Below we give more details on changes that are likely to be noticed by users in their data.

Setting data quality flags (DQ) in the occulting bar region The locations of the occulting bars in 2-D rectified long-slit data are flagged with a data quality value of 8 in the DQ extension. With MAMA data, it is possible for an occulting bar to fall on the lower or upper edge of the 2-D rectified image, in which case calstis was setting the DQ flags out of bounds. Unless this caused calstis to actually crash, it would be difficult for the user to notice any effect of this bug, since the out-of-bounds area that was incorrectly flagged is in a buffer zone, normally outside the region that contains actual data.

Indexing error for binned data fixed When extracting a 1-D spectrum from CCD data that are binned in the cross-dispersion direction, there was an error in the computation of the exact placement and size of the extraction box. The error affected the extracted values and could cause calstis to fail due to access to unallocated memory. The error did not affect extractions from non-binned data. The error is now fixed.

CR-Rejection is omitted if the No. of iterations exceeds 30 If the number of image sets is more than 30, calstis is set so that it does not do cosmic-ray rejection. Such large repeats are normally done ONLY to look for variability, so it is not useful (nor practical) to do a CR-rejection.