I'll respond with a question - why is that we can produce 64 different codon combinations, but only evolutions only has us code 20 amino acids?
I have a professor who theorized that our current genetic code evolved from another 3 base code which was palindromic. AAC and CAA would produce the same amino acid. Additionally, their complements would do the same. AAC, CAA, TTG, GTT would all be the same amino acid. This would allow you to read the DNA backwards and forwards and get the same amino acid structure (perhaps backwards, but there's evidence that it could be worked around).
If you accept a palindromic 3-base code, you'll get:

Which can represent exactly 20 different amino acids, using all 64 combinations. It's a little too pretty to be a coincidence.
I don't know how easy it would be to find on the internet, but this is all listed in this paper: Beland P. and T. F. H. Allen. 1994. The origin and evolution of the genetic code. Journal of Theoretical Biology 170:359-365.
They attempted to map what amino acids the old codons represented and there are some strikingly interesting graphs. I wish I had pictures I could link to, but the gist of it was that the 64 codons could be put on a graph where codons that produced the same amino acid are placed next to each other in a grid. The "old code" had vertical matches, where our current code has horizontal matches. Supposedly, at some point, there was a switch over in representation.
One thing to note is that this potential "old code" didn't have stop codes like we do now. It's possible that the stop codes evolved to break anything that attempted to use the old code in favor of the new interpretation.

If your DSL connection is PPPoE and not DHCP, your modem will occasionally get logged out and need to reauthenticate to get back online. This is probably the "redialing" you're talking about.