[Proj] Libproj4 manual opinion poll

[Gerald I. Evenden]

One thing quickly becomes apparent: the confusion as to what is the difference 
between proj.4 and libproj4.

In terms of cartographic projections, nearly nothing.  In terms of datum 
shifts and other miscellany: everything.  libproj4 as with the original 
proj.4 software had nothing to do with datum shifts other than the original 
distribution had a separate section that dealt only with the NAD datums.  The 
original proj program also had a mechanism for generating Tchebyshev 
polynomials that were designed for use with the mapgen map graphic system.

For the end user, the difference between running proj and lproj is trivial and 
the user can use the original report for proj when using lproj without 
getting into trouble.  There are a few more projections in libproj4 library 
and lproj than the old system.

There have been a few peculiarities introduced into the proj.4 material, some 
I can understand and some that I cannot (to put it very mildly). 

1. Alternate longitude origins.  Are not supported in libproj4 (nor original 
proj).  Some usefulness and a valid suggestion for libproj4.

2. Alternate origins for some projections.  I shudder when I see "two standard 
parallels" listed for plain ol' Mercator.  There is not even one standard 
parallel, let alone two.  There is a latitude of true scale and that only 
adjusts the k_0 factor so that the scale factor is 1. at the selected 
latitude.  If one wants to shift the origin then use x_0, y_0.  Stuff like 
that really raises blood pressure.

I am sure there are others but that is a starter list.
-- 
The whole religious complexion of the modern world is due
to the absence from Jerusalem of a lunatic asylum.
-- Havelock Ellis (1859-1939) British psychologist