Wolff proposed the following pipeline procedures and tested them out:
perl scripts call C programs that will call Fortran programs as subroutines.
While we will do most of our development in IDL, most of us will write our
final programs in Fortran.   If we can find a way to call Fortran programs
directly from perl, we will bypass the step of using C programs.  
Wolff tested out a package called SWIG (www.swig.org), a software development tool 
that will connect C programs with perl.

Wolfire downloaded and installed COBE analysis software. He created
Mollweide projection maps of DIRBE data for 8/22/1990 for DIRBE
bands at 3.5, 4.9, and 12 microns and made b = 0 crosscuts. The 
in-plane Galactic emission at l = 10, b= 0 deg is approximately
10 MJy/sr, 10 MJy/sr, and 55 MJy/sr at 3.5, 4.9, and 12 micron
respectively. He also used SPOT to obtain cirrus estimates 
(which are based on COBE 100 um maps and a model cirrus emission) 
to find
               3.5 um   4.9 um    8 um   12 um     
I_nu (MJy/sr)   2.62    3.6     75.6     56.0 

He also met with the remaining DIRBE team at Goddard (Dave Leisawitz,
Nils Odegard, and Tom Sodroski) to help understand differences between
the Kelsall et al Zodi model and Ned Wright's. Dave feels that they
do not have authorization to pass on to us the original Kelasll IDL
code, but will check to see that the numbers returned from SPOT
agree with the original code.

Whitney & Meade created test data.  Meade read two documents from the IRAC
working group website that were very useful.  These are the "Science Requirements 
for the IRAC Pipeline." and the SPIE paper by Hora et al., "Calibration and 
Performance of IRAC".  From these, she defined her understanding of the
pipeline products.  This is summarized on the GLIMPSE data processing web page 
(http://www.astro.wisc.edu/sirtf/Team/Processing/) in a document called "Test Data 
Specifications".  Whitney & Meade wrote an IDL program to read in some HST WFPC2 images
of the R136 region and create 8 test images per IRAC band.  These include  
a BCD file that has cosmic rays, dead pixels, saturated pixels and banding effects;
a photometric error image; bad pixel mask; data quality mask; ghost image; cosmic
ray mask; latent image mask; and latent image.  The program shifts and rotates
the orginal image to mimic revisits.  It will also create a set of files for
each of the four IRAC bands. We will post these images on the web this week.

Babler continued testing photometry packages, and went on vacation over the
UW spring break.