An interesting issue came my way last week. One of the problems with a large product like Enterprise Vault is the sheer number of nooks and crannies that customers, and indeed engineers, might not know about.
One of these is the view of customised shortcuts in Outlook AutoPreview. This is the couple of lines of the body that may or may not be visible in the message list, depending on your view configuration.
It's not something I use as I tend to simply read everything in the reading pane, and I find it clutters the list too much for my liking. Actually, I don't know anybody in my office that uses it either, but last week I found out that many customers from other regions do use it, so maybe it's a cultural thing or works better for different languages, but I digress.
The AutoPreview of customised shortcuts looks somewhat odd when you have characters from outside the 7-bit ASCII range, such as the good old sterling symbol, £, but more importantly non-English characters.
A customised shortcut is a surprisingly complex thing, with the body being HTML wrapped up inside RTF, with some extra RTF parts for non-html clients. This works quite well in the reading pane, and extended characters display just fine. Outlook AutoPreview, though, doesn't use the HTML view; it uses a plain text version of the body, and it is this conversion which causes the issue. Somewhere along the line, the encoding of these characters was turned to the HTML &#nnnn in both the HTML and non-HTML parts of the RTF. When the plain text body is made, it is taken from the non-HTML parts, and that sequence is not recognised as a character sequence but interpreted as a simple string of the characters it consists of.
So the good news …I've detangled those pesky character encodings, and from 11.0.1 the reading pane and AutoPreview will show the correct characters.
The bad news is that I now have all those pound signs flashing before my eyes, but my bank balance hasn't changed.