Apache wrote:
> Maybe you're getting the file in ascii mode somehow?

Well, I thought of that immediately, but an ascii
transfer would change the byte count as CR/LF pairs
are mucked with. The byte count matches, though.

The really confusing thing, though, is that if I view
the file with "less", it lists all the directories and
files in the tar archive (as if I had done a "ls -l"
on the directory tree). This tells me that it can at
least partially un-gzip the file, but I never could
manually un-gzip it.

Oh, well, it's all academic now, anyway, as I was able
to get the archive from the mirror unpacked successfully.
When I feel like messing with it, I'll do a hex dump of
both and find out what got corrupted. ("cmp" reports a
diff in the files somewhere along line 162, char # 38xxx,
so there's at least one diff in the two that's causing a
problem.)

-- 
Steve Martin, CPBE CBNT

Post a reply to this message