wishlist of things to add to nget:

- less verbose reporting modes:
    all standard information, but no percentage counters (to include statistics)
    ability to put logging/statistics information to stdout or stderr, and percentage counters to the other
- make --text=mbox filename configurable (relative(to -p) and absolute)

- verbosity levels: (<- superceded below)
    * 0: print nothing
    * 1: print group, directory, and regexp, one per line, for every regexp in
	 the options.
    * 2: as 1, but also print one line 

- format lines: have one for each possible action along the way (mostly).  Command line or .ngetrc file. For now:
    -g, -G, -p, -P, -r, -R, -d
    * then two for every article downloaded (pre and post), two for each article decoded (pre and post), and
    * one auto-update line for interactive update information.
    * newlines are not automatic, "\n" will need to be explicit
    * each status line will use the same nested format style and replacement characters as follows:
    %[<loc>][&<pad>][!][<width>][.<num> ...]<repcode>[{<subformat>[:s/<pat>/<subs>/<flags> ...]}]
    ie: "%p" might substitute the path, "%<-20r" might left-justify the regex used in a field of 20 characters,
	chopping extra characters if the width is exceeded.

	(some codes, such as header size, will not apply to every format line)
	%p	current d/l path
	%P	current temp path
	%r	current regex or expression
	%g	current groupalias
	%G	current groupname
	%d	download flags set
	%T	whether testmode is set
	%t	current time
	%i	number of bytes downloaded for this regexp/expression so far
	%I	total expected for this -r/-R
	%j	# bytes so far for this path
	%J	total expected for this -p
	%k	total number of bytes downloaded this session
	%K	expected total number of bytes to download
	%%	literal "%" character, unless <pad> is specified, in which case <fieldwidth> of those are inserted.

	(per-article codes)
	%B	size of article in bytes
	%l	number of lines in article
	%a	author
	%s	subject
	%n	part number
	%N	maximum part number
	%D	date and time
	%z	article number
	%h{<header>}	the specified header (ie, from, subject, etc.)
	
	(interactive and post-download codes) - includes per-article codes
	%R	download rate
	%b	number of bytes downloaded
	%e	time elapsed
	%E	estimated (total) time until completion (ie, # of seconds)
	%x	transfer start time
	%X	estimated (actual) time of completion

	location codes:
	<	left justify
	|	center justify
	=	outer justify (if the code specifies a field with 2 segments, segments are split
	>	right justify

	chop code:
	-	truncates line on the right if it is too long for the fieldwidth.
		Preserves column integrity.  without it, no truncation will occur.
		
    Or the system could be implemented with a heirarchy of specifiers.
    
    %[<loc>][&<pad>][!][<width>][.<num> ...]<repcode>[{<subformat>[:s/<pat>/<subs>/<flags> ...]}]
    ie: "%p" might substitute the path, "%<-20r" might left-justify the regex used in a field of 20 characters,
	chopping extra characters if the width is exceeded.

	Top hierarchy letter codes:
	p[cdt]	paths: c.urrent, d.ownload, t.emp
	R		current regexp or expression
	g		groupaliases
	G		group spec (dotted)
	c.Tif? :	conditional: T.est code enabled, i.n midfile do check, f.ile exists do check
	s[<unit>].-=afrptsl	file/stream/string size (in unit blocks: characters, Kb, etc.)
	r[<unit>[/<timeunit>]].afrptsl	transfer (d/l) rate (uc: average, lc: current snapshot)
	l[<unit>].afrptsl	lines so far (lowercase) or total expected (uppercase) for s type sub specifiers
	n.pP		number: current p.art, total P.arts
	d.ca		date/time specification: c.urrent, a.article
	t.afrptsl	time duration: same sub-specs as for s
	h.aispdAISPD	header shortcut extractors (uppercase: entire line, lower: header word + ":" excluded):
			a.uthor, article i.d, s.ubject, p.ath, d.ate of post, etc...
	{}		header line extraction code
	//		header match extraction code (any match from within header will trigger inclusion)
	
	<track> is a trackable unit, separated with . for current, = for total, and - for remaining
	    a.rticle
	    f.ile
	    r.egexp same downloaded files combined
	    p.ath same
	    t.emp path same
	    s.ession same
	    l.og same

	<unit> is one of:
	    unspecified (sized to fit field width, but will append abbrev.  unit as specified below)
	    b:	bits
	    B:	bytes
	    k:  1000 bytes
	    K:  kilobytes (1024)
	    m:  10^6
	    M:  2^20 (mega)
	    g:  10^9
	    G:	2^30 (giga)
	    t:	10^12
	    T:	2^40 (tera)
	    
	(some codes, such as header size, will not apply to every format spec)
	%p.c	current base path (where nget was initiated from)
	%p.d	current d/l path
	%p.t	current temp path
	%R	current regex or expression
	%g	current groupalias
	%G	current groupname
	%c.T?<setformat>:<elsformat>	whether testmode is set
	%c.f?:	download flag f set
	%c.i?:	download flag i set
	%d.c	current date/time
	%d.u	date of current group update
	%s.r	total number of bytes downloaded for this regexp/expression so far
	%s.R	total expected for this -r/-R
	%s.p	# bytes so far for this path
	%s.P	total expected for this -p
	%s.s	total number of bytes downloaded this session
	%s.S	expected total number of bytes to download
	%%	literal "%" character, unless <pad> is specified, in which case <fieldwidth> of those are inserted.

	(per-article codes)
	%s.A	(expected) size of article in bytes
	%l.A	(expected) number of lines in article
	%h.a	author
	%h.s	subject
	%h.i	message id
	%h.p	post path
	%{<header>}	the specified header (ie, from, subject, etc.)
	%n.p	part number
	%n.P	maximum part number
	%n.o	[x/X] part display (part x of X total)
	%n.a	article number
	%d.p	date and time of article postage
	
	(interactive and post-download codes) - includes per-article codes
	%r.t	(pseudo-)current transfer rate
	%r.T	average transfer rate
	%s.a	total number of bytes downloaded so far
	%l.a	number of lines downloaded so far
	%t.a	time elapsed
	%t.A	estimated (total) time until completion (ie, # of seconds)
	%d.p	transfer start time
	%d.P	estimated (actual) time of completion

	location codes:
	<	left justify
	|	center justify
	=	outer justify (if the code specifies a field with 2 segments, segments are split
	>	right justify

	chop codes:
	-	truncates line on the right if it is too long for the fieldwidth.
		Preserves column integrity.  without it, no truncation will occur.
		On two segment fields, truncates at division
	*	chops an extra character to insert a * to show truncation.
	.	chops 3 extra characters to insert a ... if to long.
	+	inserts width worth of + if line to long, fortran style.  Not recommended