MAPGEN and GRASS

[Gerald I. Evenden]

>From grass-lists-owner at max.cecer.army.mil Mon May 31 22:08:15 1993
>From: Andrew Hertz <hertz at geog.psu.edu>
>Subject: MAPGEN and GRASS
>Date: 	Mon, 31 May 1993 21:36:03 -0400
>Hi,
>
>I have several MAPGEN/GRASS questions...
>
>I just downloaded and installed GRASS4.1 and I noticed that
>the mapgen and proj versions embedded within GRASS don't 
>seem to match the most up to date stuff on charon.er.usgs.gov.
>GRASS seems to be using mapgen version 4.0 and proj versions 3.1, 4.5, and 4.6.  On charon... is mapgen version 4.1 and proj
>version 4.1.3.  Do you suggest getting the latest from charon
>and using it in src.related.mapgen?  And can you explain the
>three versions of proj in GRASS and what versions 4.5 and 4.6
>correspond to?  

The GRASS release is a snapshot of what is current at the time when
CERL makes up its release.  MAPGEN, consisting of the basic mapgen
directory and graphics subsystem, has been fairly static, so the GRASS
release should be  fairly close to charon's.

Proj, however, has been quite dynamic since the release of 4.x and thus
CERL's release quickly goes out of date.  I expect release 4.2 to come
out late this summer and will, among other updates, include datum
translation procedures.

One other element that confuses charon vs. GRASS, is that charon's
PROJ.4 is designed to be as ANSI (and now ISO, when I get some docs
on this) as possible.  Because of CERL's current commitment to K&R
code, they translate PROJ.4.

>The next set of questions relate to the notorious GRASS/MAPGEN
>makefile interfaces.  I believe I've gotten mapgen built under
>GRASS 4.1, but there's an SCS subdirectory which I believe
>contains additions/upgrades from the Soil Conservation Society
>that is giving me fits (along with lots of their stuff in
>src.contrib).  The only problem I had with make'ing the other
>mapgen stuff is that it was expecting some directories to exist
>that I had to manually create.  The directories were:
>$GISBASE/mapgen/bin
>               /gfonts
>               /interface
>               /man
>               /xgen
>
>Thanks,
>Andrew Hertz
>hertz at geog.psu.edu

Dept. of Agriculture's Soil Conservation Service has done a lot of
"value added" work in interfacing MAPGEN to GRASS.  Largely the
efforts of Marty Holko.  Also I am sure that install procedures
are different because of both SCS and CERL efforts to make them
compatible with general GRASS methods.  Unfortunately, I am not
familiar with the GRASS installation, so cannot be of too much
help in this area but I agree that auto generation of required
directories should be corrected.

As far as charon's ftp anon distribution is concerned, it is for
a non-specific, general case of stand alone installation and operation.
PROJ.4, in particular, is used in many non-GRASS sites.  Thus I
must maintain an installation method that is as simple as possible
and not loaded with extraneous material of a complex GRASS installation
script.  For DOS users, I just distribute the executable with
suggestions for the location of ancillary support files and
environment setup.

Hopefully, this will clear up the reasons for the discrepancy
between CERL and charon distributions.  Unfortunately, it does
not change the fact that differences exist.

Gerald (Jerry) I. Evenden   Internet: gie at charon.er.usgs.gov
voice: (508)563-6766          Postal: P.O. Box 1027
  fax: (508)457-2310                  N.Falmouth, MA 02556-1027