© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
35a32c5
(Get the code!)
looking at the lsprof of 2.0, it is because it is using the method from 1.18 and applying it directly to the 2.0 data format, which is bogus.
Specifically, the 1.18 method is to read in the entire inventory, then filter out the data which is not in the subdirectory, then compute the remaining delta.
In 2.0 the proper way to do it is to compute the delta, then filter out what is not in the subdirectory.
Reading the whole inventory in 2.0 is slower than in 1.18, but the tradeoff is that we can compute deltas *without* reading the whole inventory, which makes it a lot faster when done properly.
I remember raising the issue in the past, obviously the fixes never got pulled into the codebase.
Thanks for poking us on this bug.