 |
 |
 |
 |
MFI version 3.4E was made available 8/23/93. Below I describe a newly
discovered bug and a workaround. This bug is likely also present in
3.4D released 8/17/93. Versions prior to that lacked the 'Runs' feature.
The bug takes effect when enough text is entered into a run description
to fill a full line-width. The result is that the program saves its
configuration (into the file MFI.CFG) in a corrupted form. On the next
call of MFI from DOS, typical symptoms are a configuration with 0 channels
per histogram, 0 dots per dot plot, and 128 cycles of Savitzky-Golay
smoothing at level 0 or 128. (These are all illegal values.) MFI then
fails to work.
The workaround is to avoid using long run descriptions. If you keep your
descriptions to less than 2/3 of the space available, there will be no
problem.
This bug will be fixed in the next version of MFI. Sorry if any of you
have been inconvenienced by it (as we have!).
I would like to hear from any MFI user who finds the memory requirements
of MFI to be an inconvenience. This will help me prioritize future
enhancements.
Also, to those who registered as new users of MFI, if you've had time
to decide whether you find the program useful, and for what, let me
know. Again, this will help prioritize enhancements. If you have
specific suggestions for enhancements, I'll be happy to listen, though
I can make no promises. (Remember, the source code is available for
anyone really ambitious.)
/*- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Eric Martz, Professor of Immunology emartz@microbio.umass.edu
Dept Microbiology Voice: 413-545-2325 FAX: 413-545-1578
Morrill IVN 203, Box 35720, Univ Massachusetts, Amherst MA 01003-5720
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -*/
|
|
|
|