Chandler Engineering Meeting, Thursday April 19th 2007

Present: Katie, Sheila, Aparna, Grant, Philippe, Andi

Agenda

• Preview Schedule: Engineering perspective, what's missing? what are the risks? how can we mitigate them?
• Performance: Status of last 2 weeks progress. Areas still in need of work? Promissing directions? Also, issue with the PPC numbers, desperately low and unlikely to get to the "acceptable" level of the other platform. How do we handle that?
• 0.7Release Bugs triage: Let's discuss the email sent by Sheila on the list yesterday
• QA/Build: Heikki and Aparna to brief us on the recent QA/Build meeting

Notes

Preview Schedule

• • Do we want to extend the Perf period? Should we keep the momentum? Should we stop the period as planned and get back to bug fixing?
  • Idea: continue for 1 more week while marking more bugs as blockers so that dogfood can start for real while making more perf improvements.
  • Depends of course about what we'll achieve in this coming week. We'll assess this next week

Performance

• • Real data set: Morgen blocked by bkirsch (the obfuscator needs to clear real addresses and replace by test addresses). Philippe to contact Brian on this.
  • PPC targets: we should have different acceptable target for that platform. Heikki to modify the report table.
  • Grant to post a new perf work summary to the list

0.7Release Bugs triage

• PPD did a first pass but didn't modify the Bugzilla records
• Decision: use the priority flag, Sheila to go through all the 0.7release bugs and prioritize them, Philippe, Katie and Aparna to monitor and triage also after Sheila's done. We'll clear the target milestone field for the controversial ones. Those will be reviewed in Bug Council.

QA/Build

• Sharing and migration: lots of issues, 0.6.1 won't be finished by next week (never been successful doing one publish/subscribe cycle)

-- PhilippeBossut - 19 Apr 2007