Apparent issue with interrupted AutoMatch operations on large databases

@TomH I know that your question wasn’t addressed to me, but I will answer it anyway; I have practically given up on automatching to FS. I only have 60,000 people in my tree, but it takes me a full day to get through the whole database, and it is only towards the very end that it starts matching any significant number of people. What’s more, there is a serious risk that the process will get stuck part of the way through, in which case I have to start again from the beginning, having had almost no benefit.

What’s worse is that RM9 does not have RM7’s function to auto-match when it finds hints (although I know that this is on the list for development) and manual matching is far, far, far far, far slower in RM9 than in RM7. So linking any but a very few new records to FS is completely impractical. I find this extremely unsatisfactory.

In my humble opinion, its integration with FS is a key selling point for RM, so failing to make it work properly seriously undermines RM’s value.

I think it is a reasonably high priority to fix all these things

  • Ability to start auto-match at a certain RIN
  • Ability to pause and re-start auto-match (which is essentially an application of the point above when you think of it)
  • Option to auto-match on finding web-hints
  • A manual match process as efficient as that in RM7

However, I know that development resources are very limited and the wish list very long, so I proposed running auto-match from the highest RIN downwards as I thought that this would be the easiest quick win.

2 Likes

AutoMatch runs starting at the lowest record number of unmatched people. Have you tried letting it run overnight.

When I wrote ‘a full day’, I should have written ‘24 hours’. One overnight is not enough.

@TomH

My workaround for Automatch has been to do anything FamilySearch related on a PC instead of a Mac.