Is there a reason why Copy if not exists is implementated in your S3 implementation? Because Cloudflare R2 supports it with their S3 implementation, so I'm wondering why didnt you guys do it in Lakefs

Hi Ion, I'm not sure why we don't have it. For a while we (I) blocked put if-not-exists out of concerns that this could block future performance improvements. I assume we may have done the same for copy. There may be some concern about MPU copy on the S3 gateway, though. If you open a bug we'll prioritize.

Hey Ariel, here it is :) https://github.com/treeverse/lakeFS/issues/7642