abcde.CDDISCID/
	discid
		Text file with the output of the 'cd-discid' program in it.
		Since the directory is partially named after the discid,
		if this file does not exist the directory is assumed to not
		be a valid abcde temp dir and will be recreated.

	status
		Text file describing the status of the abcde run.
		Status blurbs are one per line, and should represent a
		unique completed action. A status check should consist of a
		grep for ^blurbname$ to see if it exists or not. Certain
		blurbs (as documented below) may contain content in the form
		of 'blurbname=content' to denote a choice a user has made.
		In this case a grep would be made for '^blurbname=' and the
		portion of the line after the = sign read.

		List of variables:
		cddb-querycomplete
		cddb-read-X-complete
		cddb-readcomplete
		cddb-choice=X
		readtrack-TRACKNUM
		encodetrack-TRACKNUM
		tagtrack-TRACKNUM
		movetrack-TRACKNUM
		playlistcomplete
		encode-output={silent,loud}
		exit

		TRACKNUM is the number of the track in question, starting
		with 1, not 0.

		Any thread may append to this file. It is never truncated,
		but cleaned when the entire directory is wiped.

		The -a commandline option to abcde specifies a
		comma-delimited set of actions to be done. The specific
		correlations are as follows:
		
		cddb
			cddb-query-complete
			cddb-read-X-complete
			cddb-choice=X
		read
			read-track-TRACKNUM
		encode (implies read)
			encode-track-TRACKNUM
		tag (implies cddb, read, encode)
			tag-track-TRACKNUM
		move (implies cddb, read, encode, tag)
			move-track-TRACKNUM
		playlist (implies cddb)
		clean

		These are to be made user-accessible through the -a
		(action) option.

		A user might specify '-a encode' to quickly read in a CD
		borrowed from a friend on a system without internet access.

		This user could later type '-a move' to finish the process
		once internet access or some other source of CDDB data (such
		as avian carrier or shortwave numbers stations) is reached.

		clean is a special action that is not recorded in the status
		file, its purpose is to delete the directory that contains
		the status file. :)

	cddbquery
		Text file as returned by the CDDB server in response to the
		initial CDDB query. See www.cddb.com for info on the format
		of this file.

	cddbread.X
		Text file as returned by the CDDB server, after conversion
		to Unix linefeeds. See www.cddb.com for info on the format
		of this file.

		cddbdata.1 is the first CDDB lookup, cddbdata.2 is the
		second, etc. This allows for multiple CDDB responses.

	trackTRACKNUM.wav
		WAV data of a track from the cdrom drive.

	trackTRACKNUM.{mp3,ogg}
		MP3 or Ogg Vorbis data after conversion from the
		corresponding .wav file. This file is tagged and then moved
		to its final destination.

	log.commandname.instance
		Log of commands which are run in the background yet have a
		high chance of failure due to misconfiguration, target
		command nonexistance, disks being full, networks down, or
		whatever. Includes:

		wget
		cdrom reader program
		encoder program
		tagger program

		These logs should be human-readable and are not meant to be
		parsed.

		If a command returns a status of 0 it is presumed to have
		run successfully and the log will be deleted. Otherwise the
		contents of the logfile are sent to the user and progress on
		that particular step is halted, or abcde exits fatally if it
		is a critical step (f.e., cd-discid could not be read).

		Clarification on the filenames of the logs:

		The commandname is how abcde calls the executable, whether
		it be via absolute path or whether it depends on the user's
		path (this is abcde's default) with forward slashes replaced
		with underscores. The instance number is the number of times
		this program has been run in this abcde session, starting
		with 1.

		For example, when /usr/local/bin/gogo is run to encode the
		first track, the command's stdout and stderr is logged to
		log._usr_local_bin_gogo.1. gogo's output for the next
		track is logged to log._usr_local_bin_gogo.2.
