|
Solutions Overview Access to ASP.Net Client/Server Java |
MS-Access to DBMS Conversion Metrics
The following charts illustrate the differences in project duration and accuracy based on the complexity of the MS-Access application as defined below. The figures have been compiled from hundreds of real-world MS-Access conversions that Bunker Hill has performed since 1994.
Manual versus Automated Conversion Metrics
The first graph depicts the number of weeks for a full-time MS-Access and part time DBMS specialist to complete conversions based on application complexity. As the complexity of the application increases, the completion time for the manual approach takes 10 times longer for databases of average (5) complexity and soon reaches infeasible levels.
Time to Completion versus Application Complexity
Conversion Accuracy
The second chart depicts conversion accuracy based on application complexity. Notice that even relatively simple databases lose accuracy, and therefore data integrity, for all but the simplest of conversions.
Complexity Factor
The complexity factor is based on a combination of MS-Access application components and conversion goals.
MS-Access Application Complexity Factors:
- Number and size of objects (tables, queries, forms, reports, VBA code, etc.)
- Advanced database design techniques employed (relationships, validation rules, Input Masks and Lookups used, etc.)
- Advanced query techniques employed (non-ANSI standard functions, like IIf(, Format(, and domain aggregates, special features like top values, alias substitutions, user-defined functions, etc.)
- Database and application design complexity and quality
- Case-sensitivity, performance, and Access/DBMS interaction issues
Conversion Goals/Alternatives:
- Replicate MS-Access table structures and optionally export data
- Convert (re-platform) MS-Access queries to run on the DBMS server. Note that a complexity level 6 and higher assumes query conversion requirements
- Create new Access front-ends that retain the look-and-feel of the original applications but operate with the DBMS backend
- Note that most of the technical issues discussed apply to MS-Access dependent Visual Basic conversions as well.
For example, a database with 10 simple tables but no front-end or query conversion requirements is assigned a complexity factor of one, and applications with hundreds of tables using advanced techniques and requiring query and front-end conversion are assigned a complexity of ten.
Conclusion
As the charts illustrate, manual or semi-automated conversion methods, when feasible, require much more time to implement and sacrifice accuracy and application integrity. Manual conversions also require much more rigorous and granular testing. Even the most stringent parallel testing may not uncover some logical errors with potentially disastrous post-production consequences.
