README    -- this file

MAKEFILE  -- compile both sources

wdbread.c -- (see source code) converts an EBCDIC WDBII dataset (20 char recs)
	     into two output files: map.index and map.data, while producing
	     a large amount of line and summary detail. The first file provides
	     bounding box and # of vertex information into the second file
	     (map.data), this latter file contains (dx,dy) byte pairs. wdbread
	     will fragment long vectors as needed to provide this format, and
	     will report this on the log it produces on stdout.

	     Multiple runs of this program will place (overwrite, if you're
	     not careful) map/index pairs for the 16 or so volumes which
	     comprise all of WDBII (arranged by continents and detail).
	     All .index and .data files may then be "cat"ed into one giant
	     file which represents the entire world. This is possible by
	     design: the skip offsets in the .index file are relative byte
	     counts in each companion .data file, so each file can be
	     appended to in tandem.
	     
wdbplot.c -- given location and scale produces UNIX PLOT output. The last
	     section of source code provides a simple means to customize
	     the software to direct drive other devices, although we most
	     often use Tektronix output (via plot). Other suggested fixes
	     would be to "trigger" based on vector rank, which changes as
	     a function of political boundary, river, coastline, etc. More
	     ambitious users might wish to encode more general map projections:
	     Mercator and Rectangular equal area would be the most simple
	     to employ because the clipping borders would remain roughly
	     unchanged.

Author's disclamer: I have no intent in modifying or upgrading this ancient
code: if you can make big bucks on it, go for it. If you want to discuss
digital cartography from a less nuts and bolts level, drop me a line sometime!

-- Alan Wm Paeth (awpaeth@watcgl.waterloo.edu, awpaeth@watcgl.waterloo.cdn).
