From IRC: See https://metacpan.org/source/RI[…]esultSet.pm#L1594 count is cached at the dbic level here which is why recursing count on the same resultset should be very very fast. https://metacpan.org/source/RI[…]esultSet.pm#L1229 next gets really clever ;)
https://bugs.koha-community.org/bugzilla3/show_bug.cgi?id=20819 --- Comment #79 from Marcel de Rooy <m.de.rooy@rijksmuseum.nl> --- there's loads of optimizatino in there regarding caching rows see _construct_results It seems that count is not cached in my test (so recursing not fast) and next is cached. I am not triggered right now to replace the next in Auth.pm. And for one next we are talking about milliseconds. -- You are receiving this mail because: You are watching all bug changes.