Dear All
Today a new set of test versions have been placed on the Http://www.epidata.dk/testing.php for general download.
=============================================================== Changes to Manager:
* HTML Report window stays open. * Added additional reports. * Fixed bug on MAC where valuelabels could not be deleted at all in the ValueLabel Editor. * Fixed bug on Windows where valuelabels, comparison and "write valuelabel to" definitions could be added randomly to a field.
Changes to EntryClient:
* Added a WorkTool bar to main form. * Fixed bug where skip records (F5/F8) would cause program crash if dataset was empty. * Fixed bug where jumping to a field would not work under certain circumstances. * Fixed long standing bug where pasting text into uppercase field was not automatically made uppercase. ===========================
The bug fix since last test versions are essential to Windows users, which in the previous version could mess up use and definition of Value Labels and field definitions. Another issue was only relevant for Mac users. Please notice that files produced with the previous version might contain errors.
This underlines the fact that supporting multiple operating systems effectively creates a lot more work and complications, since the same problems - now fixed - was not seen in a Linux computer.
It also underlines the need for a batch testing procedure, which we are developing now, such that functionality simulating user behaviour will be implemented sometime soon. We plan to use the scripting engine to automate tests as shown below. More on this later. The simulations will be done on all supported operating systems. Once we have a few scripts developed they will be posted on the web such that users can enhance and extend the tests.
regards
Jens Lauritsen EpiData Association
List of tasks for simulator:
1. Create user simulation and gui automation tests. 1. Create a macro/script template for benchmarking tests. 1. create datafile structure (all field types first, then a new section with all field types) 2. define value labels 3. extend definition of datafile with more "checks" from field property page 1 4. extend definition of datafile with more "checks" from field property page 2 (extended) 5. extend definition of datafile with more "checks" from field property page 3 (jumps) 6. extend definition of datafile with more "checks" from field property page 4 (calculations) 2. Import/export data between the versions of epidata (in different formats) and check import/export success. - specifically import and export of bromar, lifetable and tiramisu datasets - import as reg file or stata file (where labels are attached directly). 1. Create a scrip to automate this 1. data integrity checks: e.g. go to record no xxx and compare the actual data with the imported value. Check data type (int,float,string,Date-time) 2. Check inconsistencies between xml/rec formats (if any) and see how gui interprets the xml data