[bind10-dev] A-Team Sprint Planning Meeting 26 Jan 2011

JINMEI Tatuya / 神明達哉 jinmei at isc.org
Wed Jan 26 18:49:22 UTC 2011


Sorry, I overslept this morning and missed the planning call...(I'll
use two alarms next time).

At Wed, 26 Jan 2011 16:34:49 +0000,
Stephen Morris <stephen at isc.org> wrote:
> 
> Following the A-Team sprint planning session today:
> 
> * A new sprint milestone has been created
> * (http://bind10.isc.org/milestone/A-Team-Sprint-20110209) and all
> * tasks not marked as complete have been moved to it.  This has also
> * been marked as the "Current A-Team Sprint" on the "Sprint
> * Progress" page (http://bind10.isc.org/wiki/SprintProgress).

I have a proposal: it would be better to separate milestones for
statistics and non-statistics related tasks because different and
disjoint sets of developers work on these.  It would be misleading,
e.g., to see 8/10 tasks are closed in the sprint progress bar:
http://bind10.isc.org/milestone/A-Team-Sprint-20110209

because e.g., it could be possible that all of the 8 completed tasks
are statistics-related and there may have been no progress on the
generic authoritative work.

I understand that separating these may introduce an additional
administration overhead for you.  If this is a concern, I volunteer to
maintain the milestone/progress pages.

> * An A-Team task backlog milestone has been created
> * (http://bind10.isc.org/milestone/A-Team-Task-Backlog).  If you
> * identify any additional tasks for the resolver, please create them
> * and add them to this page.  We'll discuss them (and maybe estimate
> * them as well) at the next sprint planning session.

This is nice, I wanted to have a storage like this (except that "for
the resolver" should be "for the authoritative server/statistics" :-)

> Michal will create tasks for tackling performance and add them to
> the current sprint.  We'll do the estimating for these by email -
> when the tasks have been created, I'll send round an email asking
> for your estimates then take the consensus and/or arrange for a call
> to discuss them

I don't object to the idea, but I'd like to know the specific,
concrete requirement of year-2 optimization (including whether we need
to do anything within year-2 in the first place) as I said in the
"Profiling & performance sugestions".  I'm expecting this information
from Shane.

---
JINMEI, Tatuya



More information about the bind10-dev mailing list