15:01:26 <dmcbride> #startmeeting colorado release
15:01:26 <collabot> Meeting started Tue Apr 12 15:01:26 2016 UTC.  The chair is dmcbride. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:26 <collabot> Useful Commands: #action #agreed #help #info #idea #link #topic.
15:01:26 <collabot> The meeting name has been set to 'colorado_release'
15:01:49 <dmcbride> #chair frankbrockners
15:01:49 <collabot> Current chairs: dmcbride frankbrockners
15:02:11 <frankbrockners> #info Frank Brockners (Genesis, FDS)
15:02:20 <ChrisPriceAB> sorry folks cannot attend, attending an EU thing this afternoon/evening
15:02:32 <dmcbride> #topic roll call
15:03:05 <anac1> #info Ana Cunha (Yardstick)
15:03:11 <bryan_att> #info Bryan Sullivan
15:03:15 <jose_lausuch> #info Jose Lausuch (functest)
15:03:20 <trozet> #info Tim Rozet (Apex)
15:05:22 <fdegir> #info Fatih Degirmenci
15:05:52 <sofiawallin> #info Sofia Wallin
15:07:17 <Julien-z_> #info Julien
15:08:50 <jose_lausuch> dmcbride: reduce the volume of your speakers maybe :)
15:09:20 <bryan_att> what's driving the 2.5 months after code freeze that most projects have claimed they need?
15:09:34 <bryan_att> that's almost half the release cadence
15:14:27 <bryan_att> what does "projects are associated with scenarios" mean?
15:15:43 <bryan_att> projects should be able to claim (and demo) compatibility with scenarios, but not require work by the project "team" (whatever that is)
15:16:10 <trozet> bryan_att: is there a gtm for this?
15:16:30 <bryan_att> hang on, I'll get it
15:17:37 <bryan_att> ID 971789629
15:18:03 <jmorgan1> https://www.gotomeeting.com/join/971789629
15:18:12 <trozet> bryan_att, jmorgan1: thanks
15:19:00 <bryan_att> +1 to reducing the number of scenarios, or at least not blowing it up by permutating features on top of basic scenarios
15:20:23 <trozet> I would like to see less run time for functest/yardstick, functest takes 2.5 hours per scenario
15:20:50 <B_Smith_> What is the gotomeeting ID for colorado planning
15:21:15 <bryan_att> B_Smith_: https://www.gotomeeting.com/join/971789629
15:21:22 <trozet> yardstick takes 1 hr when it passes
15:21:32 <jose_lausuch> trozet: we have plans for colorado to reduce execution time
15:21:50 <B_Smith_> thanks...could not find on meeting page
15:22:02 <frankbrockners> #link https://www.gotomeeting.com/join/477545189
15:22:21 <trozet> jose_lausuch: sure I'm just saying 3.5 hours of testing for a 50 minute deployment... # of scenarios is not the issue
15:22:22 <frankbrockners> 6am PT
15:22:38 <trozet> thats just my opinion anyway
15:24:16 <bryan_att> trozet: # of scenarios is not the issue *now* but would be if we dramatically increased them by requiring all features be represented in a scenario, thus require a unique CI/CD run and in combination...
15:24:34 <bryan_att> we would need dozens of PODS
15:25:03 <anac1> yes, yardstick takes about 1h
15:28:09 <bryan_att> we need to apply an agile process to testing, so it's natural that test development parallels actual testing. We should not try to enforce a waterfall process by making a "test cases developed" milestone early in the release.
15:28:49 <billyo> #info Billy O'Mahony
15:29:28 <Simona_> #info Simona Coppola (Intel)
15:30:40 <bryan_att> the difficulty in predicting the complexity of test failure is a key reason why we have to have test dev in parallel to testing.
15:32:56 <anac1> bryan_att: yes, that's what we want to apply, test in agile SW development to catch faults early
16:00:30 <dmcbride> #endmeeting