Re: [openstack-dev] [release][stackalytics] possibly breaking bug closing statistics?

2015-11-11 16:38 GMT+03:00 Thierry Carrez : > > date_fix_committed is probably not set if we directly switch the bug to > "Fix released", and that is what we plan to do now with Launchpad bugs. > > We might therefore need a backward-compatible patch to Stackalytics so >

Re: [openstack-dev] [release][stackalytics] possibly breaking bug closing statistics?

Excerpts from Ilya Shakhat's message of 2015-11-11 17:17:58 +0300: > 2015-11-11 16:38 GMT+03:00 Thierry Carrez : > > > > > date_fix_committed is probably not set if we directly switch the bug to > > "Fix released", and that is what we plan to do now with Launchpad bugs. > >

Re: [openstack-dev] [release][stackalytics] possibly breaking bug closing statistics?

Ilya Shakhat wrote: > Doug, > > You are right, there should not be any changes in the stats. Bugs are > mapped to release only by date, target milestone is not taken into > account. For resolved bugs metric Stackalytics uses 'date_fix_commited' > field, for filed bugs metric - 'date_created'.

Re: [openstack-dev] [release][stackalytics] possibly breaking bug closing statistics?

On 12 November 2015 at 03:17, Ilya Shakhat wrote: > 2015-11-11 16:38 GMT+03:00 Thierry Carrez : >> >> >> date_fix_committed is probably not set if we directly switch the bug to >> "Fix released", and that is what we plan to do now with Launchpad bugs.

Re: [openstack-dev] [release][stackalytics] possibly breaking bug closing statistics?

Doug, You are right, there should not be any changes in the stats. Bugs are mapped to release only by date, target milestone is not taken into account. For resolved bugs metric Stackalytics uses 'date_fix_commited' field, for filed bugs metric - 'date_created'. Thanks, Ilya 2015-11-10 0:53

[openstack-dev] [release][stackalytics] possibly breaking bug closing statistics?

Stackalytics experts, When we roll out the launchpad process changes described in my earlier email [1], we will no longer be targeting bugs as part of closing them. Thierry raised a concern that this might break the way stackalytics calculates closed bug statistics for a cycle. Looking at the