Logbook parser now looks for mugshots when adding people. If it finds an image, it adds a new photo model instance linked to that person. If it finds an html page, it adds the body text of that page as the person's "blurb" and then adds the mugshot.
My interface web design isn't working, sorry. Need a blank slate on that one.
Also, the urls for the person model now allow lookups using their actual first and last name in the url instead of the primary key. We should do similar things for all models, maybe even using the url as a search which results in a page with multiple model instance results if the input is ambiguous (e.g. they only enter a first name).
