Last time we processed the Digsby, Trillian and Pidgin logs and saved them as a unified YAML file for each contact.
Now let's start seeing who I talk with most. A naive way to do this would be to just simply sort by file size the YAML files. A small problem is that some contacts have longer IM names (such as thebestcatalin), other have shorter names (such as b0gdiy). That's a difference of 8 characters, so 8 bytes, which over say 100.000 lines exchanged become 0.76 Megabytes. A 100.000 line file has about 2 Mb, so it would be an error of 38%.
And there are a few other reasons to …continue.