Row Labels 1830 1840 1850 1860 ALEXANDER, John-10 1 1 1 ALEXANDER, John-17 1 FERGUSON, Martha-18 1 1 KEEBLER, Jacob-50 1 1 1 KEEBLER, John-51 1 WOODS, Levi-16 1
Tom, this discussion might better go on your SQLite wiki, but I'm thinking of going the other direction. Which is to say, for U.S. censuses I'm thinking of creating user defined fact types called census1790, census1800, census1810, etc. through census1940. These user defined fact types would be exact clones of the built-in census fact type. For example, there is nothing that would tie the census1790 fact type to the year 1790 except for the name of the fact type. Then I would have a script that changes all census events for the year 1790 to census1790 events for the year 1790 and so forth for each U.S. census year. Nothing would change in any report that would come out of RM.
Why would I do such a strange thing? Because then I could profitably put census events into People View. At the present time, it's a big waste of time to try to put census events into People View because census events are effectively duplicate events and only one such duplicate can be displayed in People View. With the new scheme, I could put any particular U.S. census year into People View.
And of course, on the back end, I would want to restore each of the census1790, census1800, etc. events back to standard census events. It's probably going to be a few weeks until I have time to put the script together (even though it's a pretty simple script). So maybe with a little luck (and if this idea tickles your fancy), you might have this code done within the next five or ten minutes. You are way faster with SQLite than I am. This idea is even a good candidate for RMTrix. And if it doesn't tickle your fancy or if you don't have time, I'll do it myself in a few weeks.