Estimation and Measurement Telecon, 2009-09-25

See Weekly Meeting Logistics for telecon information.

Attendees

AndyBerner, ArthurRyman, LawrencePutnamJr, LeeFischman, ScottBosworth, SteveAbrams

Minutes

1. Alternative XML Format for Fact Tables - SteveAbrams

ACTION: Steve will present his alternative at the next meeting on 2009-10-02

2. Summary of Reestimation Use Case Discussion - AndyBerner

• Andy summarizes the discussion on the wiki page and added questions which we answered
• the purpose of reestimation is to provide revised estimates to completion that take into account the actual measurements to date and any other changes in the project
• the new estimates are used for two main purposes: portfolio review and project control
• the portfolio manager is interested in the updated cost and risk of the whole project since this may affect the business case
• the project manager is interested in the predicted values of key metrics (e.g. features completed, source code size, defects discovered, etc.) over time, together with control bounds, so that variances can be detected and diagnosed (analogous to statistical process control)
• Larry and Lee confirmed that their tools either already provide control bounds, or could do so - we will revise the draft EMS 1.0 spec based on implementation experience to validate this
• the whole project metrics, e.g. cost, are derivable from the time-bases metrics, and both are probability distributions so that risk can be evaluated
• reestimation is performed whenever appropriate, e.g. when a significant change occurs in the project, or a some logical point when new measurements are available, e.g. at the end of an iteration - but this is not determined by EMS 1.0 - it's dependent on the PPM tools to request reestimation
• any participant can create new Scenarios to be estimated, e.g. the estimator can create new Scenarios if the current Scenario becomes infeasible
• the PPM tools adopt the Scenario that has the most merit, and this becomes part of the new project baseline
• EMS 1.0 provides durable, secure, storage of its resources so the PPM tool can simply reference estimates via URL, or they may copy them - that is a PPM tool implementation detail and outside the scope of EMS 1.0

3. Software Estimation Use Cases Status - LawrencePutnamJr

When can we declare this document done?

• Larry and Andy volunteered to complete the Reestimation use case, target completion 2009-10-02, at which time we can declare the Use Case document complete.

ACTION: Larry and Andy will schedule a telecon, and invite Lee, to complete the Reestimation use case. Target completion date is 2009-10-02

4. Primer Status - ArthurRyman

When can we declare this document done?

• Arthur discussed how to stabilize the Primer
• Arthur recommended to back out the replacement of ESLOC with Scope in the Initiating scenario since it had ripples as described in email, but make it clear that many types of size input are allowed, e.g. function point, story point - this was agreed to
• Arthur will include the "effort by role" in the Primer since it adds useful realism
• Arthur recommended that we move the Primer into a higher level of change control, changing its status to, e.g. "Review Draft"
• Significant changes to documents in Review Draft status should be vetted in advance with the workgroup and made in cooperation with the document's designated editor - comments should be made via email or in the document's Comment box, not inline
• We need to wrap up the Primer asap and proceed to the Specification this quarter, aiming for implementations in 4Q

5. Other Business

• Arthur will be probably be in transit on 2009-10-09. Any volunteers to chair the telecon, or should we skip it?

Comments

Add your comments here: