If it was anything other than a basic yahoo account (webmail only, no POP or IMAP) I could have used the mail-notification package

#!/bin/bash

#requires packages libnotify-bin, fetchyahoo

# include settings from $HOME/.yahoomail, format as shown below
# USERNAME='my-yahoo-email-account'
# PASSWORD='my-password'
. $HOME/.yahoomail

MESSAGES=$(/usr/bin/fetchyahoo --quiet --username="$USERNAME" --password="$PASSWORD" --spoolfile=/dev/null --newonly --onlylistmessages)
COUNT=$(echo "$MESSAGES" | wc -l)

if [[ $COUNT -gt 0 ]]
then
    /usr/bin/notify-send "New Email" "You have $COUNT new messages"
fi

exit 0

I added it to the System->Preferences->Session and now when my dad logs in he doesn’t have to check for emails, the emails will check on their own.

I finally found some documentation of the SBF file format. It has made the work a lot easier.

I had realised before that the data was split into multiples of 128B blocks but I didn’t realise that this block size was specified in the document info at the beginning of the file (i had no idea what that data was previously). My previous attempts at parsing relied on 0×80 appearing at the start of new records, unfortunately it also sometimes appeared as part of the block data so any attempt to split the data that way resulted in some records being cut in half when they shouldn’t have been.

The new script reads the first 60 bytes of the file to get the file info (such as block size) and then reads the file in block by block. The first 4 bytes of each block (the rest is data) are a pointer to the next block in the record with some bit flags showing whether the block is the first block of a record (this was what I saw as 0×80).

After struggling with trying to extract data from an unknown binary file format I now realise how important open and well documented file formats are.

I went to several presentations:

• Packaging Perl Applications

  Dealing with issues surrounding packaging Perl code as a standalone unit.

• The Goo

  A productivityapplication someone built for himself, aimed to be thought process friendly

  Brought up a lot of interesting points about interface design and thought processes

• Accessing databases with Perl

  Had some cool stuff about some Perl modules that would pick up the structure of your database and create objects to perform common relational operations on them.

• MighTyV

  A walkthrough of the developmental process for the creation of a website to show TV schedules. It was one of the best presented talks I have seen, you could tell that the presenters were very passionate about the subject.

• 3D programming with Perl

  This talk really surprised me, the amount of code required to do OpenGL stuff in Perl is tiny. A good intro for beginners to simple OpenGL stuff.

• Programming Red flags

  This was a walkthrough on how to improve your coding style. It was very interesting and made you think about structural and readability issues when coding a project. It was part of a larger talk but had to be cut down to fit in the time slot. I would like to see the full length version of that talk. I think it should be mandatory for all programmers to see this presentation.

I’m also adding visitor logging so I can see how popular it is.