Feature: rewindable source / setting audio delay

[jedi7]

HI, the idea is to have rewindable source (waterfall), with possibility to select at which time offset user wants decode audio. Usecase: User see an activity in waterfall, but is too slow (mostly) to switch to that frequency. So he can stop the device/waterfall and select from which time he wants play the audio. Or can select source delay, so he will have time to switch to proper frequency, where is visible activity.

NOTE: It will take bigger amount of memory (like 200MB for 10s buffer) More analysis will be needed.

[BatchDrake]

Copy paste of notes from our previous conversation:

Could we unificate two UIs for a feature that is very similar and I didn't have time to implement?
Like for instance, for file sources, provide some kind of slider to navigate and define the loop offset
Okay so how about this, we define a new widget, e.g. timeline widget
This timeline widget has a music player style slider and some time ticks
For file sources, the ticks are fixed, and the slider handle moves according to the progress in the file source
You can define looping intervals with, I don't know, two extra handles, in the beginning and the end.
For real-time sources the handle is fixed but the ticks are moving
Initially, the handle is in the rightmost side of the slider and you can move it to the left up to certain limit, which is defined by the buffering limit you configured
That way you basically define the amount of delay with respect the real time source that you want
In addition that you can set that up as well in the waterfall
The waterfall gives you a more visual way to set this delay while the slider lets you somehow fine-tune it
For real time sources, you use a method in suscan_source_t (accesible via suscan_analyzer_t) named set_delay, or set_source_delay
For file sources, instead you define a _set_loop_limit and _set_offset