First Release of Mesort
This is the first release of Mesort, deployed at https://kommunikative-figurationen.de/mesort .
To access this project you are required to contact one of the project members:
Technically we need to create a user via command line or the newest admin area, then the invited user has to set a password (min. length 6) to access.
The main feature of Mesort is to create studies for sorting tasks: this is made from general information, tokens and questions with related answers. Tokens can be chosen from the presets or uploaded from the researcher's device, also:
- general information is required.
- a minimum of one token is required.
- questions are not required.
There is currently no limit to the number of studies. A study can be deleted before an associated interview is saved; if an interview is already added to the study only an admin can delete that study via command line. When you delete a study, you delete also all the corresponding token files.
Interviews are done with the given data from the study, the sorting is saved as a high resolution screenshot. Each token saves its x and y position in pixel, percentage position, distance from the center and in which circle it is positioned. The interview structure so far consists of:
- Pre-sorting questions.
- Discuss Sorting phase.
- Post-sorting questions.
Questions are saved if you go back, if you go forward to discuss the sorting and back to the sorting itself, the tokens are in the starting position. You can end the interview at any time and no data is saved.
A Researcher can consult each interview data and export the sorting by clicking on the image of the sorting.
Single interviews cannot be deleted.
The functionality of the frontend is mostly on Vue components:
- overviewuser.vue - not used, kept just to reuse the table reference in the future
The backend is built in Laravel in an MVC pattern. Static methods are added to the models to improve readability but the code still needs continuous optimization. An additional paper about how and why I wrote the code in this way is at this day being written.
Using Php insights, a tool that measures the quality of the written code, Mesort scored this:
This is the current database structure:
The website is meant to be used in landscape mode and optimized for the latest tablets, with a minimum width of 1024px. Less than that, we can not assure the whole functionality.
There is already the predisposition for a multi language platform, but no other languages are yet available.
An API interface is not deeply implemented since there are no external requests to access the data.