palila_gui - v1.2.1

The main goal of v1.2.1 is the inclusion of the user manual. This version contains a visual overhaul of the navigation system. The demonstration screen can now be easily customised to meet experiment needs.

Release Notes for v1.2.1:

Added

• The demo parameter can now be defined as a block to define a custom demonstration screen.
• A first version of the user manual is now included.
• Added the example PALILA file from the manual.

Changed

• The 'Please enter your participant ID' part of the welcome message is now always visible.
• text is now an optional parameter for the Annoyance question type.
• An open question (how_found_other) is added to the questionnaire for when participants select 'other' in how_found.
• Made the "continue" and other navigation buttons static on the screen.
• Changed the default questionnaire's cold_magnitude question to MultipleChoice.
• The default questionnaire no longer uses manual split = yes.
• The configuration in gui_dev.palila has been made more relevant.

Deprecated

• Experiment configuration information will be removed from the README. This section will not be maintained from v1.2.0.

Removed

• Removed the default_questionnaire.pdf document. This information is now included in the manual.

Fixed

• An issue with the recognition of floating point numbers has been fixed.
• Without the [questionnaire] block, the GUI still showed the 'back to main questionnaire' button, which has now been removed in this case.
• The repetition index in the question identifier no longer appears when repeat = 1.

- Python
Published by spockele 10 months ago

palila_gui - v1.2.0

This update contains major back-end changes to the code structure, mainly targeted at duplicate and redundant code. This version also contains some fixes, deprecations and removals in line with planned developements.

Added

• Annoyance audio question type, containing the standardised 11-point annoyance question.

Changed

• Revamp of the progress bar: static regardless of screen, turns green when part is completed, resets when returning to main questionnaire.
• Removed final semicolon (;) from the multiple choice, multiple answer question type.
• Removed AttributeError from MultiMulipleChoiceQQuestion, because it was not relevant.
• Added way to enter multiple unlock_condition items by using a semicolon.
• Cleanup of the root directory. Moved Python support scripts to their own folder and modified the .bat scripts.
• FOR DEVELOPERS: this update contains major changes to the Questionnaire and Audio Screen logic, and their underlying management classes. The main changes are the equalisation and merger of both systems. Please refer to the commit history to understand these changes.

Deprecated

• The Questionnaire MultiMultipleChoice question type will be removed in future versions. Multiple-choice-multiple-answer questions can be created using the MultipleChoice type with multi = yes
• Questionnaire question IDs should now be defined in the brackets. They will always be standardised to the [part_id]-questionnaire-[question_id] format.

Removed

• The deprecated keywords dependant and dependant condition are now fully removed. They are now defined in the locked question and renamed: unlocked by and unlock condition.
• PointCompassAQuestion type has been removed, because it is unnecessarily complex.
• filler keyword removed from input system, to reflect desired GUI layout behaviour.

Fixed

• When returning to the main questionnaire from the end screen, the full questionnaire is now accessible.
• Pressing play on an audio now locks the continue button, to avoid the audio playing in the next screen.
• Fixed the demo screen at the start of the GUI.

- Python
Published by spockele about 1 year ago

palila_gui - v1.1.2

Release to alleviate vulnerability in the configobj dependency.

Changed

• Updated Python package dependencies to latest versions.

Fixed

• Changed configobj version to 5.0.9 in response to fix of CVE-2023-26112 ReDoS attack vulnerability.

- Python
Published by spockele over 1 year ago

palila_gui - v1.1.1

Bugfix version of v1.1.0.

Fixed

Hotfix for a critical navigation issue that makes v1.1.0 unusable. This issue was related to the incomplete removal of the unused pre-navigation function of the PalilaScreen class

- Python
Published by spockele about 2 years ago

palila_gui - v1.1.0

First major update of the PALILA GUI. This version is designed to iron out the kinks of v1.0.0 and to resolve some of its shortcomings. It also builds up to merging the codes that manage the audio and questionnaire questions, since they are very similar under the hood.

Known issues

• When replaying a sound sample, participants can continue to the next screen before the end of the sound sample. The running audio will continue playing while on the next screen.
• When returning to the initial questionnaire from the end screen, participants can only see the first of the questionnaire screens.
• The MultiMultipleChoiceQQuestion (questionnaire multiple choice, multiple answer) question type is not compatible with the system of dynamically unlocking question based on the answers of previous ones.

Planned changes for the next release

• A manual with a visual overview of the different options in the configuration file.
• Simplification of the README, mainly removing the description of the configuration file, since that will be in the manual.
• Fixes for the above-mentioned known issues.
• Bug fixes.

Added

• The number of replays of sound samples is now recorded in the output file:

  • For audio screens with 1 sample: <part name>-<audio name>(_<repetition index>)-replays
  • For audio screens with 2 samples: <part name>-<audio name>(_<repetition index>)-replays-left and <part name>-<audio name>(_<repetition index>)-replays-right

• A QuestionManager class to the questionnaire system to get closer to the system of audio questions.

• A ButtonAQuestion superclass for all question types that use buttons to answer (MultipleChoice, IntegerScale, etc.)

• A MultiMultipleChoiceQQuestion class to allow for multiple choice multiple answer questions in Questionnaires.

• Progress bar at the bottom of the screen to show progression through the experiment.

• A script to easily and quickly set up a new experiment.

Changed

• Code restructure to make the audio and questionnaire question systems more uniform.

  • Answers are now stored directly in a QuestionManager.answers, which is the Layout that holds the Question widgets.
  • Changed the multitude of functions to trigger an answer change to one singular change_answer() function.
  • The change_answer() function is now supplemented by a type-specific trigger function per question type.
  • ChoiceButton classes are now functionally the same.
  • questionnaire.py and questionnaire.kv have been split into questionnairequestions and questionnairescreen.
  • Logic for splitting the questionnaire over multiple sceens is moved to an external function questionnaire_setup().

• Change in the logic behind the keywords dependant and dependant condition:

  • They are now defined in the conditionally unlocked question and renamed: unlocked by and unlock condition.
  • All question types now allow for conditional locking/unlocking.

• An empty questionnaire block now results in no questionnnaire screen appearing at all.

• Repetition numbering of questions is now 1-indexed instead of 0-indexed.

• Timer now starts when leaving the welcome screen instead of when leaving the startup questionnaire.

Deprecated

• The keywords dependant and dependant condition will be removed in future versions.
  • This version supports the old system where these are defined in the question that conditionally unlocks another.
  • Future versions will only support the new system as described in change 2.

Removed

• N/A

Fixed

• Spinner Audio questions did not record their answer to the output file. This is fixed.

Full Changelog: https://github.com/spockele/palila_gui/compare/v1.0.0...v1.1.0

- Python
Published by spockele about 2 years ago

palila_gui - PALILA GUI v1.0.0 - Zenodo

Zenodo Release version of the GUI for the Psychoacoustic Listening Laboratory at the TU Delft faculty of Aerospace Engineering.

- Python
Published by spockele about 2 years ago