Jimmy's weblog

As mentioned in a previous blog, I have recently removed Microsoft Windows from my work desktop in favour of Apple’s OS X and Linux. During this move I had to transfer all my documents and email across to my new OS X machine which is my primary desktop.

Transferring my documents was pain free with one exception. I copied them all to a networked Windows 2000 server share and then copied them back via Apple File Protocol (AFP). Everything worked well except all file names were capitalised and severly truncated (to about ten characters if I remember correctly). Fortunately connecting via Server Message Block (SMB - Windows filesharing protocol) retained file name information.

The problems started when I tried to transfer my email. The University of Melbourne (where I work) recommends and supports Eudora as an email client but I had been using Microsoft Outlook XP instead. Other than silently hiding attachments to compensate for severe security problems, Outlook had worked well for me. Unfortunately Outlook wasn’t so keen to give me up as a user and attempted to use proprietary stands to lock me in.

Outlook uses a proprietary binary data storage format called PST. It is almost impossible to find a definition of the acronym PST, let alone find out anything about its format. Like most Microsoft programs, Outlook makes it relatively easy to import data from other formats but impossible to export to any useful format. Outlook also has the horrible habit of storing all your attachments within the data file needlessly inflating its size and increasing the likelihood of corruption.

“No problem” I thought. I’m fairly computer literate and work in the IT industry so I’ll search the Internet and find a script / application / method to allow me to convert the PST file to a standard format to import in to Entourage (Microsoft’s OS X email client). “How hard can it be?”

Short answer? Very hard until you think out of the box. A variety of scripts exist to export from PST however I could not find one that would work reliably. However despite Microsoft’s attempt at lock-in, Outlook must be able to speak the standard email protocols of POP3 and IMAP (and SMTP though that isn’t relevant in this situation).

POP3 is no use however IMAP certainly is. POP3 allows your client to download messages from a mail server but it doesn’t allow you to publish anything back or create folders on the server. IMAP stores your messages on the server and allows you to create folders and move messages between them. Outlook supports IMAP so I was able to copy my folders and messages to the server and then download them when I had made the switch to OS X

I had planned to switch to using Entourage (also from Microsoft) but after my PST experience I decided to do a little research before I commited to a new client. It turns out Entourage also uses a Microsoft data format just as Outlook does (though not the same format which would have made my transition much easier).

I settled on a combination of Apple’s Mail, Address Book and iCal for my needs. Lest you think that I have simply “drunk the cool-aid” as many Apple fans do, let me emphasise that my decision was made on the basis of open standards. While the Address Book does use a binary format for storage, it makes it very easy to export all addresses as a single VCard (RFC 2739). Mail uses the standard Mbox mail format (Postmark followed by email formatted according to RFC 822) and iCal uses iCalendar (RFC 2445).

Working in a largely Microsoft server environment (though generally in an open source role) I have always been aware of vendor lock-in on the server side and never really considered it on the client side. This experience has made me think far more about the clients I use for my work.

After moving to open standards with my communication and scheduling tools, my single point of reliance on proprietary standards is now Microsoft Word and I plan to move to a combination of Open Office with its XML format and other forms of XML publishing.