INDEX | ORGANISATION | TIME LINE | PEOPLE | STORY | RESEARCH METHOD | COMMENTARY | FINDINGS | HELP | BOTTOM

Phase I (Sep 92 - Jun 93): Quality assurance review

'Areas of concern

The development method used is driven by user demands and it is well known that control is a problem. The team's initial plans recognised this, but there are still concerns over the way changes were controlled:

- Workshops did not seem to show people the impact of changes.

- The team did filter out some changes and defer others, but was this process rigorous enough? Severe time pressure (which firmer change control could have avoided in part) meant taking short cuts in system testing and cutting out the period planned for consolidation and review.

- The database grew in an ad-hoc way.

There was nobody with time available to enforce standards, although the team had recognised they are important and defined them at the outset.

Unit testing was not documented. System tests addressed unit test issues in part as is necessary with prototyping, but did not produce records of test data, even for only the more complex areas.

Although live load was three months ago, there is no technical documentation of the system. This is mainly because of resource problems and the need to finish some deferred work and enhancements, but the documentation work now planned for phase 2 must be given priority.

Although the team wrote procedures for controlling maintenance work, they are not being followed so there is poor control over promoting source and object code to live libraries.

There was poor control over the reinsurance module and it had to be rewritten.'

X028 page 004.01 Teddy Document 10/11/93

Social influence: Organisation: Resource issues
Technical influence: Process: Methodology


Recording
Story
Social influence
Technical influence

INDEX | ORGANISATION | TIME LINE | PEOPLE | STORY | RESEARCH METHOD | COMMENTARY | FINDINGS | HELP | TOP

© Clare Tagg 2000