Project 
======= 

Make lunar distance tables for a given period. For every date the
tables contain the true geocentric lunar distances to selected bodies
(Sun and suitable set of planets or stars). The bodies are selected
subject to 1: the body (other then the sun) is far enough from the sun
to be visible 'somewhere on earth'; 2: the lunar distance is less than
about 120 deg, and 3: the angular velocity of the moon is greatest
with respect to these bodies, adding to accuracy. Lunar distances are
given for every 3 hours of UT. No tabulation if the moon is too close
to the sun for observation.

Method
======
Items marked $ are quantified below.
For every date, do the following:
1. If date is $too close$ to New Moon, skip this date.
	 Implementation: This can be done in advance. Using Emacs function
	 'phases-of-moon', make a table of New Moon instances in UT in or
	 near the intended range of data. List the dates to skip according
	 to the rule below. Pass this list to the program. The program
	 checks every date against this list.
2. Calculate a table of true geocentric lunar positions for every 3
   hours UT.
	 Implementation: Let 'aa' do this. Use 'awk' to extract a table
	 containing:
		hour, decimal RA, decimal Dec.
3. Calculate approx geocentric positions at 12 UT for all
   $bright$ bodies $near enough$ to the ecliptic and $far enough$ from
   the sun. 
	 Implementation: Let 'aa' calculate the positions for Sun and
	 Planets. Take star positions from file adapted from 'star.cat' in
	 the aa directory (ie bogus entries and Dec>60 deleted). Discard
	 Venus if its RA differs less than 1 hour from the sun; discard
	 other planets and stars below 3 hours. List for the remaining
	 bodies their aa-number, decimal RA, decimal Dec.
4. Tabulate the lunar distances to these bodies at 12UT, 15UT,
   and the difference in 3 hours. These are approx values only used
   for decision of the bodies to include. Discard all entries with
   distance > 120deg.

	 Implementation: For each of the two lunar positions and for each
	 listed body, evaluate 
		distance = arccos ( sin Moon.Dec sin S.Dec 
			              + cos Moon.Dec cos S.Dec cos Delta RA )
     Make a list of aa-number, distance at 12 UT, diff. dist., EW; the 
     last field to indicate of Moon is East or West of body.
5. Select 5 entries: (4+sun if its dist<120 or 5 ex sun) with largest
   differences from the table. Check that we have at least one
   easterly and one westerly distance.
	 Implementation: If the Sun is still listed, select it. Sort the
	 remaining entries in the distance list on the
	 diff.dist. field. Select the first entry marked E and the first
	 entry marked W. Complete selection to 5 bodies from the top of
	 the list.
6. For these bodies, calculate the exact lunar distances for the 
   hours (00, 03, 06, 09, 12, 15, 18, 21, 24). Also calculate the
   prop.logs (we need the 24UT calculation for the last prop.log).
	  Implementation: For each body in the selection, do the
	  following: 
	  -if the body is a star, let 'aa' and 'awk' get the true
	  (precessed etc) coordinates for 12UT. Calculate the lunar
	  distances using the stored table of lunar positions and the
	  distance formula mentioned above.
	  -for Sun or Planet, let 'aa' and 'awk' calculate a full set of 
	  coordinates for each of the 9 specified times. Proceed as above.
	  -calculate the prop.log = log 10800 - log |dist(t) - dist(t+3)|
	  for t=0(3)21.
	  -store results in format time, distance, prop.log.
7. Format the table of lunar distances and prop.logs in LaTeX.
		date | Body1  PL | Body2  PL | ...
		---------------------------------------------------
		00UT |           |           |
		03   |           |           |
		06   |           |           |
		09   |           |           |
		12   |           |           |
		...  |           |           |
		21   |           |           |
		----------------------------------------------------

		date | Body1  PL | Body2  PL | ...
		---------------------------------------------------
		00UT |           |           |
		03   |           |           |
		06   |           |           |
		09   |           |           |
		12   |           |           |
		...
	  


Quantification
==============
Too close to New Moon
  reference: http://www.chowk.com/bin/showa.cgi?kshaukat_sep2397
  The moon will not be seen unless at least 10-12 deg from the sun at
  sunset/sunrise.  Quite arbitrarily, we'll omit tabulation if the
  New Moon occurs within 24 hours from 12UT of the date in
  question. Thus, skip the date of New Moon and if NM occurs before
  12 UT, skip the previous date; else skip the next date.

Bright bodies
  Selected Stars, Venus, Mars, Jupiter, Saturnus

Near the ecliptic
  Keep it simple: |Dec| < 60. Then, we might aswell take just all
  Selected Stars.

Far from the Sun
  Either RA differs more than 3 hours (Venus:1 hour) or Dec differs
  more than 45 deg. But if Dec differs so much, the star is too far
  from the ecliptic anyway. This is an intuitive guess.