It's pretty well spelled out in the pack documentation. C for unsigned char (byte), n for unsigned 16-bit short in network order, N for 32-bit long in network order, a for "arbitrary binary data".

Don't be confused by the 32-bit CCCC's, since an IP address is 4 8-bit numbers, not a 32-bit number.

Don't be confused by the 32-bit CCCC's, since an IP address is 4 8-bit numbers, not a 32-bit number.

It's actually 32 opaque bits. The dotted form is just a representation. The separation into bytes in completely arbitrary. But if you're going to display in that form, it might be useful to extract the IP as 4 bytes.

an IP address is 4 8-bit numbers, not a 32-bit number

s/is/is normally represented as/;

$ host perlmonks.org
perlmonks.org has address 66.39.54.27
$ perl -le 'print ((66 * 256**3)+(39 * 256**2)+(54 * 256)+27);'
1109866011
$ perl -le 'print unpack('N',(gethostbyname("perlmonks.org"))[4]);'
1109866011
[download]

Many web browsers accept the decimal value of the IP address: http://1109866011/

s/is normally represented as/is encoded in an IP header as/; # :-)

Now it's making more sense... Thanks for the help, gmargo!