Raw sockets support is not something that you turn on or turn off. It's
part of the TCP/IP stack. I mean I guess they could have shipped a
crippled TCP/IP stack with Windows 2000 Pro and left raw socket support
in Win2k Server, but then you're talking about maintaining two codebases
for a problem that just simply is not a big deal.
I understand the argument that Gibson was trying to make. All I'm saying
is that it's stupid to bash Windows for having a feature that's part of
TCP/IP.
I don't think that your average Linux user would know enough to use or
need raw socket support either.
Brian Weeden wrote:
Raw sockets itself is not a problem. For a given Unix system the
majority of people who have the power to use raw sockets know how to
do it properly.
Gibson's beef was that you now have millions of copies of Windows 2000
with raw sockets on by default and every single copy could be infected
by malware/viruses that abuse raw sockets. He argued that there was
no good reason to include them in Windows 2000 because no average
windows user knew enough to use them properly or needed them. And it
left a huge hole that could be exploited. Which it was by every worm
and virus that rampaged the net in the last few years.
Doesn't anyone actually listen to what people say before they spout off?
