The reason I've modelled the database like this, is to save space. I've created a utility which converts genealogical data from the Gedcom format to a MySQL-driven database. When dealing with genealogy it's fairly common to have individuals with more than one name; either you're not sure about the person's name, and want to refer to more than one names, or the person have a marriage name etc.

That name vs. lastname "bug" was a typo, and it doesn't affect the real problem here.

For simplicity, I could have had a name table looking like this:

person_id    mediumint unsigned not null, /* references 'person' */
firstname    varchar(255) not null,
lastname     varchar(255) not null
[download]

...but that's not really what I want. As I said, I want to save disk space, and by not duplicating data in the database I save a lot of space in this example.

However, after some reading I discovered Class::DBI::Join which seems to do what I'm after. The only problem is: I still can't get it to work. Any examples on using that class would have been great, as the example doesn't do it for me.

Thanks again!