FC (Fay) Stegerman

I finally had it happen locally, once. CPU fan was blowing hard. So some kind of busy loop or something. Sadly no way to debug when it's this infrequent.

After this change, the only errors ignored during import should be duplicate groups (which is fine), unless the data being imported does not conform to the expected format somehow, which...

> unless the data being imported does not conform to the expected format somehow It looks like the parser is handling that already. So I can't guarantee any overlooked edge...

There could be an NPE from invalid data. But that should be caught by the importer and cause the import to fail.

> I'm wondering, wouldn't it be cleaner to change how the id database value for cards is generated open card creation (like: unix timestamp instead of incrementing by 1 for...

Also note that only catima exports contain IDs. If we want "intelligent merging", it seems good to apply that to all other import formats as well, not just catima. For...

> prefer existing data / prefer imported data / don't check for conflicts A fourth option would be to prefer latest `lastused`, but that seems a bit brittle.

> I can add that to this PR if you'd like. Or we could merge this as-is and I'll make a follow-up PR for that. Let me know what you...

actually, I think this would not be too hard after all: * ask / prefer existing data / prefer imported data / don't check for duplicates (i.e. what it does...

> severely complicate multi-device sync That wasn't something I was considering in this context. But good point. We should address that. > (Temporary hack) Drop all data before import #1389...