In the track header (editor window, left pane) is a button labelled p (for "Playlist"). If you click on this button, Ardour displays the following menu:
Playlists are created with the name of the track of which they are associated, plus a version number. So, the first playlist for a track called "Cowbell" will be called Cowbell.1. This name will be used to define the names of any regions added to the playlist by recording. You can change the name at any time, to anything you want. Ardour does not require that your playlist names are all unique, but it will make your life easier if they are. Suggested examples of user-assigned names for a playlist might include Lead Guitar, 2nd take, vocals (quiet), and downbeat cuica. Notice how these might be different from the associated track names, which for these examples might be Lead Guitar, Vocals and Cuica. The playlist name provides more information because it is about a specific version of the material that may (or may not) end up in the final version of the track.
If you are going to rename your playlists, do so before recording new material to them.
It appears that recorded regions are not named after the playlist, but after the track.
It is entirely possible to share playlists between tracks. The only slightly unusual thing you may notice when sharing is that edits to the playlist made in one track will magically appear in the other. If you think about this for a moment, its an obvious consequence of sharing. One application of this attribute is parallel processing, described below.
You might not want this kind of behaviour, even though you still want two tracks to use the same (or substantially the same) playlist. To accomplish this, select the chosen playlist in the second track, and then use New Copy to generate an independent copy of it for that track. You can then edit this playlist without affecting the original.
One of the uses of playlists is to apply multiple effects to the same
audio stream. For example, let's say you would like to apply two
different non-linear effects such as distortion or compression to the
same audio source (for linear effects, you could just apply them one after
the other in the same track).
Create a new track, apply the original track's playlist, and
then apply effects to both tracks independently.
The same result could be achieved by feeding your track to multiple busses which then contain the processing, but this increases the overall latency, complicates routing and uses more space in the Mixer window.
Using Playlists for takes is a good solution if you are going to need the ability to edit individual takes, and select between them, but you won't be compositing multiple takes together.
Each time you start a new take, create a new playlist with p > New Later, you can Select your way back to previous or later takes as desired.
If you want to record multiple takes and then "comp" between them, it is probably better to simply record each successive take on top of the others in "layers" and then edit them using the layer tools, explained later.
The same approach as for takes is useful when you are recording or
editing content in multiple versions, such as dubbed movie dialog in
several languages, and you want all versions on the same track, to
get the same processing.
Select the appropriate language before exporting the session.