Author Topic: [FAQ] Merging PVD Databases  (Read 1980 times)

0 Members and 1 Guest are viewing this topic.

Offline CRSquared

  • Member
  • *
  • Posts: 1
    • View Profile
[FAQ] Merging PVD Databases
« on: July 16, 2010, 02:54:36 pm »
Hi all,
I know this sounds weird, but I have well over 30k titles, and, although I've gotten them all into the database, it'll take forever to import prod info. I want to split it up and have a friend work on half of it, but I can't figure out how. I tried exporting to csv, then back again, but it adds a bunch of semicolons in among the commas, and the whole thing was a mess. Fortunately, I'm just starting this so I can afford to screw up the database now, but when I have thousands already entered it could be problematic. Any suggestions?
« Last Edit: September 22, 2010, 01:26:14 am by rick.ca »

Offline rick.ca

  • Global Moderator
  • *****
  • Posts: 3241
  • "I'm willing to shoot you!"
    • View Profile
Re: Merging PVD Databases
« Reply #1 on: July 16, 2010, 06:26:48 pm »
Welcome, CRSquared.

As you've found, PVD does not yet have a facility to split and merge databases. It has been requested, and will probably be added at some point. While exporting and re-importing works fine, it's not really a practical solution to your problem. If you want to split your database, it would be much easier to make two copies of the entire database and then delete from each one the records you don't want. That might be a sensible approach—if there were a logical way to do the split (i.e., each contains fundamentally different items), and you were content to use them separately until a merge feature is available (an indefinite wait). Otherwise, running two databases only means more work and it's own problems—especially with respect to managing credits.

You might consider installing Firebird Server, so multiple users can connect to the same database. This would work well if your friend is local and can connect via a LAN. It would also work connecting over the Internet, but I don't know if the performance would be good enough to make heavy-duty maintenance work (e.g., mass updates) practical.

Considering the number of titles, you should consider carefully how to maintain credit data. It's probably impractical to download all credits and maintain filmography data for all people. Downloaded credits can be restricted by selecting "...Except actors (from main page only)" in the plugin configuration. This limits actors to the first 15 credited. You can also select "Do not add people from awards if they do not appear in credits" or, maybe better, forgo awards data all together. For the people who are added, you'll probably need to be selective about which, if any, are to include filmographies. To illustrate the issue... My database includes about 2,000 movies. Actors are limited to "main page" and people are not added via awards, but there are still 31,000 people in my database. Since I have no practical basis for deciding which of these should include filmography data, they all do. As a result, there are 236,000 additional ("invisible") movie records in my database (i.e., movies in these filmographies that are not otherwise in the database). You might find your database a bit sluggish if you allow 3.5 million records to be added. ;)

Now I'm curious. In what circumstances do you need a database of 30,000 titles? It seem a higher number than you could own, be able to view, or even consider viewing in a lifetime. Is it some kind of special purpose of thematic database? If not, maybe there's a way to limit the number (or logically split the database—as suggested above).
« Last Edit: September 22, 2010, 01:25:41 am by rick.ca »