OpenJDK bug database: DRAFT Developer Workflow

[Brian Goetz]

> For instance, if we have two sub-states for New (unqualified and qualified, the distinction being initial triage
> ensuring that the information in the bug report is sufficient to do something with it), we could have a desired
> SLA (service level agreement)  on the time to move between these.  If we are consistently failing to meet this
> in a particular area then it shows us we need more investment there (or we need to decide we can accept a
> less aggressive SLA).   This kind of query can easily be automated.  Maybe this is exactly the kind of boundary
> system you are worried about, though?

What I'm worried about is mostly the discovery effort of individual 
managers trying to craft their own queries, and each getting it slightly 
different.  Substates increase the likelihood of this, which leads to 
the call for centralized boundary systems.  If such queries can be 
canned and reproduced to dashboards, or the constituency for them is 
small, then it probably works fine.