Comment 3 for bug 1383539
Revision history for this message
| Barry Warsaw (barry) wrote Re: phased update support does not give idempotent answer for each (machine,update) | #3 |
© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
dbe715f
demo site
(Get the code!)
... Okay, so we rip out phased % calculation from index build-up, which means we calculate the winning path as if we *were* going to update to the target. Then we calculate the winning path as if phased % is 100. Then we use the target build number of the winning path in the pu calculate and if that leads us to a device-% that is > phased-%, we simply report that there's no update available.
This makes sense, and although there's a subtle semantic difference from what was happening, it's probably what was originally intended. The implementation change is not too risky.