Limesurvey offers a wide set of question types, and a number of distinctions are completely irrelevant for the Koord2ool, e.g., whether the user will select "one out of X given options" via a drop-down or with buttons.
Internally, koord2ool will have a number of different code-paths to process the responses into data ready for graphing:
single choice
yes/no questions
select one of multiple pre-defined options
This is drawn be counting the answers for each option an generating a graph for each option. (plus n/a)
Freetext answer
various types of textfields
This is likely to be a good choice for a default code-path
This is drawn by listing the users on the y axis, and drawing dots with mouse-overs containing the answers
Limesurvey offers a wide set of question types, and a number of distinctions are completely irrelevant for the Koord2ool, e.g., whether the user will select "one out of X given options" via a drop-down or with buttons.
Internally, koord2ool will have a number of different code-paths to process the responses into data ready for graphing:
This is drawn be counting the answers for each option an generating a graph for each option. (plus n/a)
This is drawn by listing the users on the y axis, and drawing dots with mouse-overs containing the answers
What's definitely needed is an easily extended configuration that maps each question to the right koord2ool code-path.
The first iteration could be an in-code list of survey-types for each code-path.
Second iteration might be a config-file.
Third iterations might look at the answers themselves to select the optimal code-path. This might be needed to detect numerical answers.