In my tests with the performance suite I'm seeing 5-10% improvements in the relevant parts of performance-test/T00-new.sh. That's not nothing, but it does require an API change, so if a few people could test these patches on real world large scale "notmuch new after mv/rm" I would appreciate it. I'm particularly curious about results from spinning rust disks.
_______________________________________________ notmuch mailing list -- notmuch@notmuchmail.org To unsubscribe send an email to notmuch-le...@notmuchmail.org