All of the current logged oven data is stored in FITS files (older oven data is stored in "old IRAF format" image files). Some of the really old oven data may have endian issues. It is stored big endian (as it was written on a SPARC machine), but truly old IRAF had no mechanism to manage endian issues. The data from the MMT casting may be from this era. This is nothing we intend to deal with here, but it worth mentioning in passing. In fact from this point on, we will only discuss FITS files. A final comment is that the recommendation was made to use IRAF running on a SPARC machine to convert any old image format files to FITS. This would be a bit of a job and has not been tackled. Very likely by the time anyone is reading this, those machines will be long gone and some old files will be in hand someone wants to read.
Python has a number of packages to read FITS files. Something called "PyFITS" exists, probably written at Space Telescope. It is deprecated, being replaced by astropy.io.fits. There is also a package for python3 called "fitsio" that is available as a fedora package. It is a perfectly good alternative. I am inclined to use astropy.io.fits because my expert advisor Tim says that it is what he has used without problems for over 10 years. The following gives an excellent discussion of how these various python FITS packages relate to each other:
There is something called Astropython, but it is best ignored. It has been static for 5 years or so. Astropy is under active development.However, I can just use the command line "xpaset" to send things to ds9 via XPA. This takes data on stdin, making it easy to pipe to it. This solves my problem (but ximtool is clueless about XPA, so this locks things to DS9, which is not a bad thing necessarily.
XPA stands for "X public access".