One of these days I need to write a presentation. I once started a CGI one (after all, I wrote a book on it, so should be able to do an talk on it!), but never finished. I wrote one (and gave one) on all the bits of an email, especially header info and ecodings. I didn't have much Perl in it, just more of a "How to look at raw email and understand it, so you understand how to parse it or what a parser is doing when it parses it". I also gave one on I18N and G11N, in relation to emails (this was part 2 to the headers) and the web. Covered a lot of dealing with encodings, double byte characters, UTF-8, etc... While I was working at Verio I learned a lot about email structures, encodings, charsets, etc... since I had to write a "globalized" web based email system. I wrote some modules which do various header parsing/creation which works nice with non-ascii charsets (concentrated on iso-2022-jp, Shift_JIS, iso-8859-1 and UTF-8... but uses iconv for whatever conversion is needed). Why not Encode.pm? Wasn't considered stable at the time (and we were using 5.6.0). And the existing email parsing modules didn't seem to work right, or do what I needed as far as getting a folded iso-2022-jp subject converted to UTF-8 for web display.
Maybe I should revisit the talks and add some Perl-fu into them. YAPC will be here before I know it! They are interesting on their own, but would need more Perl... or maybe not.