> > Dave Wreski
>
> Nice conversation you've been listening to! If it would be impossible to
> route a class C address over the internet than we would neither be able
> to read eachothers e-mail nor to surf the www, since many ISPs give to
> their clients IP-addresses out of a pool of a class C net.
uhm.. not entirely true. it depends on what you mean by routable. if he
meant "announcable via bgp to the routers of the net" then no, most class
c's are not. they are announced as part of their providers network. read
up on cidr, bgp, ASes, and big-ass internet routing in general.
> Class A means that the first digits of an IP-adress determins the
nope. class A means that the first bit is 0 (1.foo -> 127.foo). though i
think the iana has a huge chunk of that reserved.
> Class B means that the first 2 digits of an IP-adress determins the
> network address: nnn.nnn.xxx.xxx. So you can have 254*255 different
nope again. class b is when the first 2 bits are 10 (128.foo -> 191.foo)
> Class C means that the first 3 digits of an IP-adress determins the
now this one i'm not real sure of.. isn't it the first 3 bits are 110?
don't the multicast ("class d") start at 224 (111). sounds right.
but all of this "class" garbage is depricated. people need to get in the
cidr mindset wherein a network is descirbed as addy/sigbits where addy is
the address and sigbits are the number of bits that make up the network
portion. as an example, what you called a "class a" is really a /8
because the first 8 bits are the network. this also makes subnetting make
more sense when you're not thinking in the class mindset.
i'm willing to clarify if people actually care..off the list if possible,
unless it relates to linux somehow :)
-- zach