Develop timeline feature enough to complete first video project

[Venryx]

Tasks:

• MS imported audio-files persist between page refreshes. (while still being stored locally)
• MS you can assign a point in each audio-file to a timeline-step, as its "start time" in audio track. (stored locally only, but with export of all this info as a JSON file, for eventual re-import/restore functionality to prevent data-loss)
• MS start-times of steps for audio-file show up on that file's waveform ui.
• MS step durations can be derived from the start-times assigned in the audio files.
• MS playback of timeline also plays the associated audio segments.
• MS the timeline/audio panel can import multiple audio files at once, and have a button to auto-assign them to the timeline steps based on matches between the step-ids and the filenames.
• MS step->audio associations let you assign a volume-multiplier as well.
• Add a visual "talking-time tracker" feature, which is visible to editors, and in the video export (if/when enabled). [note that some segments should be excluded, eg. intro at start -- perhaps marked using the full/left/right positioning flags for now]
• Fix the issue where focus-nodes can apparently be "active" (and thus affect auto-scrolling), even when the given nodes are not actually visible in the main/non-load-helper map. [optional extra task: make-so a warning is given when the steps' fields try to do this]
• Rework the timeline Editor and Playing panels to have the switching be less jarring. (probably get rid of the "tab" approach, and instead have more granular toggles for specific functionality)
• Replaced node-reveals system with a more expansive step-effects system.
• MS the "time_relative" field of step-effects is processed/not-ignored. (probably requires a moderate rework, to have the accessors all based on time or reached-effect-index, rather than reached-step-index as it is now)
• Improve performance of processing that occurs while timeline is progressing. (for some cases at least, it's still not quite as fast as it was before the rework -- but the cleanliness [+reliability] improvements outweigh the likely slight-perf-loss)
• Start work on Electron wrapper app, for making interaction with the local filesystem more streamlined.
• MS electron app offers a streamlined way to record "takes" of audio (per step), pass them to the voice app, then receive the results + store them + assign them.
• MS new step-specific audio-file system is fully integrated. (replacing old audio-files-in-map-folder approach)
• MS electron app has streamlined screenshot-recording functionality. (todo: observability on the saved file/file-ranges)
• MS electron app has easy way to stitch the recorded screenshot-images and audio files into a combined video.
• Fix the following frame-recording issue: The frontend occasionally "moves on" and displays the next frame prior to the current frame having been fully captured -- causing "Frame10.png" to show the contents of frame 11, meaning frame 10 and 11 show the same contents, meaning slight jitter in video.
• Fine-tune the timings of the timeline node-effects, to make sense with the timings of the characters' speech. [project-specific]
• Add new step-effect type: show on-screen message

Likely tasks:

• MS TakeUI lets you do basic edits of a recorded audio file. (just trimming for now probably)
• MS the timeline panel scrolls itself to keep the timeline's current-time arrow in view. (by default anyway)
• Expand the node-effect struct to allow for (at least) unfocusing of all nodes under a given path.
• Make a "combined view" in audio-panel, letting you see (and change) time-assignments of each timeline-step, along with the waveforms of all of the audio-files that get triggered throughout the timeline.
• Add a button to copy the list of ids of each timeline-step (either full-ids or the first X characters of each id).

Possible tasks:

• Expand electron app to allow easy integration with audio tools. (for more advanced editing)
• Maybe rework audio-clip system to be just a form of "timeline-step effects".
• Add way to mass-change the volume-multiplier of step->audio entries based on patterns. (eg. text contains string)
• Extend timeline-step audio controls row, to allow custom ordering of the audio start-time assignments based on quality. (and perhaps with a star-rating system for knowing when a step has a "good enough" best entry)