RM10 Drag & Drop crawls on large database

RootsMagic
, ,
1

Have you experienced slow progress when doing drag’n’drop of a large set of people? I tried doing one of everyone in a database with over a quarter-million people, 121k families, 0.9 million events, 48k places, 340k sources, 367k citations, 1k media items. The progress indicator stepped usually 5 people at a time. usually 10-20s between steps. So, if progress to completion were proportional to the number of people, it would take a week or two to complete! My computer is not super but it’s no slouch - Intel i5-6200U, 16GB RAM, SSD.

Using drag’n’drop is frequently recommended by Support to clean out unused items and detritus from a database and as a panacea for other issues but solutions that take hours, let alone days, are not practical.

I had developed an equivalent procedure using SQLite for RM7 databases 8 years ago that was very fast:

RootsMagic 10 needs to do better than it is at the moment, by a couple of orders of magnitude.

That’s a request for enhancement, @rzamor1 .:grinning_face:

2

Confirming request reported to development.

3

My understanding is that operation times should be fairly independent of database size- if the database has the proper indexes in place.

4

What’s that mean in practical terms, Richard? A small database of 1600 people on the same computer was DnD’d in ~15s - that’s over 100/s compared to 2-4s per person on the large database.

I agree that proper indexes are needed and can make a big difference. That might be the case here - back in RM4, I identified the need for an index that Bruce implemented but it was a simpler case than this one is.

Another factor, I suspect, is how much operation is taking place in the application rather than within the queries.