[daip] SPLIT

[Eric Greisen]

Bryan Butler writes:
 > 
 > i have a vague memory that there used to be an option in SPLIT
 > to allow failed solutions to be passed.  but it appears that
 > this option has gone away in recent versions of SPLIT.  is this
 > the case?  
 > 
No there never was a deliberate plan to pass uncalibrated data to the
user when most was calibrated - that was a DEC-10 "feature".  WE did
have a bug that caused failed solutions to be ignored in the middle
and the cal from the last good time to continue to be applied for as
long as it took to get a new good solution - we fixed that bug a year
or more ago.

Eric