hence, we are loathe to make any specific
recommendations. But after eliminating those packages ruled out by the
foregoing issues, the following are at least potentially worth
consideration:
AVG Anti-Virus Free
AVG Anti-Virus Professional Edition
ClamAV ,
ClamAV for Windows, and
ClamWin Free (which are all related, but different)
Sophos Anti-Virus
There may be other worthy candidates; but these currently seem to be the best bets.
[2] For the home or very small office user, entry-level models such as the D-Link
DI-804HV, Linksys
BEFSX41v2, and Netgear
RP614,
will usually prove sufficient.; if WiFi (IEEE 802.11a/b/g) wireless
LAN connectivity is needed, such as for a roaming laptop, then models
such as the D-Link
DI-824VUP, Linksys
WRT54GL and Netgear
WGU624 would
be more appropriate. Please note that these references and links are
provided solely as examples for your convenience, and do not constitute
a specific recommendation; there may be other makes/models which offer
equal or superior performance and/or value.
[3] For confirmation and further background information on this, see
this article by
Steve Atkins of SamSpade.Org and Word to the Wise.