There are currently hard-coded logos (Levvel) and texts in the frontend and player that should be configurable by the user of 360Composer. This would allow everyone to "brand" 360Composer as they wish.

Currently, a Scene stops playing at the moment where an annotation is set in the Scene editor. Add an option so that the scene creator can choose whether the Scene stops at the annotation or continues playing until the end.

Currently multiple parts are made by the students themselves with potential security risks while this is already solved by numerous extensions for Flask. In this issue we'll keep a list of potential Flask extensions that can replace code from the students.

• Hash password creation could be replaced by bcrypt. This provides better hash-algorithms which we don't have to maintain ourselves.
• Authentication is now created from scratch while that could be solved by Flask-login

This applies to the annotation editor, as well as the scene player.

I'm not sure if it's possible to make it work, perhaps display a message to the user asking to disable resistFingerprinting.

Currently it is very clear that it has been made by multiple people with different ideas on how to implement the UI design. For example the LoginForm component has its own custom made style while other components make use of MUI. This makes the user experience highly irregular and should be made consistent. I would suggest to use MUI consistently throughout the application.

Add a facility that selects the optimal video bit-rate/resolution based on the performance capabilities of the client device (: the player).

It would be a great asset to have an "undo" facility in the online editor...

There are cases where it would be useful to sue a Scenario that was created before as part of another Scenario. In that case it would be useful to be able to add that Scenario as if it were a Scene.

There are many situations in which it would be useful to add 3D objects to a 360 camera scene so that they appear to be part of the scene. For example; adding a 3D model of a machine to a filmed factory floor.
This is a potential Pandora's box: once this would be added, chances are high that plenty of other features are going to be asked for next: add animation, physics, interaction types, etc. (basically: add all functionality of Unity/Unreal)!

The current front-end software does not have any way to remove users (there is a way to add users) or change password, or an option to reset a password.

The current "analytics" functionality records events as they occur during a session; as described in section 4.2.2 of Emil van Veen's thesis): "1) the start and end of a timeline, 2) the start and end of a scenario, 3) the start and end of a scene, and 4) the occurrence of a key interaction moment". The problem is that the records that are produced are not "ACID" structured, which makes robust analysis hard. For example: "start" and "end" events of a timeline, scenario or scene are not always balanced (i.e. there is no matching "end" with every "start", which can happen when a client is interrupted). It is also difficult to tell apart sessions and interaction, which is wanted in the dashboard by Sharon Monk.

Currently the database password does not except special characters, my local postgresql has a random 32 character random master password which is not accepted as password and gives me this error:

pg_isready: error: invalid percent-encoded token:

I tried to URI encode the password but then docker-compose trips over the percentage sign which you can then escape with double percentage signs but then the database trips on it again. This needs a proper solution because special characters are commonly used in passwords. For now I simplified my local db password because there is no sensitive data in it anyways.

There is currently no way to control where a text annotation appears in a Scene. Add an ability by which a scene creator can control this, either at a fixed location in a Scene or (relative to) the current viewpoint.

It would be nice to be able to zoom in/out of a scenario using a mouse scroll wheel.

There is currently no way to delete a project, only to add a new one.

Currently, the user needs to confirm the continuation from one Scenario into the next in a Timeline. Add an option by which this interaction can be disabled so that the Timeline immediately continues into the next Scenario.

For analytics purposes it would be nice to be able to record where a end-user is looking while playing a Scenario. This can be used to monitor if the end-user is acquiring the correct information from the Scenes before answering the questions.

It is often useful to be able to create a new Project from an existing Project and build on that.
Likewise: in the editor there is currently no way to make a copy of a Scene, Scenario or Timeline.

The video players in the scene editor is based on Babylon while the XR player is built on A-frame. This results in discrepancies, such as different locations of annotations. It would be good to use the same code for both video players.

In the current implementation for each video fragment a separate <video> element exists in the A-Frame assets list. This causes all videos to be loaded at the same time, which may cause an excessive memory usage especially for larger number of videos or longer/higher quality (i.e. 4K) videos. This may cause problems especially on weaker devices.

Currently, all Scenes start with the same camera viewpoint. Add an ability by which this initial camera angle can be set.

The scenario editing environment does not scale with the available screen space. This is inconvenient if you're creating a complex scenario that runs off the screen. You can scroll through a scenario by dragging the background of the scenario, but is not completely clear for someone new to the editor. In that case it would be helpful to show scrollbars.

Now every asset, scene, scenario or timeline only have one related project which should should be a one-to-many relation. This way duplicating projects or re-usage get a lot easier.

Add an annotation/interaction possibility that controls the progression through a Scenario based on where a user is looking. This would result in a possibility for a scenario creator to connect Scenes based on the region in which the end-user is looking at a Scene.

In the Scenario editor; the lines that connect Scenes are often invisible if they pass through other Scenes. This quickly results in difficult to understand "spaghetti", but sometimes causes important side effects. For example: a Scene output that goes into its own input is possible, but is completely invisible and can therefore not be deleted!

Currently, all Scenarios in a Timeline are randomized. In some cases it is necessary to exclude Scenarios from randomization, for example: the "introduction" scenario should always come first, or the "exit" scenario should always come last, or a scenario that introduces the next scenarios should always be at a specific moment.

Add a possibility to select the scenarios that are randomized.

In the scenario panel: if you click on the trashcan in a scenario, a pop-up appears "Are you sure you want to remove this scenario?", but if you then press "Remove" the scenario is not deleted and a red message appears "Unable to delete scenario.".

Currently there is a Alembic workflow by changing everything by hand. This created a severe difference between the Flask-sqlalchemy models and the database migrations. One automatic migration already showed about a 100 differences between the models and the database (from the migrations directory).

There are situations where the end-user requires an optimal Quality of Service or has limited connectivity and is unable to reach the server backend. For these cases it would be nice to have an "export" facility that contains a completely self-contained package that can be installed on the client device and that has the same functionality as the on-line version.