Re: Disk 18e weirdness

[John McKendry <mckendry@mediaone.net>]

I have a bad 18e also. A few observations:
Paul (I think) wrote: 
> The interesting stuff starts at the IMAGETYP line, as it
> doesnt seem padded to 80 characters.
> 
> IMAGETYP= 'dark    '           / dark frame     RTMP= '8.921381'
> 
> The next bit of weirdness is in the CCDTEMP field which has spurious
> data.
> 
> CCDTEMP = '-23.5^@^@^H^@^@^@^H^@4932'
> 
 What I noticed is that the funniness seems to happen at 2048-byte boundaries.
If you open the file with a plain old text editor (EMACS, e.g.) and tell it
"forward 2048 characters", you end up on the IMAGETYP line, in the blanks following
the comment. Do it again in a well-formed file (not one of the busted ones) and you 
end up in the CCDTEMP line, right at the position where you see the funny ^@^H stuff
in the bad files. Assuming 2K buffers, some step of the copy process failed to copy
the first 276 bytes of the second buffer (part of the IMAGETYP record, all of the
TELESCOP, OBSERVER, and OBSERVAT records, and the first 4 bytes of WATERTMP), and
inserted 8 bytes of binary noise at the beginning of the third buffer. After that
we're in the binary data and it's hard to see exactly what's going on. But the same
pattern applies to all the corrupt files I've looked at. Whatever it is, it's
systematic.
 No idea why it would do that, though.
 If someone can make a good set of 18e files available for ftp, I will have my son
make good copies and distribute them.

John