Question on performance

[lly-c232733]

Hello!

I'm diffing some large nested structures and running into really long run times with v3.

Interestingly v1 of this package returns in mere seconds with the correct changelog result.

However, I'm really looking for patching support that comes with v3. I've tried using options like 'FlattenEmbeddedStructs', 'DiscardComplexOrigin', and 'DisableStructValues' individually but nothing gets me to the performance that comes with v1

Is there an option or set of options in v3 to align performance with v1? What could be taking so much more time in v3?

Thanks for the help