Statmaster Revamped to use WS Library
In 2007 Statmaster was put in place to address some concerns and handle data automation. Those concerns were:
- Eliminate the need to carry a notebook computer to do spreadsheet data processing
- Minimize the use of human data entry errors by automating how data is processed
- Allow information to be updated in real time and provide an easy means to make data accessible on the web
While the original Statmaster application addressed requirement #1 by replacing the need for a notebook computer with a PDA, the remaining 2 issues were not satisfactory addressed. The second edition made use of the Wolffysoft Library and evolving standards to make the product more human friendly.
 |
 Read the Statmaster v1.0 Announcement
|
|
|
The Wolffysoft Library is built on the principle that the more basic of building blocks used, the more advanced solutions you can develop. Using this principle the back end was replaced with an easier to use system. In addition the Webblaze project was being prototyped so Statmaster could take advantage of this project to post results to the web. By making the back end data storage more user friendly if an error occurs it is easy to access the information to correct the mistake. The following table lists the improvements and the reason behind the improvements over the original version.
| Data Storage Improvements | |
Enhancement |
Results |
| Binary file format converted to XML | Should bad data get inserted in the back end storage, having a human readable file that can be edited in a number of possible applications allows for mistakes to quickly be modified. This mitigates the need to have an advanced editor built into the application. |
| Match result information seperated from player performance | Although XML was being used once loaded the data is kept in a virtual database format. This allows for the data to be stored as close to the fifth normalized form as possible and maintaining minimal memory use requirements. |
| Minimizing User Error | |
Enhancement |
Results |
| Rules are now objects and are executed based on one click | Once the user assigns results to each match, the data from that match is then passed in to the rules processor. Currently rules are fixed and coded into the application but follow a standard interface. |
| Recursion is key | The primary means used to assemble results is recursion. The trade off for the additional stack space is faster computation of the results with minimal instructions resulting in performance gain. |
| Rules are executed as a stack | Since rules are now standard they are loaded into the rules stack. The results of a match are passed through to each rule in the stack which can then assign proper processing criteria to each result. |
| Providing Results on the Web | |
Enhancement |
Results |
| Webblaze add on object process and render results | Because the back end information is stored in XML it is easy to create a Webblaze object to read the result and convert the information into formatted HTML for rendering in a standard web browser. Simply upload XML files to the Webblaze server. |
When the above disciplines and design preferences are combined into the application accuracy and performance benefited. By using the Webblaze model to render the data for the web the results could be customized using the template/page rendering engine to customize how the results are presented. This allowed easier access results by all participants who could then chart their own progress.
|
|
| Read about the Webblaze project
|
|
|
